►
From YouTube: Optimism’s OP Stack by Karl Floersch | Devcon Bogotá
Description
Visit the https://archive.devcon.org/ to gain access to the entire library of Devcon talks with the ease of filtering, playlists, personalized suggestions, decentralized access on Swarm, IPFS and more.
https://archive.devcon.org/archive/watch/6/optimisms-op-stack/
Introducing the OP Stack — a standardized tech stack for L2 chains.
Introducing the Superchain vision — the inevitable unification of the distinct L2 chains into a single, horizontally scalable, super-sequenced, Superchain!
The OP Stack is how we realize the Superchain vision.
Speaker(s): Karl Floersch
Skill level: Intermediate
Track: Layer 2s
Keywords: scalability,multi-chain,superchain
Follow us: https://twitter.com/efdevcon, https://twitter.com/ethereum
Learn more about devcon: https://www.devcon.org/
Learn more about ethereum: https://ethereum.org/
Devcon is the Ethereum conference for developers, researchers, thinkers, and makers.
Devcon 6 was held in Bogotá, Colombia on Oct 11 - 14, 2022.
Devcon is organized and presented by the Ethereum Foundation, with the support of our sponsors. To find out more, please visit https://ethereum.foundation/
A
A
Why
hello
wow
that's
so
much
easier
to
be
loud?
Oh,
my
goodness!
Okay,
so
before
we
have
any
slides
I'm
just
going
to
tell
you
a
little
about
me,
I'm
Carl,
I'm,
a
contributor
to
the
optimism,
Collective
I
work
at
op
labs
and
today
we're
going
to
talk
about
the
op
stack.
It
is
op
and
the
super
chain
which
you
will
hear
once
and
you
will
never
forget,
because
it
is
a
sticky
meme,
sticky,
meme
now
well,
my
slides
have
arrived
just
in
time
like
Uber,
Eats,
okay,.
A
Okay,
this
is
gonna,
be
a
fun
talk.
Let
me
tell
you,
let
me
tell
you
so,
as
usual,
I
have
a
bajillion
bajillion
slides,
so
in
2018
we
set
out
I
use.
We
very
liberally
we
set
out
to
build
plasma
group
plasma
group
was
a
non-profit
focused
on
scaling
ethereum
with
layer,
2
technology.
We
wanted
to
unlock
tons
of
applications,
but
jillions
of
transactions
per
second,
and
so
you
know
we
sailed
out
on
ethereum
cyberspace,
following
vitalik's
Doctrine.
A
Following
his
advice,
we
knew
what
to
build,
and
so
we
went
and
in
fact
we
did
some
cool
stuff.
Last
Devcon
we
talked,
we
introduced
optimistic
roll-up
and
we
introduced
the
unipig,
which
was
like
the
first
implementation,
and
it
was
pretty
sweet.
If
you
don't
ask,
if
you
do
ask
me
so
ethereum,
we
did
good
things,
good
things
for
ethereum
public
goods.
This
was
open
source
software.
It
already
got
forked
immediately,
but
guess
what
happened?
A
We
could
not
get
funding
for
it
because
we're
a
non-profit
and
ethereum
did
not
have
the
public
goods
funding
infrastructure
that
we
needed
to
support
the
building
that
we
were
doing
to
for
the
public.
Good,
no
bueno,
no
bueno
at
all.
Anyway,
until
we
formed
optimism
and
we
set
out
to
not
only
scale
ethereum
but
also
fund
public
goods
fund,
ethereum,
public
goods,
optimism,
public
goods,
build
the
community,
push
the
community
forward
and
so
that
we
did
not
realize
it.
A
In
my
humble
opinion,
maybe
maybe
you
know
trillions
if
we
go
multi-planetary
billions
of
AI
really
more
than
that
trillions
of
transactions
actually
weigh
more
than
trillions.
That's
for
sure
all
of
these
things
promoting
humans
over
capital
and
impact
making
impact
profitable
right.
We
need
to
be
expressing
what
we
want
as
impact
and
make
that
profitable.
That's
public
goods
funding
really
at
the
end
of
the
day,
so
optimism
Collective.
That's
why
we're
here.
But
that
is
not
the
story:
let's
go
Leroy
Jenkins
the
rough
start,
so
this
is
the
whole
journey.
A
whole
journey.
A
Rough
start
right.
We
formed
some,
you
know
terraformed
some
land
and
ethereum
cyberspace
and
you
know
created
our
little
your
little
community
and
built
an
ethereum
Layer
Two
that
was
evm
compatible
and
you
know
it
was
pretty
cool.
But,
to
be
honest,
it
was
not
easy.
Not
at
all.
We
had
built
a
roll-up,
we
had
built
a
fault
proof,
but
it
was
complicated.
It
was
hard
to
maintain.
It
was
difficult
to
onboard
new
people.
This
was
not
the
you
know,
easy
sailing
that
we
thought
we
were
gonna
do
we're
like.
A
A
We
happen
to
be
eating
dinner
and
we
saw
in
the
back
corner
of
our
eye.
George
hotz
literally
the
you
know,
founder
of
comma
AI
builder,
of
really
really
good,
open
source
production
software
and
we're
like
George
come
through
and
it
turns
out.
He
knows
ethereum
and
like
new
ethereum,
Layer
Two
is
pretty
insane,
so
you
know
George
on
his
little.
You
know
Journey
kind
of
stopped
by
our
little
island
said
you
know.
What's
up
and
he
looked
at
our
code
and
I'm
not
gonna
lie.
He
gave
us
some.
A
You
know
it's
a
hard
pill
to
swallow.
He
said
one.
He
said
you
have
a
hundred
times
the
lines
of
code
that
you
actually
need
for
this
two
every
line
of
code
should
be
worth
at
least
a
hundred
dollars.
That
means
the
cost
of
maintaining
it.
The
cost
of
people
learning
it.
It
is
expensive
to
write
lines
of
code,
even
a
comment
and
finally,
the
goal.
Your
goal
should
be
to
Upstream
to
ethereum
and
build
software.
A
A
That
was
the
beginning
of
the
formation
of
our
North
Star
towards
Simplicity
and
standardization,
and
so
of
course,
as
jiohad
does
he
he
left
and
then
chapter
two.
We
finally
had
our
North
Star,
and
so
we
could
start
measuring
progress
towards
it.
We
were
going
to
boil
our
code
down
to
its
Essence.
We
were
going
to
reduce
our
death
diff
to
less
than
1
000
lines
of
code.
A
That
was
our
basically
our
whole
road
map
at
the
time,
and
so
we
went
to
coding
our
little
gerbil
selves.
You
know
all
these
are
durables
well,
they're,
really
hamsters,
because
hamsters
actually
are
cuter
than
gerbils
anyway,
I'm,
sorry,
gerbils.
Okay,
we
started
making
improvements
to
our
L2
Island
right.
We
terraformed
the
land
further
it
really.
We
like
started
making
real
progress,
simplifying
our
code.
A
It
was
easier
and
we
became
fat
gerbils
off
of
our
great
successes,
and
so
we
had
made
some
serious
progress
and
in
2021
Q4
we
released
our
evm
equivalence
upgrade
and
that
thing
has
served
us
well.
We
have
been
evm
equivalent
for
basically
a
year
now
and
fast
as
ethereum,
and
it
has
been
nice.
It
was
a
beefy
release,
it
was
fantastic,
but
our
job
was
not
over.
We
still
hadn't
finished
because
Joe
hot
came
back,
he
just
popped
in
and
he
said
all
right.
Let
me
tell
you
something
else.
A
Well
done,
you
got
rid
of
a
lot
of
your
code
that
must
have
hurt.
That
is
hard,
but
your
proof
system
is
way
too
complicated.
Still
you're
going
you're
going
off
the
deep
end
here,
I
can
help
and
so
what
he
did
well,
he
contributed
Canon.
Just
this
really
simple
proof
system,
and
why
did
it?
Why
was
it
so
simple?
Well,
it
was
because
it
separated
the
client
software
from
the
proof,
software
from
the
proof
layer.
A
You
compile
the
client
down
to
the
proof
and
in
fact
vitalik
at
the
same
time
was
suggesting
the
same
kind
of
architecture,
but
George
actually
took
that
and
created
a
you
know,
pretty
pretty
solid
version
that
people
have
been
playing
with
ever
since,
and
that
showed
us
something
else
that
we
hadn't
thought
about
before
that
showed
us
the
power
of
modularity.
The
fact
that
he
was
able
to
achieve
this
evm
equivalence
was
because
he
split
it
out
into
two
different
components,
and
so
right,
evm
equivalent
fall
proof
less
than
4
000
lines
of
code.
A
The
client
compiles
to
the
proof,
and
it
showed
us
that
power
now
we
had
a
North
star,
but
for
our
goal
of
reducing
our
get
diff
down
to
less
than
a
thousand
lines.
Now
we
had
a
path
to
achieving
it,
and
so
we
also
at
the
same
time,
happen
to
be
listening
to
some
nice
like
streaming
episodes-
and
this
is
one
of
my
favorite
moments
in
in
like
every
podcast
that
I've
listened
to.
Basically,
it's
just
this
line
is
when
things
are
beautifully
done.
Usually,
there's
a
well
thought
out
set
of
abstraction
layers.
A
It
has
the
thinking
about
the
problem,
space
slicing
it
up
and
then
going
at
it
that
mentality.
It
changes
the
game
for
software,
development
and
architecture.
So
you
know
he's
off
again
and
next
up
the
last
Mile
right.
We
had
just
made
this
10x
line
of
code
reduction
we,
but
we
still
had
5x
too
many
lines
of
code
and
get
it
was
crazy
and
we
still
had
10x
too
many
overall.
A
So
what
did
we
do
well
turns
out
that
we
started
coding
and
we
noticed
that
eth2
was
coming
to
the
same
realizations
that
we
were
having
as
well.
They
realized
that
they
needed
to
modularize
the
E2
clients
and
make
them
talk
to
each
other
via
the
engine
API.
They
realized
that
the
currently
it
was
too
tightly
coupled
too
monolithic,
and
so
what
did
they
do?
They
took
the
client
software
and
they
split
it
further
into
consensus
and
execution.
That's
why
we
have
these
consensus,
clients
and
execution
clients
so
well.
We
saw
this
eth2
consensus.
A
We
were
like
we
could
just
write
a
roll-up
consensus,
client
and
boom
slot
it
in
in
the
same
way,
and
now
we
can
reuse
all
of
these
other
execution.
Clients
like
Geth,
like
Aragon,
like
Etc,
there's
so
many,
and
this
gives
us
that
multi-client
ecosystem
that
is
absolutely
critical
for
securing
a
chain
which
has
billions
of
dollars
we
need.
A
We
cannot
stop
with
one
client
unless
you're
going
to
formally
prove
your
client
and
good
luck
and
anyway,
formal
proofs
are
like
two
clients
anyway,
okay
point
being,
we
finally
had
our
golden
ticket
to
a
less
than
one
thousand
line
of
code
diff
in
Geth,
and
so
we
got
to
building
in
2022
to
do
terraforming,
building
cool
things
and
boom
in
our
Bedrock
release
our
upcoming
Bedrock
release.
We
finally
have
that
commit
that
single
commit
that
turns
up.
You
know
normal
death
into
opgath
with
less
than
a
thousand
lines
of
code.
A
So,
let's
you
know
8
21,
less
than
a
thousand.
We
achieved
our
goal.
We
got
to
have
a
little
hamster
celebration,
gerbil
celebration:
we
include
all
rodents
in
this
talk
the
Bedrock
release.
Finally,
we
had
created
this
code
in
a
simple
modular
way:
we
split
up
consensus,
execution
and
the
proofs
and
we
renamed
the
proofs
to
settlement
because
it's
more
Mimi
and
boom.
We
had
achieved
that
goal
of
Simplicity
and
standards,
but
then
something
unexpected
started
happening.
A
I
really
didn't
I.
Did
this
from
memory?
I
have
not
refreshed
the
actual
sound
I
just
know
the
dirt
sound
anyway
Rattata,
okay,
no,
we
started
getting
forked
literally
all
over
the
place.
It
is
absolutely
mind-boggling
right.
We
multiple
people
set
out,
took
our
code
base,
forked
it
set
up
their
own
settlements
and
further
it
got
crazier.
A
Some
were
modifying
evm,
we're
modifying
to
evm
and
the
execution
layer
in
Geth
and
some
were
modifying
the
actual
roll-up
part
of
it
to
make
it
a
plasma
so
that
you
don't
post
the
data
from
like
all
of
these
forks
and
I'm.
Not
talking
about,
like
you
know,
Pet
Project,
Forks
I'm,
talking
about
like
Mega,
you
know
you
look
at
l2b,
you
got
the
little
Ops
shout
out.
We
got
some
Forks
in
the
room,
so
we
had
set
out.
We
had
created
the
consensus
layer,
execution,
layer,
split
and
then
things
got
crazy.
A
Guess
what
next
guess?
What
next?
Well,
if
you
were
at
a
talk
like
five
hours
ago,
you
would
know
what
next
someone
literally
took
the
code
base
for
like
forked
it
and
made
a
three-day
Survival
game
that
is
fully
on
chain,
literally
playing
something
that
is
unrecognizable
brand
new
software
and
and
literally
every
single
block
like
like
I
just
gotta
momentarily
like
we're
not
talking
about
like
oh
yeah.
A
You
know
Pokemon
roll
up,
it
doesn't
make
sense.
It's
like
ridiculous.
What
you
can
really
do
with
this
code.
This
guy
really
is
the
limit
right.
You
have
the
rust
program.
They
turned
Canon
into
rust.
All
of
these
things
because
we
were
Building
open
source.
We
were
building
foundational
modules
that
stand
alone
and
are
elegantly
built.
A
So
very
we
got
all
these
new
Forks,
all
these
new
little
Islands
out
in
ethereum
L2,
and
then
our
North
Star
of
Simplicity
started
bearing
fruit
great
exciting.
So
and
thus
this
begins
a
new
era.
It
creates.
It
was
time
for
Branding,
because
our
code
base
had
no
name.
It
was
just
the
code
base.
You
know
it's
like
that's
kind
of
boring,
so
we're
like
hey.
Let's
turn
this:
let's
flip
it,
let's
make
it
cool,
let's
make
it
fun,
let's
call
it
the
op
stack.
A
A
It's
our
open
source
software
stack
designed
to
power,
blockchain
ecosystems
built
for
the
optimism
Collective,
so
that
includes
that
includes,
of
course,
what
we've
been
talking
about
this
whole
time,
the
whole
chain,
software
right,
the
decentralized
stack
required
for
spinning
up
chains,
the
consensus
layer,
execution
layer,
more
layers
to
come
literally.
This
is
just
the
beginning.
Additionally,
it
also
requires
a
governance
layer,
things
that
we're
building
such
as
the
identity
layer,
the
retroactive
public
goods
funding.
A
All
of
these
things
need
to
be
built
in
a
modular
way
so
that
they
can
be
reused
and
all
on
top
of
ethereum.
Of
course,
where
are
we
right
now
come
on?
What's
on
my
shirt?
Now
we
are
in
Devcon.
Actually
that
was
a
great
transition.
I
did
not
mean
to
make.
We
are
in
Devcon
in
in
case
you
didn't
know
talking
about
the
op
stack,
and
so
this
is
kind
of
the
time
that
we're
just
really
talking
about
it,
and
that
means
that
this
story
is
bridging
into
the
future.
A
Okay,
now
we're
in
the
future
gal,
okay,
so
everyone
in
this
future
in
our
future,
everyone's
future,
contributing
back
to
the
stack
right
with
shared
modules.
This
is
the
lesson.
I
literally
came
from
a
talk
on.
You
know,
practical
pluralism.
Everyone
is
able
to
share
bass
players
share
modules,
that's
how
Innovation
happens
in
this
space
and
we
can
all
benefit.
We
can
literally
push
the
ecosystem
forward,
and
so
there's
lots
coming
down
the
pipeline
literally
so
much.
We
have
this
really
nice
code
base
that
we're
releasing
very
soon
with
bedrock.
It's
super
hype.
A
It's
clean,
it's
like
the
cleanest,
but
there
are
things
that
we
can
make
improvements
to,
such
as
introducing
multi-client
fault
proofs
such
as
introducing
multi-client,
ZK,
proofs
and
critically.
This
means
that
deploying
a
you
know
deploying
one
of
these
op
chains
we're
calling
them.
It
doesn't
tie
you
to
a
single
proof
system.
It
doesn't
tie
you
to
a
single
technology.
You
can
have
a
ZK
roll
up
with
a
seamless
upgrade
so
that
no
one
even
has
to
notice
you're
just
lowering
the
withdrawal
times
in
the
background
right.
A
These
kinds
of
things
happen
because
you
have
clean
apis
where
you
can
separate
and
plug
things
in
and
additionally,
we
got.
You
know,
plasma
consensus
for
some
ZK
plasma
for
super
scalable
op
chains,
and
you
know
we
got
a
Pokemon
execution
apparently
for
ZK
Pokemon
plasma
Bitcoin,
you
know
whatever
you
know
the
sky's
the
limit
really,
so
everyone
can
benefit
from
these
shared
modules,
really
nice,
so
one
person
creates
a
little
Greek
statue,
which
is,
of
course,
a
great
contribution
to
the
op
stack
and
boom.
Now
it's
an
OPI
second
boom.
A
Everyone
can
adopt
it,
no
problemo,
it's
great
and
by
the
way
and
retroactive
public
goods
funding
right
we're
not
doing
this
and
the
whole.
The
whole
reason
why
we
got
into
this
thing
was
because
we
wanted
to
contribute
back
to
the
open
source
software,
the
stack
that
and
then
make
sure
that
we
would
get
paid,
even
though
we
were
building
public
goods,
and
so
of
course
you
know
people
got
to
get
paid
for
their
their
great
Greek
statues.
So
with
everyone
contributing
everyone
gets
paid
everyone's
happy.
Let's
go
plasma
groups.
A
A
Experimentation
cannot
should
not
must
not
equal
fragmentation
at
the
same
time
as
we're
experimenting
we
need
to
unify.
We
need
to
bring
the
user
experience.
The
developer
experience
everything
together,
because
it's
better
right.
So
that
means
that
we're
standardizing
op-chain
security
right
when
you
reuse
these
modules.
You
don't
have
all
these
crazy,
Bridge
hacks
literally
every
week,
because
everyone
is
running
the
same
software
bending
fitting
from
the
same
Lindy
effect
of
that
software.
It's
all
you
know
proven
formally
proven
multi-client
whatever
it
is.
A
We
need
to
standardize
because
it
makes
everyone
stronger
additionally
standardizing
off
chain
tooling.
We
noticed
this
with
the
evm
equivalence
upgrade
everyone
got
to
use
their
tooling
from
before
and
standardizing
option
governance,
everything
from
building
governance
modules
to
even
introducing
people
into
the
collective
through
an
identity
layer
so
that
we
can
have
democratically
run
governance.
All
of
those
things
are
absolutely
critical
for
people
for
us
to
share
as
a
community,
and-
and
it's
important
to
note
that,
just
because
you
have
two
different
options-
does
not
mean
that
they
cannot
communicate
atomically
with
one
another
right.
A
You
can
send
a
transaction
to
One
op
chain
and
have
a
flash
loan
Atomic
Swap.
This
is
something
that
is
very
possible
and
it
is
enabled
when
everyone
is
running
the
same
software
and
even
shares
a
sequencing
layer.
So
no
bridging
who
bridging
bridging
is
Lane
I,
don't
like
bridging
I
like
doing
things.
That's
not
a
doing
thing,
that's
a
doing
thing
to
do
a
thing
anyway,
very,
very!
That's
how
we
design
things
I
mean
really.
It
is
actually
in
some
sense.
Okay,
the
op
sack
is
not
about
pushing
people
apart.
A
It's
about
bringing
people
together,
so
we
can
all
have
a
gerbil
party
together
and
so
talking
too
loud.
Now
the
final
evolution
right,
we're
doing
all
these
standards.
We've
got
all
this
composability
we're
getting
to
the
point
where,
like
the
software
stack,
is
maturing,
it
is
this
one
thing
we're
all
contributing
to
what
does
that
mean?
That
means
that
we
are
starting
to
form
a
superstructure
not
just
defined
by
this
or
that
you
know,
opt
chain
doing
this
thing.
A
This
saw,
you
know
we're
actually
bringing
people
together
with
the
network
effects
of
cooperation,
to
form
something
that
is
entirely
new,
and
what
is
that
thing?
Well,
if
you
remember
the
beginning
of
the
talk,
you
might
start
to
find
out
as
the
islands
configure,
the
super
chin
we're
back
baby
the
super
chain,
I
I
told
you
I
was
gonna,
do
it.
That
would
be
weird
if
I
didn't
anyway,
the
super
chain
birthed
through
the
op
stack
through
these
standards
through
the
shared
compatibility,
the
atomic
composition
shared
sequencer
set.
A
We
start
seeing
this
thing:
the
replacement
of
the
web
2
infrastack
with
decentralized
Alternatives
that
can
power
horizontally,
scalable
applications
like
we're
talking
about
the
internet's
computation-
and
this
is
not
multi-chain
right
options
are
not
app
chains.
Everyone
is
a
part
of
the
collective
we're
all
part
of
this
network.
Additionally,
this
is
not
the
mono
chain
right
because
at
the
same
time
as
we're
all
interconnected,
we're
also
not
you
know
the
same
right.
There
is
still
multiplicity,
there's
still
biodiversity
and
it's
horizontally
scalable,
so
instead
it's
all
connected
by
the
same
Technical
and
social
fabric.
A
It
is
not
the
multi-chain.
It's
not
the
model
chain.
Tell
your
friends
y'all!
It's
a
super
chain:
let's
go
but
wait,
but
wait.
I
just
went
Ham
on
some
tech.
Let
me
just
real
quick,
bring
it
back
to
why
we
are
here
because
I
forget
that
literally
I
get
so
into
the
tech.
It's
so
much
fun
anyway,
but
we
are
here
because
we
are
forming
new
technological
social
institutions.
We
are
forming
the
optimism
Collective.
Why
we
want
humans
to
be
valued
over
Capital.
We
want
identity,
citizens
Etc.
We
want
impact
to
be
profitable.
A
We
want
to
build
the
optimism
Collective
and
just
contextualizing
this.
Why
now
right?
Why
is
this
so
critical?
Now?
It's
because
we
are
entering
a
new
era
in
the
internet
and
as
a
species
period,
because
I'm
gonna
tell
you
right
now:
I'm
a
species
in
the
internet.
Right
now,
my
life
is
crazy.
So
today
we
are
living
in
a
world
where
profit
maximizing
corporations.
They
are
the
ones
that
are
operating
our
our
social
public
infrastructure
right,
I
can't
post
on
the
internet
without
going
through
a
privately
owned,
Corporation.
Also
we're
so
literally.
A
We
are
confronted
daily
with
global
coordination,
failures
from
conflict
to
biological
stuff,
like
it
is
getting
terrifying
and
we're
doing
this
in
this
crazy,
Melting,
Pot,
World,
very
misaligned,
all
while
we're
about
to
birth,
a
super
intelligence,
that's
going
to
look
at
our
economy
and
be
like
oh
yeah,
so
you
the
way
that
the
things
that
we
reward
in
this
system
is
extraction,
and
you
know
bad
behavior
right
that
this
is
terrifying
to
me.
Our
future
is
not
guaranteed.
We
need
to
build
this
thing.
A
A
So
that's
why
we're
building
the
op
stack
as
an
open
source
public
good
right,
we're
funding
public
goods
retroactively.
We
need
to
be
doing
this
on
a
massive
scale.
Our
social
institutions
for
funding
public
goods
are
not
adequate
and
we
need
to
be
providing
people
identities.
So
we
measure
human
utility,
not
just
you
know,
shareholder
profits,
so
we
can
maximize
impact
and
give
profit
to
people
and
that's
why
we're
building
the
optimism
Collective
and
that's
because
together
we
can
make
impact
equal
profit
and
summon
ethers
Phoenix.
A
It's
like
it's
like
the
it's
like
the
rationalist
god
that
will
like
give
you
money
for
making
it
come
to
be,
and
you
know
you're
You're,
Building,
public
goods
and
you're.
You
know
being
a
good
person
anyway.
It's
just
like
our
the
pull
towards
Utopia,
our
better
selves,
ethers,
Phoenix.
So
that's
how
right
we
started
out
as
building
out
we
built
out
this
op
stack,
because
we
realize
that's
how
we
push
open
source
forward.