►
Description
Moderated by Owen Fernau of The Defiant, this Graph Day 2022 panel was joined by Rahul Sethuram of Connext, Sara Reynolds of Uniswap Labs, Eric Tang of Livepeer, Roneil Rumburg of Audius, and Mike Sall of Goldfinch, as they engaged in a curated conversation regarding the way in which web3 protocols are transforming the internet. Panelists explored questions such as: What is a protocol? Is a token fully necessary in order to incentivize continuous development? Should protocols be developed based on governance or or should the code never be altered? Are there things we should do differently regarding our communication with stakeholders and deployment of tokens?
Watch more Graph Day 2022 talks, panels, and announcements here: https://www.youtube.com/playlist?list=PLTqyKgxaGF3SWt8ipQw-Ul9HQvmxulbUu
Keep up with The Graph on social media!
Time Stamps
0:00 Introductions
1:23 What is a protocol?
4:20 Do you need a token or a protocol to push the project forward
9:27 Upgradability
16:28 What have learned about Protocol development?
Twitter: https://twitter.com/graphprotocol
Instagram: https://instagram.com/graphprotocol
LinkedIn: https://www.linkedin.com/company/thegraph
GitHub: https://github.com/graphprotocol
Website: https://thegraph.com
A
B
C
A
B
A
protocol
is
a
standard
way
for
a
set
of
actors
to
coordinate
around
a
shared
set
of
resources
or
a
system,
and
it's
important
to
kind
of
say
it
like
that,
because
you
know
you
don't
have
any
concept
of
like
smart
contracts
or
off
chain
on
chain.
Anything
like
that,
because
if
you
think
about
it
like
tcpip
is
a
protocol,
http
is
a
protocol.
Also,
the
graph
is
a
protocol,
so
uniswap
is
a
protocol,
so
any
of
these
things
have
to
be
categorized
in
a
protocol.
So
it's
important
to
have
like
a
very
generic
definition.
C
Yeah
and
I
think
kind
of
adding
on
there
to
talk
about
protocols
in
web
3.
Specifically,
I
always
like
to
say
it's
sort
of
a
series
of
smart
contracts
that
enable
on-chain
actions.
So
you
know
in
uniswap's
case
the
uniswap
protocol.
Is
you
know
this
sort
of
suite
of
you
know
non-upgradable,
persistent
contracts
that
together
create
an
amm
and
and
the
on-chain
actions.
It
allows.
Are
this
sort
of
peer-to-peer
market
making
and
swapping
and
and
just
to
kind
of
bring
it
back
to
web3?
E
F
F
Yeah,
I
would
echo
a
lot
of
the
same
things
here.
The
the
I
think
about
it
as
an
important
thing
here
is
having
a
common
set
of
rules
that
everyone
is
agreeing
to,
and
then
you
can
build
communities
on
top
of
it,
but
we
all
need
to
like
have
this
set
of
rules
that
can
coordinate
everything
that
we're
doing
and
you
need.
A
Do
you
need
a
token
to
kind
of
coalesce
all
those
incentives
together,
because
yeah
I
mean,
I
think
a
bitcoin
is
the
first
time
I
really
heard
that
word
get
thrown
around,
but
then
at
the
same
time,
http
and
the
other
kind
of
web.
One
protocols
didn't
really
have
a
token.
So
do
you
guys
feel
like
to
incentivize
continuous
development?
You
need
a
token
or
do
you
need,
like
a
company
or
an
equity
bearing
organization
that
is
kind
of
incentivized
to
push
the
project
forward.
B
I
think,
in
my
opinion,
a
protocol
is
not
a
company,
it's
a
public
infrastructure,
so
it's
public
goods
protocol
directly
should
not
make
money.
I
think
it
should
like
allow
other
actors
within
the
protocol
to
make
money,
but
I
think
protocols
themselves
shouldn't
be
making
money.
They
should
be
public
infrastructure,
public
goods,
and
because
of
that,
I
think
it's
very
important
to
have
a
token,
because
I
think
tokens
are
like
the
best
mechanism
that
we
have
today
to
like
fund
ongoing
development
of
a
protocol
without
having
to
like
charge
protocol
fees
or
something
like
that.
C
Yeah
and-
and
I
think
in
in
unisop's
case,
you
know
the
the
uni
token
is
a
way
to
give
power
to
the
community
and
to
users
in
the
future
success
of
the
protocol-
and
you
know,
we've
seen
a
lot
of
success
there
with
you
know,
for
example,
you
know
governance
recently
enabled
the
one
bit
fee
tier
on
v3,
which
you
know
in
a
couple
of
months.
You
know
you
saw
the
protocol
essentially
gained
dominance
in
a
lot
of
the
stable
coin.
Disabled
corn
market
share.
C
D
Yeah
yeah,
I
would
say
the
only
case
where
you
don't
need
a
token
would
be
if
you're
super,
confident
that
you
got
the
protocol
right
the
first
shot,
and
then
you
never
need
to
change
it
again.
Unless
you
just
deploy
it
and
you're
done,
then
you
don't
need
you
don't
need
a
token,
and
if
you
need
to
iterate
on
the
protocol
and
continue
to
make
changes
and
have
this
community
around
it
and
and
coordinate,
then
the
token
is
probably
the
most
effective
way
of
for
for,
like
a
coordination
point.
E
That
are,
you,
know,
kind
of
in
in
the
hands
of
the
world,
but
I
think
in
in
web
one
times,
actually
there
were
centralized
standard
setting
bodies
like
w3c
and
others
that
were
able
to
you
know
there
are
multiple
versions
of
http,
for
example,
right
and
having
ssl
and
and
all
these
other
things
be
developed
on
top,
for
example,
that
required
some
kind
of
you
know
disparate
parties
to
coordinate
with
one
another
to
agree
upon
that
standards.
You
needed
browser
makers
to
support
it.
E
Then
you
needed,
you
know,
like
the
apache
web
servers
of
of
the
world,
to
support
it
as
well,
but
I
think
you
know
those
those
standards
have
languished
over
time
because
of
this
tragedy
of
the
commons
problem.
Right,
like
what
incentive
do,
google
and
facebook
and
whomever
have
to
continue
to
pay
into
those
things,
their
incentive
is
actually
to
make
sort
of.
You
know
their
own
solutions.
These
problems
that
are
not
open
and
not
standards
compliant
because
it
locks
people
into
their
systems
right.
So
so
I
think
it
is
possible.
A
F
B
A
And
it's
a
mixed
bag
because
bitcoin
like
at
its
very
heart,
needs
the
token
or
else
the
security
of
the
network.
Isn't
there
so
I
mean,
I
guess
it's
yeah,
it's
it's
a
bit
of
it's
a
bit
of
everything
and
all
right.
Just
a
question
like
when
you
guys
have
we
kind
of
touched
on
this
already,
but
when
you
have
a
protocol,
do
you
think
of
it
as
like?
Are
you
constantly
developing
it
based
on
governance
or
your
own
choices?
A
Or
do
you
think
it's
like
you
deploy
the
code
and
then
you
let
it
go.
I
mean
even
bitcoin
is
under
continuous
development,
but
like
http,
I'm
not
sure
if
http
is
still
like
being
worked
on
or
something
like
that.
So
what
do
you
guys?
Think
of
as
decentralization
like
does
decentralization
mean
totally
immutable
code?
You
you
submitted
it
like
sarah.
I
know
you
said
that
unit
swap
like
v2.
A
B
B
Try,
like
you
said
so,
then
you
might
have
some
issues
which
end
up
being
like
huge
user
experience,
issues
where
you
have
to
move
liquidity
from
one
contract
to
another
contract,
and
you
have
to
coordinate
like
a
decentralized
set
of
actors
to
make
these
migrations
like
in
our
case,
it
was
hard
because
we
have
like
these
routers
who
are
liquidity
providers
who
are
running
software.
We
have
to
get
them
to
pull
their
liquidity
out
of
the
old
contract
and
the
new
contract.
B
We
will
go
through
like
dow
governance
to
upgrade
things
and
I
think
it
can
work
actually
if
you
use
an
upgradeable
system,
but
I
think
it's
very
important
to
have
like
time
locks
governance.
Things
like
that
like,
if
you
give
people
the
chance
to
exit
the
protocol,
if
they're
not
happy
with
the
upgrade,
then
I
think
it's
a
fair
to
say
that
upgradability
can
be
part
of
the
protocol.
C
Yeah-
and
you
know
I
think
in
in
uniswap's
case,
you
know,
immutability
is
sort
of
a
key
feature
of
the
suite
of
smart
contracts
and
and
really
that's
just
to
guarantee
that
you
know
something
like
social
collusion,
couldn't
change
the
code
in
a
malicious
way.
You
know,
I
I
think
another
way
to
think
about
this
is
that
upgradability
sometimes
may
you
know,
sacrifice
the
trust
that
a
user
has
in
the
code.
You
know
if
somewhere
down
the
line,
that
code
might
change,
and
so
that's
been
like
a
really
key
thing.
D
Yeah
for
for
life
here
it's
been
five
years
since
we've
been
working
on
it.
It's
always
been
upgradable,
mostly
because
we
we
knew
that
you
know
from
the
very
beginning.
This
was
going
to
be
a
long-term
project.
We
mapped
out
like
a
seven-year
plan
when
we,
when
we
first
started,
and
we
knew
we
needed
to
have
multiple
phases.
So
in
order
to
take
account
for
that,
we've
always
had
upgradeability.
D
I
do
think
for
for
a
project
like
live
pr
where
you
know
we're
trying
to
provide
video
streaming
infrastructure.
Just
given
that
video
streaming
is
constantly
changing.
The
streaming
protocol
itself
is
constantly
changing.
The
needs
are
constantly
changing.
We
I
do
think
the
protocol
itself
will
also
need
to
be
upgradeable
over
time,
but,
of
course,
with
clear
governance,
principles
and
and
all
the
things
that
were
mentioned.
E
Yeah,
I
think
one
of
one
of
the
things
that's
most
interesting
or
really
different
about
on-chain,
like
crypto
crypto
native
protocols,
from
like
a
traditional
protocol,
is
that
there's
kind
of
this
like
singleton
aspect
to
it.
The
code
that
is
running
that
single
incantation
of
the
system
on
chain
is
the
the
protocol
in
a
sense
whereas,
like
when
that
term
was
created,
it
was
actually
referring
to
like
a
document
right,
a
document
that
explained
how
you
know.
Two
machines
could
talk
to
one
another
or
something
like
that.
E
E
Audience
has
been
since
so
the
network
launched
formally
two
years
ago
and
all
of
the
on-chain
bits
of
the
network
and
then
all
the
off-chain
code,
which
is
effectively
the
only
acceptable
versions
of
that
off-chain
code
being
run
by
node
operators,
are
managed
by
an
on-chain
registry.
All
of
that
is
controlled
by
the
the
token
holders,
via
this
dow
structure.
So
just
simple,
just
a
straightforward
place
to
start-
and
it's
worked
well
for
us
so
far,
but
it
just
it
means
that
upgrading
requires
a
lot
more
coordination.
F
Yeah
they're
definitely
trade-offs
and
when
I
think
about
immutability,
it's
also
highly
affected
by
the
kinds
of
activity
that's
happening
on
the
protocol,
so
think
about
uniswap
and
you
do
a
single
transaction
and
then
you're
done
it's
pretty
self-contained
in
that
transaction.
It
doesn't
matter
if,
if
that
code
isn't
changing
after
that
fact,
then
goldbench,
which
is
upgradable.
F
There
are
businesses
in
brazil
and
nigeria
and
indonesia
who
are
taking
out
three
to
four
year
loans
through
goldfinch,
and
so
the
community
is
going
to
need
to
keep
adding
new
features
and
making
these
kinds
of
loans
work
in
better
ways,
and
so
the
upgradability
is
really
important
because
has
to
evolve
and
change
over
time.
Yeah.
D
D
Here,
we've
actually
done
both
types
of
upgrades
where
we've
changed
the
underlying
logic
of
the
smart
contract
via
a
governance
process,
but
we've
also
done
like
the
recent
upgrade
for
the
protocol
is
moving
from
layer,
1
ethereum
to
arbitrary,
and
that
was
deploying
a
whole
new
set
of
smart
contracts.
Then
helping
people
migrate
their
stake
from
l1
to
l2
right
where
you
have
both
protocols
running,
but
people
can
kind
of
move
between
the
two
cool.
A
Cool
all
right,
and
what
do
you
think?
What
have
are
the
things
that
have
stand
out
stood
out
in
terms
of
like,
since
I
mean
I
joined
the
space
in
2017
ish,
like
that
we've
learned
in
the
past
four
years
I
mean
luna
is
kind
of
like
a
depressing
thing
that
I'm
like
did
we
learn
anything,
but
we
definitely
have
so
I'm
just
like
like
what
have
you
learned?
What
have
you
learned
as
far
as
protocol
development
in
the
past
and
like
looking
forward?
A
Are
there
things
that
you
think
we
should
do
differently
in
terms
of
how
we
communicate
with
stakeholders
how
we
choose
to
deploy
tokens
yeah
just
like
moving
forward,
because
I
mean
obviously
it's
the
beginning
of
a
bear
market
and
you
know,
might
be
a
bear
market
for
a
couple
years.
So
that's
why
this
way
yeah.
B
F
That's
why,
with
goldfinch,
we
started
with
going
to
emerging
markets
and
providing
loans
in
those
places,
because
we
could
clearly
see
the
need
and
then
how
do
we
use
the
interoperable
systems
of
of
defy
and
crypto
to
like
serve
that
need
and
that
sets
up
a
stronger
foundation
versus
like?
I
don't
know,
other
other
things
that
maybe
will
you
know,
don't
do
as
well
afterwards,
yeah
for
sure
yeah.
E
Honestly,
it's
it's
a
bit
frustrating.
I
feel
it's
frustrating
that
we
haven't
yet
or
doesn't
feel
like
as
a
community
we've
yet
learned
how
to
onboard
like
larger
groups
of
users
and
more
mainstream
users
to
being
able
to
use
these
tools
engage
in
these
ecosystems.
I
I
think
too
much
of
what
we
do
is
like
building
for
each
other
and
building
for
this
same
existing
set
of
people
that
have
already
been
in
in
things,
but
I
I
think
it's
been.
It's
been
really
really
cool.
D
Yeah
for
sure
I
definitely
second
that
the
the
kind
of
I
think
the
value
of
what
web
creates
it's
much
more
crisper
now
right
than
like
you
know
years
ago,
when
we
first
developed
or
we're
developing
these
protocols,
like
one
of
the
big
lessons
that
I've
learned
in
protocol
development
itself
is
like
thinking
about
the
pace
of
change
and
thinking
about
the
technical
choices
that
we
make.
You
know
looking
at.
D
We
can't
look
at
things
from
a
five-year
time
horizon
where
we're
making
these
changes,
because
you
just
don't
know
what
happens
on
the
basis
based
our
infrastructure.
So
I
we
always
look
at
things
from
a
say,
like
eight
to
15,
eight
to
16
months
time
horizon
right,
where
you
know
that
the
technical
choices
that
you
make
in
that
time
frame
can
carry
you
through
that
period
and
then
you
have
to
make
new
choices.
C
Yeah,
that's
really
interesting,
because
I
think
I
sort
of
have
the
opposite
experience.
Where
a
lot
of
the
things
that
we
build,
you
know
we
really
try
to
build
them
for
the
long
term.
You
know
these
are
things
that
should
exist
forever
and
live.
You
know
on
chain
forever,
and
so
there's
sort
of
that
theme
underlying
a
lot
of
what
we
build
but
kind
of
thinking
about
the
future.
C
I
think
you
know
something
I'm
I'm
more
excited
about,
especially
with
the
uniform
protocol
is,
I
think,
we're
seeing
a
lot
more
of
like
a
multi-chain
cross-chain
world
unfold,
and
you
know,
especially
with
sort
of
the
open
sourcing
of
you,
know
the
protocols,
deployment,
scripts
and
documentation
and
stuff
we'll
start
to
see
the
protocol
live
on
a
lot
more.
You
know
l2s
and
other
chains,
and
things
like
that.
So
that's
I
think
something
I'm
really
excited
about
going
forward.