►
From YouTube: Graph Day Keynote by Yaniv Tal
Description
Yaniv Tal, Graph Protocol's co-founder and Project Lead, gave the keynote at Graph Day in San Francisco on January 25th, 2019. This presentation discusses the past, present, and future of Web3, and announces the launch of Graph Explorer and the Hosted Service. Yaniv and Jannis Pohlmann, co-founder and Tech Lead give a live demo of the Graph Explorer.
Sections:
0:00 Intro
1:25 The pros of Web2
2:44 The cons of Web2
5:13 Monopolies
5:49 Network effects
9:55 Web3
16:08 Benefits for developers
19:12 Benefits for users
22:02 Opportunity
23:00 The platform isn't ready/missing pieces
27:46 Building infrastructure
29:57 Roadmap
31:15 Graph explorer
47:00 Outro
Follow The Graph on social media
Twitter: https://twitter.com/graphprotocol?s=20
Instagram: https://instagram.com/graphprotocol
LinkedIn: https://www.linkedin.com/company/thegraph/
GitHub: https://github.com/graphprotocol
Website: https://thegraph.com
#web3 #dapps #thegraph
A
Right,
thank
you
all
so
much
for
coming
a
graph
day.
We
really
believe
in
web
3.
We
think
web
3
is
a
single
most
important
thing.
We
could
be
working
on
right
now,
but
it's
not
the
easiest
time
for
the
industry.
2017
was
an
exciting
year
for
the
CRISPR
crypto
space,
but
also
a
time
of
irrational
exuberance.
This
exuberance
brought
in
a
lot
of
capital
and
talent
and
attention
which
are
all
wonderful
things,
but
it
also
perpetuated
some
questionable
ideas.
A
A
What,
if
we're
10
years
too
early
well
today,
I
want
to
blow
that
narrative
out
of
the
water
I
want
to
share
what
exactly
is
wrong
with
the
way
we
build
software
today,
what
is
web
3
and
why
is
it
going
to
be
a
significantly
better
platform,
but
for
users
and
developers,
and
what
are
the
steps
that
we
need
to
take
to
get
there?
I
want
to
discuss
the
concrete
problems?
I
think
we
need
to
solve
and
I'm
going
to
share
what
the
graph
is
contributing
to
this
effort.
A
So
what
is
wrong
with
how
we
build
software
today?
Well,
actually,
let's
first
start
with
the
good,
we
have
to
remember
technology's,
been
moving
really
fast.
We
just
finished
learning
how
to
build
software
that
people
actually
want
to
use
software.
That's
fast,
beautiful
and
easy
to
use
software,
that's
smart!
As
developers
we
were
figuring
out.
A
A
A
A
The
server's
can
generally
scale
to
meet
demand
and
they're
generally
a
few
other
services
over
here.
All
of
this
running
in
the
cloud.
There
are
a
few
problems.
First
problem:
any
of
these
components
can
go
down
either
on
purpose
or
on
accident.
You
really
have
to
trust
these
companies
to
continue
to
exist
and
to
operate
these
servers
correctly.
If
the
company
goes
out
of
business
gets
acquired
if
the
product
gets
shut
down,
or
they
just
decide
to
change
product
directions,
you,
the
user,
are
out
of
luck.
The
second
problem,
these
interfaces
change
all
the
time.
A
If
the
database
table
changes,
the
server
can
break
if
the
server
code
changes,
the
rest
API
can
break
since
there's
nothing
that
guarantees
that
there
are
no
breaking
changes.
Making
software
updates
requires
really
careful
coordination,
and
this
makes
software
maintenance
are
difficult
and
costly.
A
Third
problem.
Whoever
operates
this
infrastructure
has
all
the
power.
If
they
want
to
change
a
record,
they
can
drop
into
the
database
done
if
they
want
to
take
your
private
information
and
sell
it
to
the
highest
bidder
done
if
they
want
to
wait
till
a
lot
of
people
are
using
the
app
and
then
go
in
and
change
the
rules
on
them.
That's
easy!
No
one
can
stop
them
from
doing
whatever
they
want.
A
So
these
are
problems
and
what
it
boils
down
to
is
that
you
really
need
to
trust
these
companies.
Trust
them
with
your
time.
Trust
them
with
your
data
with
your
money
with
your
security
and
if
the
last
few
years
have
taught
us
anything,
it's
that
we
can't
trust
these
people
here
are
just
a
few
headlines
from
the
last
few
years,
and
there
are
hundreds
like
these
I,
don't
even
try
most
new
applications
anymore,
because
I
don't
really
want
to
spend
my
time
and
energy
on
an
application.
That's
probably
going
to
get
shut
down.
A
But
all
of
this
trust
means
is
that
we
end
up
with
monopolies.
I
had
fewer
companies
that
have
more
and
more
power.
We
get
monopolies
in
verticals
like
search
social
retail,
video
ride-sharing,
even
business
apps
like
sales
monopolies,
resulted
in
rent-seeking,
behavior,
restricted
freedom
of
choice
and
increased
prices
for
consumers.
It
means
that
people
have
less
control
over
how
the
things
in
their
lives
work.
Of
course,
Trust
is
just
part
of
the
problem.
There's
another
force
at
play
here.
Network
effects
with
both
data
and
liquidity
and
marketplaces
network
effects,
accrue
further
concentrating
monopolies.
A
These
network
effects
are
a
real
force.
It
might
as
well
be
physics.
Imagine
if,
when
we're
moving
out
west
as
a
nation,
we
just
handed
out
states
to
the
first
credible
settlers
that
could
erect
a
few
buildings.
Here,
you
take
Nevada,
you
can
have
Oregon
you
take
California
in
perpetuity
for
life.
That's
essentially
what
we've
done
here,
that's
web.
A
So
how
did
these
network
effects
work?
Well,
if
I'm
a
service
provider,
then
I
also
get
to
manage
the
data.
You
remember
the
web
architecture,
since
all
of
that
is
really
complicated.
To
maintain
somebody's
got
to
run
that
infrastructure
and
that
somebody's
also
going
to
get
to
control
and
manage
data
now,
once
we've
given
them
that
data
their
incentive
is
to
create
a
moat
to
protect
their
business.
A
This
is
when
they
destroy
the
user,
experience
to
drive
up
profit
and
because
of
switching
costs
are
so
high.
It
can
really
ratchet
this
up
and
there's
nothing.
We
can
do
if
that
wasn't
enough,
once
they
drove
up
monetization,
they
decided
to
sell
the
company
to
Microsoft.
They
can
do
that
now.
The
new
owner
has
access
to
all
the
data.
They
can
change
the
rules.
However,
they
want
and
there's
nothing
we
can
do.
A
A
Another
problem
with
this
architecture
is
that
it's
a
honeypot
for
hackers,
a
single
lone
developer,
can
make
a
mistake
that
can
expose
everybody's
private
information.
These
are
just
a
few
breaches
from
the
last
few
years,
and
I
recognize
that
users
don't
seem
to
care
enough
about
privacy
to
change
behavior,
it's
one
of
those
things
that
just
kind
of
sneaks
up
on
you
and
it's
hard
to
know
when
to
make
a
change,
but
were
the
engineers
we're
responsible
for
this?
We
need
to
fix
this.
A
A
Efficient
marketplaces
ensure
that
users
are
getting
a
competitive
service
at
the
lowest
price,
as
some
people
might
want
these
protocols
to
be
free,
a
problem
is
we
can't
guarantee
these
other
properties
like
stability,
without
a
way
to
cover
the
cost?
If
we
want
this
infrastructure
to
be
sustainable
and
we
don't
want
to
create
perverse
incentives,
then
payment
needs
to
be
built
in
it
need
to
be
secured.
In
most
cases,
that
means
Byzantine
fault,
tolerant
or
roughly
resilient
to
malicious
actors.
A
A
A
Let's
take
a
look
at
the
web
3
architecture,
it
starts
with
a
set
of
decentralized
infrastructure
protocols.
These
protocols
provide
the
core
building
blocks
and
primitives
on
top
of
the
hardware.
They
include
things
like
block
chains
and
storage
networks,
anything
that
requires
low-level
computation
in
order
to
be
efficient.
Now
current
blockchain
designs
require
that
computation
is
replicated
on
tens
of
thousands
of
machines
and
that
obviously
doesn't
scale
but
we're
in
the
early
days
of
experimenting
and
exploring
this
design
space.
A
Just
this
year,
we're
going
to
see
the
release
of
a
new
class
of
protocols
in
polkadot
and
cosmos,
and
with
these
protocols
each
application
can
have
its
own
block
chain
and
these
block
chains
I
can
connect
together
to
benefit
from
pooled
security
and
interconnectivity,
so
I
think
we
should
expect
to
see
continuing
experimentation
and
much
bigger
improvements
here.
On
top
of
the
infrastructure
protocols,
we
have
what
I'm
calling
logic
protocols.
These
can
be
built
on
smart
contract
platforms
and
they
encode
a
set
of
rules
that
allow
you
to
solve
a
coordination
challenge
amongst
participants.
A
A
good
logic
protocol
should
tackle
the
smallest
piece
of
functionality
that
it
can
so
that
it
can
be
reused
as
a
modular
building
block
anytime
you've
got
a
value
exchange
between
a
buyer
and
seller,
consumer
and
provider.
A
natural
contention
arises
where
the
interests
of
one
party
might
may
differ
from
the
interests
of
another.
In
all
of
these
cases,
a
logic
protocol
can
encode
what
a
fair
exchange
would
look
like
who's
allowed
to
perform.
What
action?
A
A
And
then,
on
top
of
these
logic
protocols,
you've
got
the
DAPs.
An
adapt
is
just
an
application.
That's
built
on
top
of
these
web
3
protocols.
Now,
because
all
of
these
protocols
are
open,
they're
going
to
produce
a
tremendous
amount
of
open
data,
that
data
means
that
a
developer
can
come
in
and
tap
into
that
to
build
applications
that
previously
would
have
been
out
of
reach.
This
data
also
means
that
we
can
have
interoperability
at
the
data
layer.
A
Every
time
we
take
a
piece
of
logic
or
data
and
push
it
down
into
the
protocol
layer.
We
create
a
new
modular
building
block
that
can
be
leveraged
by
the
next
developer
and
the
next
we
go
from
having
hundreds
of
thousands
of
individual
siloed
machines
each
one
that
lives
and
dies
with
the
lifespan
of
that
individual
project
to
having
a
giant
global
machine
that
persists
and
is
available
to
all
developers
can
tap
into
these
networks
without
fear
of
them
breaking
or
changing
under
their
feet.
A
Think
about
how
much
duplicated
work
we
have
across
every
software
company
setting
up
infrastructure
databases,
managing
users
and
accounts
duplicating
features,
bootstrapping
networks,
there's
going
to
be
orders
of
magnitude,
efficiency
gains
in
the
shift,
and
all
of
that
is
only
possible
because
of
the
guarantees
of
openness,
security,
stability
and
verifiability
that
are
force
it
every
layer
in
the
stack.
So
this
is
all
really
exciting,
but
how
do
we
get
people
to
actually
use
this
stuff?
Where
are
the
users
well
in
a
capitalist
society,
everyone's
free
to
make
decisions
and
their
own
best
interest?
A
A
These
are
the
things
that
developers
care
about
when
they're
deciding
what
platform
they
want
to
build
on,
I.
Think,
whatever
platform
solves
for
these,
the
best
is
going
to
win
the
lion's
share
of
developers.
So,
let's
see
how
we're
doing
let's
compare
web
2
to
where
we're
at
with
web
3
today
and
where
we
might
be
able
to
get
to
as
we
build
more
of
the
infrastructure
and
the
platform
matures
so
building
compelling
apps.
A
You
can
certainly
do
that
on
web
2,
but
you
are
also
limited
to
what
you
can
build
yourself
and
what
you
trust
third
party
API,
is
to
provide
the
daps
on
web.
3
are
primitive
today
and
the
UX
is
not
great,
but
once
we
solve
those
things,
I
think
we'll
be
able
to
build
apps
that
are
every
bit
as
compelling,
if
not
more
compelling
with
web
3
ease
of
development.
Web
2
has
a
couple
decade
head
start
here.
A
The
tooling
is
still
really
immature
for
web
3,
so
it's
hard
to
build,
adapt
but
again,
I
think
with
web
3.
Once
we
build
out
the
infrastructure,
I
think
it'll
be
even
easier
than
web
2,
because
there's
going
to
be
so
many
more
reusable
components
for
developers
to
reliably
stitch
together,
maintenance.
This
is
an
area
of
the
web
2
struggles
because
you
have
to
run
the
infrastructure
yourself.
Web
3
is
already
ahead
here.
Once
you
deploy
a
smart
contract
to
aetherium,
you
don't
need
to
touch
it
again.
It
just
runs.
A
A
This
is
a
major
point
of
friction
today,
and
it
definitely
hinders
distribution
on
web
three,
but
we're
already
seeing
where
distribution
for
web
three
can
actually
become
really
compelling.
You
can
see
now
DAP
stores
inside
of
browsers,
like
coin
based
wallet
and
balance,
and
these
discovery
experiences
can
be
even
better
tailored
than
in
closed
platforms
like
the
Apple
App
Store
business
models
for
web
2
or
mostly
ad
based
or
subscriptions
with
web
3
I.
A
Now,
let's
take
a
look
at
the
motivation
for
users.
These
are
the
things
that
users
care
about.
Let's
see
how
we're
doing
for
great
user
experience.
Web
2
is
already
ahead,
and
this
is
an
area
that
we
need
to
catch
up,
so
I
think
we
we
can
absolutely
do
it
and
we
just
need
to
you
know,
set
set
the
bar
high
and
get
there
more
features.
So
this
is
an
area
I
think
web
3
can
really
shine.
Have
you
seen
any
public
issue
trackers
where
people
can
request
features
for
products?
A
Many
of
these
have
hundreds
or
thousands
of
feature
requests,
and
the
problem
is
a
single
product.
Team
can
never
hope
to
build
all
of
the
features
that
users
want
once
we
open
up
development
so
that
anyone
can
come
and
contribute
and
get
paid,
I
think
we'll
see
a
lot
more
features
getting
built,
easy
setup.
This
is
another
place
where
web
3
can
really
join,
because
we
can
have
data
portability
between
applications.
A
I
think
a
lot
more
people
will
be
able
to
feel
comfortable
participating
in
these
marketplaces
and
even
though
I'm
sure
it's
going
to
take
a
while
I
think
that
will
allow
these
markets
on
web
3
to
eventually
dwarf
web.
To
you
know,
price
and
cost
is
an
interesting
thing
to
consider
with
web
3.
It's
certainly
expensive
today,
because
the
protocols
don't
scale,
and
so
you
have
to
pay
a
lot
in
network
fees
where
this
ends
up
I.
Think
it's
too
early
to
tell
you
have
some
opposing
forces.
A
On
the
one
hand,
you
have
the
increased
cost
of
the
additional
security
requirements
and,
on
the
other
hand,
you
had
the
downward
price
pressure
of
competitive
markets
and
economies
of
scale
so
for
cost,
I
think
we're
just
going
to
have
to
wait
and
see
and
then
finally,
for
stability
because
of
the
guarantees
of
the
platform
web
3
is
already
ahead.
So
again,
when
we
look
at
this
there's
some
good
news
and
some
bad
news.
The
bad
news
is
that
we
suck
I
mean
it's.
A
A
There
are
about
23
million
developers
worldwide.
Most
of
these
developers
do
not
have
the
same
opportunities
that
we
have
here
in
San
Francisco,
it's
not
as
easy
for
them
to
go
out
and
raise
a
two
million.
She
drowned
or
a
$10,000,000
series
a.
But
you
know
what
they're
every
bit
as
capable
of
writing
code.
That's
the
opportunity
there
may
be
a
hundred
thousand
developers
at
Facebook,
Apple,
Amazon,
Netflix
Google.
They
may
seem
like
giants
now,
but
they
can't
compete
against
23
million
developers,
not
with
a
platform
that
enables
large-scale
coordination
consistently.
A
A
It's
clear
that
the
platform
isn't
ready,
but
if
we
want
to
learn
about
launching
platforms,
there's
a
great
example
that
we
can
look
at
the
iPhone
is
arguably
the
most
successful
platform
of
all
time,
but
some
of
you
might
remember
when
it
launched
a
lot
of
people,
thought
it
would
fail.
They
thought
that
nobody
would
want
a
phone
that
didn't
have
a
physical
keyboard.
What
are
the
use
cases
for
this
thing?
They
said,
so
what
did
Apple
do?
A
A
We
need
better
and
more
protocols.
It's
still
impossible
to
build
a
lot
of
the
kind
of
decentralized
apps
that
we
would
like.
So
it
needs
to
be
possible
and
easy
Eric
is
going
to
be
speaking
about
the
design
of
the
live.
Pure
protocol
is
a
great
example
right
after
this
talk,
blockchains
are
still
slow
inexpensive,
but
if
we
can
get
block
times
to
around
3
seconds
finality
to
around
6
and
transaction
cost
to
under
a
cent,
I
think
we'll
be
able
to
use
block
chains
for
a
lot
more
applications.
A
A
We
need
to
create
abstractions
to
better
communicate
things
like
finality
to
users,
and
this
idea
that
users
should
manually
approve
every
transaction
I
think
makes
sense
for
large
purchases,
but
is
never
going
to
work
for
state
updates
in
web
3.
These
are
just
a
few
examples.
Tara
is
going
to
be
speaking
about
fixing
DAP
you
acts
later
today.
A
A
A
A
If
we
can
make
it
so
that
anyone
can
build
a
DAP
and
get
paid
competitively
to
do
that,
that
is
really
going
to
enable
the
wave
of
development
that
we
want
to
see
so,
as
I
think,
are
the
missing
pieces
and
I
think
once
we
have
these
built,
we'll
be
able
to
say
that
the
platform
is
ready.
A
lot
of
these
I
think
we
can
do
in
the
next
12
to
24
months.
A
A
A
As
soon
as
you
start
building
an
application
on
a
blockchain
like
aetherium.
You
quickly
realize
that
it's
really
difficult
to
get
the
data
that
you
need
to
your
front-end
application,
say:
you're,
building
a
digital
goods
marketplace
the
users
they
want
to
buy
and
sell
digital
good
are
probably
going
to
want
away
to
search
and
filter
for
specific
kinds
of
items.
Maybe
they
want
to
sort
to
find
the
most
valuable
items
that
are
for
sale.
A
Traditionally,
there
have
been
kind
of
two
ways
to
try
to
solve
this
problem.
One
you
can
either
build
a
centralized
proprietary
indexing
server.
Their
problem
is,
then,
you
lose
the
guarantees
that
we
care
about
like
stability
and
verifiability.
The
second
option
is
to
try
to
process
everything
on
the
client,
but
depending
on
how
much
activity
or
smart
contract
has
trying
to
sync
all
of
that
chain.
Data
and
the
client
can
take
minutes
two
hours
to
even
days.
Those
were
the
only
two
options
available
until
the
graph.
A
You
start
by
defining
something
we
call
a
sub
graph
and
the
sub
graph
defines
what
data
sources
you
want
to.
Listen
to
how
you
want
to
transform
that
data
and
what
schema
you
want
to
use
to
query
that
data
once
you've
built
that
sub
graph,
you
deploy
it
to
a
graph
node
and
that
nodes
start
subscribing
to
the
underlying
data
sources
when
an
event
is
triggered.
It
processes
that
data
in
a
way
MVM
and
that
allows
computations
to
be
fast
and
deterministic
from
there.
A
A
A
We
want
to
make
it
so
anyone
can
run
a
graph
node
to
get
paid
to
provide
indexing
and
query
processing
services
to
the
network,
and
we
want
to
build
in
those
guarantees
around
things
like
stability
and
verifiability,
to
make
sure
that
users
can
verify
that
the
data
they're
getting
back
is
correct.
Brandon
Ramirez,
our
co-founder
and
research
lead,
is
going
to
be
sharing
new
details
about
our
decentralized
network
later
today.
A
A
The
graph
Explorer
is
a
place
that
developers
can
come
to
find
all
of
the
data,
that's
being
indexed
on
the
graph
and
easily
pull
them
into
their
applications.
I'm
also
excited
to
announce
seven
amazing,
launch
partners
today.
Thank
you
today,
we're
launching
with
Dharma
compound
Eunice,
WAP,
ENS,
origin,
decentraland
and
live
Pierre.
All
of
these
amazing
protocols
are
now
available
on
the
graph,
so
anyone
who's
building
applications
and
wants
to
access
this
data
can
now
do
it
really
easily
using
graph
QL.
A
A
A
All
of
these
are
in
a
draft
state
right
now,
which
means
that
there's
still
subject
to
change,
but
we
have
this
concept
of
publishing,
at
which
point
we
enforce
it.
There
are
no
more
breaking
changes.
Let's
take
a
look
at
one
of
these
sub
graphs.
You
can
look
at
compound,
so
here
you
can
see
that
it's
fully
synched.
You
can
see
it.
What
the
latest
block
number
is
here.
This
is
going
to
live,
update
as
new
blocks
or
mind
and
the
graph
node
processes
it.
You
can
see.
A
This
was
updated
recently
and
we
support
updating
sub
graphs
with
zero
downtime
deploys
we've
indexed,
just
shy
of
5,000
entities
for
the
compound
sub
graph.
If
you
come
down,
you
can
see
the
playground
or
you
can
see
what
data
is
available
through
the
schema,
and
you
can
see
an
example
query
which
you
can
hit
play,
and
you
can
query
for
data
here.
You
can
see
the
logs,
so
you
can
see
what
events
have
been
triggered
and
processed
by
the
graph
node
and
let's
take
a
look
at
another
sub
graph.
A
This
is,
you
know,
swap
which
is
a
decentralized
exchange
that
happens
all
on
chain
and
one
of
the
great
things
about
graph
QL
is
even
if
you
don't
know
what
data
you're
looking
for
with
the
feature
called
introspection,
you
can
get
a
drop-down
and
that
allows
you
to
just
kind
of
explore
around
and
see
what
data
is
available.
So,
let's
see
how
we
can
look
up
some
information
about
transactions
that
have
gone
through
on
unis
walk.
So
even
if
I
don't
know
what
fields
exists
on
these
transactions,
I
can
look
and
I
can
see.
A
You
know,
I
could
grab
the
block
if
I
want
maybe
the
event
type
and
maybe,
for
example,
we
can
see
what
the
most
fees
earned
on
yunusov
transactions
have
been
to
the
liquidity
providers.
You
can
also
have
parameters
for
collections,
so
if
I
want
to
do
things
like
pagination
I
could
grab
the
first
like
10
and
then,
if
I
want
to
see
the
most
expensive
fees
that
have
been
paid
out,
I
can
do
sorting.
So
I
can
see
you
order
by
see.
You
know
what
I
can
order
by.
A
So
we
want
to
do
fees
and
then,
if
we
want
to
see
the
most
expensive
ones,
we
can
set
the
direction
to,
for
example,
descending.
So
if
I
had
played
there,
there
are
the
most
expensive
transactions
that
we've
seen
process
on.
You
know
swap
so
it's
just
that
easy
to
explore
around
and
find
data,
that's
being
indexed
on
the
graph
and
pull
it
into
your
applications.
A
B
Hi
I'm
gonna
stay
on
this
machine
cool.
So
you
probably
wonder
how
do
you
I
get
a
subgraph
for
my
smart
contracts
on
to
this
Explorer
and
how
do
I
push
it
to
them
to
the
hosted
service
in
general?
How
do
I
use
it
in
my
DAP
and
guess
what
we're
gonna
do
in
the
next
five
minutes?
So
the
first
thing
I'm
gonna
do
is
I'm
gonna
sign
in,
and
this
signs
mean
as
genève
I'm
on
his
laptop
and
I'm
gonna.
B
Go
to
the
dashboard,
which
allows
me
to
create
new
sub
graphs
and
I
can
see
all
my
organizations
and
I
have
an
access
token
that
will
need
first
of
all,
I'm
gonna
create
a
new
sub
graph.
I
want
to
call
it
counter.
It's
going
to
be
a
very
simple
sub
graph.
It'll
pick
a
nice
image
that
should
be
here.
B
Okay,
don't
go
too
crazy
with
this,
we
don't
want
to
be
the
next
myspace
okay,
so
we've
got
the
sub
graph
created.
What
I
will
need
for
the
next
steps
is
an
access
token,
so
you
can
see
the
sub
graph.
Is
there
on
my
dashboard?
It's
not
deployed
yet
so
this.
What
we're
gonna
do
now
so
graph
in
it
is
a
tool
that
we
provide.
It
allows
you
to
very
easily
create
a
sub
graph,
a
new
sub
graph
from
an
example
and
I'm
gonna.
B
Do
I'm
gonna
use
the
name
that
I've
used
in
the
Explorer,
so
it's
counter
is
the
name
of
the
sub
graph
and
I'm
good.
Just
I'm,
just
gonna
give
it
the
directory
that
it
should
create
a
sub
graph
in
and
what's
nice
about.
This
is
it
creates
the
sub
graph,
and
it
also
gives
me
next
steps
that
I
need
to
run
in
order
to
work
with
this
OP
graph,
deploy
it
and
so
on.
So
we're
gonna,
I'm,
actually
gonna
need
the
access.
B
Token
again,
gonna
run
this
command
to
authenticate
with
the
hosted
service
in
the
past
max
talk,
and
it
will
store
that
securely
in
your
system,
keychain
and
now
we're
going
to
do
into
the
counter
sub
graph.
As
you
can
see,
it's
very
similar
to
like
a
regular
Jase,
JavaScript
or
typescript
project
and
first
gonna
run
yarn
in
yawn
to
install
all
the
dependencies
that
we
need
for
building
and
deploying
this.
B
It
should
take
only
a
few
moments
there
we
go
and
then
what
we're
gonna
run
quickly.
I'm
gonna
explain
this.
While
we
do,
this
is
run
yawn
cogent.
This
you
can
see
this
runs
graph
Cochin,
which
is
another
tool
we
provide,
and
what
this
will
do
is
it
will
create
a
semblance
crypt,
which
is
a
subset
subset
of
typescript,
that
compiles
to
web
assembly
assembly
script
types
for
your
contract
ABI
eyes.
B
B
B
The
manifest
is
what
is
where
everything
like
comes
together,
where
your
schema
is
reference,
where
you
can
add
some
metadata
and
where,
for
every
for
every,
like
collection
of
data
that
you
want
to
index
and
have
and
be
able
to
create
that
you
create
a
data
source.
In
this
case,
we
have
a
simple
data
source
based
on
the
counter:
smart
contract,
it
lives
on
main
net
and
it
lives
at
this.
Address
uses
this
ABI
and
in
order
to
then
index
data
from
this
particular
contract,
we
define
a
mapping
in
this
case.
B
It's
an
ethereal
events
mapping
which
will
listen
to
your
theorem
events,
scan
pass
blocks
for
those
events
and
whenever
it
encounters
such
an
event,
it
will
call
an
event
handler
for
this,
and
then
that's
what
you
can
where
you
can
write
your
code
to
create
the
data
that
you
actually
want
to
query.
So
here
you
can
see
we're
using
a
counter
API,
which
is
pre
generated
from
solid
with
civility
compiler,
and
we
define
an
event
handler
for
the
incremented
event.
Remember.
B
This
is
a
counter
whenever
this
is
incremented,
it
will
emit
such
an
event
with
the
new
counter
value
and
the
handler
we're
gonna
call
for
this
is
called
handle
incremented
unless
this
lives
in
the
mapping,
which
is
an
assembly
script
file.
Let's
take
a
quick
look
at
the
schema.
It's
a
very
simple
one.
This
one
has
just
one
data
type,
so
this
is
essentially
the
data
model
for
your
sub
graph.
You
can
define
different
types.
You
can
define
fields
on
them.
B
B
We
try
to
load
a
counter
with
the
ID
default
counter,
there's
only
one
and
if
that
doesn't
exist
yet
meaning
we're
incrementing
the
counter
for
the
first
time
ever
we're
going
to
create
it
we're
going
to
set
its
counter
value
from
the
event,
and
you
can
see
this
is
all
you
know,
it's
all
auto
completed
and
typesafe,
and
then
we're
gonna
save
that
back
to
the
store.
So
it's
basically
incrementing
and
that's
basically
the
the
gist
of
what
a
subgraph
is
and
you
can
make
it
more
complex.
B
B
B
If
you
familiar
with
Apollo,
for
instance,
this
URL,
would
you
could
just
copy
that
and
use
it
straight
away?
So
let's
take
a
quick
look
at
this
and
what
the
subgraph
is
doing.
You
can
see
it's
already
synced,
like
70%
of
the
blocks.
I
mean
I've
created
this
recently.
So
there's
very
few
events
in
there,
but
it
scans
all
the
pass
blocks
up
to
the
current
block
on
the
etherium
main.
B
Yet
in
this
case
and
as
the
blocks
are
being
added
to
the
chain,
it
will
also
scan
those
four
new
events
that
are
relevant
to
our
sub
graph.
Okay,
this
might
take
a
couple
of
minutes.
Maybe
I
mean
it's
going
pretty
fast,
but
there's
probably
no
data,
yet
there's
counters
already
at
14.
So
that's
cool
cool,
okay.
So
that's
the
first
demo,
we've
created
a
sub
graph,
we've
deployed
it
to
the
graph
Explorer
on
the
hosted
service
and
we
can
look
at
it.
We
can
run
queries.
B
A
B
B
Let's
open
this
one
and
you
can
see
there
is
a
listing
scream,
but
that's
currently
not
really
doing
anything.
So
what
I'm
gonna
do
is
I'm
gonna
go
to
the
graph
Explorer
I'm
gonna
go
to
the
origin
sub
graph,
because
I
want
to
show
listings
that
out
on
the
origin,
protocol
and
I'm
gonna
work
on
my
query
a
bit.
So
what
I
want
is
I
want
the
status
of
all
of
those
things.
I
want
the
current
price.
I
want
the
currency.
Not
the
price
is
in
I.
B
B
Look
good
I
want
to
see
if
I
can
grab
any
media
files
from
this
from
these
listings,
so
I'm
gonna,
media
files
or
entities
again
so
I'm
gonna,
ask
the
ID
and
the
URL
and,
let's
say,
I,
want
the
first
50
and
I
want
to
query
the
most
expensive
one,
ignoring
the
currency
so
similar
to
what
you've
just
did,
order
by
surprise
or
a
direction
descending.
Let's
run
that
quickly.
B
There
we
go,
we
have
listings-
and
let's
put
exactly
this
query
in
to
our
app
and
now
I
hope
I
can
switch
over
to
that
fast
enough.
Maybe
there
we
go
it's
already.
It's
already
refreshed
too
fast
for
me
to
switch,
and
these
are
the
50
most
expensive
listing
on
origin,
cool
I.
Think
that's
it
I'll
hand
it
back
to
you.
A
A
A
Together,
we
can
build
a
solid
foundation
for
software,
a
new
platform
that
allows
us
to
build
farther
and
higher
than
ever
before,
and
when
we
do,
it
will
change
how
humans
cooperate
and
organize
and
create
economic
opportunity
for
everyone
in
this
global
connected
world.
On
a
personal
note,
this
is
the
most
exciting
ride
that
I
could
hope
to
be
on
and
I'm
so
thankful
that
enough
people
have
gotten
the
call
that
we
can
make
this
happen,
so
we've
got
a
really
great
day
planned
for
you
today.
Thank
you
so
much
for
coming.