►
From YouTube: Onboarding dapps and game storage to web3 - Adam Grodzki, Changwé Mutakasha, David Choi
Description
An open discussion with the leads of web3.storage on the challenges and opportunities they've observed building scalable infrastructure for web3.
A
Hey
everyone
how's
it
going
I'm
David
I
lead
the
web,
3
storage
team
on
the
product
and
business
side
and
I'm
super
stoked
to
be
here
today
and
then,
as
anchovan
pointed
out,
I
guess
I'm.
The
last
thing
we
are
the
last
thing
standing
between
you
and
lunch.
So
it's
a
ton
of
power
we'll
try
not
to
let
it
get
to
our
heads,
but
yeah
I
mean
this
is
a
pretty.
It
was
initially
pitched
as
a
community
brainstorm
to
us.
A
So
I
would
love
engagement
from
all
of
you,
all
the
builders
and
folks
doing
all
the
hard
work
out
there.
If
you
have
any
thoughts
on
the
topics,
just
raise
your
hand
and
someone
will
come
by
with
a
microphone
if
you
want
to
add
anything
to
the
conversation,
otherwise
I'll.
Let
these
two
good
folks
introduce
themselves
as
well.
A
Yeah
and
it's
a
it's
a
good
cast,
because
you
know
our
team
works
on
the
storage
info
side
of
things,
so
things
happening
off
chain.
You
know
Chang
way,
and
the
team
helps
with
the
things
on
chain
and
the
infrastructure
there
and
then
Adam
is
actually
helping
build
in
the
space
and
has
experience
firsthand
both
the
highs
and
the
lows
and
some
of
the
challenges
so
yeah
just
to
jump
into
things.
I
think
really
quickly.
It's
always
useful
to
talk
about
the.
Why
of?
Why?
B
I
think
it's
web3
has
given
people
a
very
powerful
vehicle
in
terms
of
not
just
expression
of
self,
but
also
governance
of
data,
I
think
specifically
as
well
in
gaming.
What's
also
happened,
it's
it
empowered
developers
as
well
to
have
a
new
way
of
reaching
users
and
consumers
as
well
as
also
empowering
them
as
well.
Traditionally,
big
web
2,
Gaming
studios
are
very
profit
focused,
so
you
get
completely
drowned
out
in
a
market
of
Indian
devs,
but
as
web
3
has
come
along.
B
There's
this
exciting
new
innovation,
where
users
can
now
be
part
of
the
creative
process.
They
can
also
be
it's
a
more
democratized
way
of
gaming,
not
just
from
creation,
but
also
from
ownership,
as
well
as
continuation
of
the
story
as
well.
So
I
think
what
the
infra
should
allow
for
is
to
enable
people
to
create
Without
Limits
and
that's
exactly
what
we
believe
in
at
near.
C
I
can
also
add
that
on
an
engineering
level,
because
we
in
3s
games
still
were
mostly
Engineers
once
you
get
familiar
with
the
web
tree
stack,
which
can
be
quite
intimidating
at
the
start,
you
will
find
out
that
it
is
actually
accelerating
the
web,
the
the
game,
development
process
as
a
whole,
because
suddenly,
when
you
have
this
trustless
network,
that
you
can
get
interactions
with
directly
from
the
client,
you
solve
some
of
the
biggest
issues
when
creating
multiplayer
games.
So
you
no
longer
have
to
secure
the
connection.
C
The
network
on
your
side,
you
don't
have
a
single
point
of
failure,
which
is
your
login
services
that
you
have
to
run
and
in
general
running
multiplayer
games
is
something
around
20
percent
coding.
The
game
itself
and
80
is
coding,
the
back
end
the
services.
So
when
you
start
using
web3
stock,
so
it
goes
for
ipfs,
blockchain
and
anything
else
that
goes
into
it.
You
suddenly
don't
have
to
worry
about
it
and
that
speeds
up
the
development
of
many
features
of
a
game.
C
So
like
trading
items
in
between
users,
you
don't
even
have
to
implement
that.
If
you
implement
the
nft
stack,
either
of
its
erc721
or
1155
or
any
other
of
the
newer
standards,
you
suddenly
are
compatible
with
external
exchanges
that
allow
you
to
trade
items
in
between
players
directly.
Let's
say
on
openc
and
event:
openc
is
taking
a
revenue
cut
on
all
the
trades.
It's
still
a
lot
cheaper,
even
for
the
end
user
to
use
than,
for
example,
Steam.
A
You
know
there's
something
there,
but
equally
as
powerful
as
an
accelerant
is,
if
you
can
make
the
lives
of
the
development
Studios
and
the
engineers
easier,
an
nft
could
be
cool
to
take
from
one
Universe
to
another.
But
you
know
it's
not
necessarily
Universal
that,
like
one
game
that
has
a
sword
and
you
owning
that
translates
to
all
other
games
out
there.
So
you
know
there's
still
stuff
to
figure
out
on
that
level.
A
But
if
you
make
the
developer's
life
easier
by
plugging
into
the
nft
ecosystem,
because
all
the
tooling
is
there
and
now
you
don't
have
to
focus
so
much
on
specifically
what
you
need
to
build
to
enable
items
in
your
games
and
things
like
that.
That's
equally
as
powerful!
So
I
love
that
exam
people
and.
B
I
think
also
being
able
to
allow
developers
to
express
themselves
in
different
coding
languages.
So
I
think
traditionally
correct
me
if
I'm
wrong,
but
most
of
the
games
were
probably
coded
in
C
plus
and
built
on
Unity
right
now,
I
think
that's
80
of
the
market,
but
now
we
obviously
have
I
mean
on
near.
B
We
have
rust,
we
have
facility
environments,
we
have
substrate
and
octopus,
so
we
also
have
a
JavaScript
SDK,
so
that
allows
for
Millions
more
developers
in
a
coding
language
that
they
already
understand
to
come
in
and
develop
without
having
that
barrier
to
entry.
So
I
think
it's
also
removing
some
of
those
hurdles.
Because
then,
with
experimentation
comes
the
understanding
of
what
business
models
might
work
you
you
mentioned.
Obviously
things
like
play
to
earn
and
we've
things
we've
seen.
C
Yeah,
exactly
the
same
thing
happened
to
gaming
in
general,
with
quantities
eventually
comes
quality,
so
the
more
developers
get
their
hands
onto
the
technology
as
being
more
accessible
throughout
the
years.
More
quality
products
will
rise
up
and
people
will
use
them
because
they
like
using
them,
not
because
someone
has
really
good
monetization
plan
all
right.
So
the
same
happened
to
you
know
like
the
rise
of
the
Battle
Royale
games.
C
So
there's
a
lot
of
early
products
that
we've
seen
on
the
market
that
we're
focusing
on
how
to
sell
to
as
much
people
as
you
can
see,
and
even
a
very
bad
implementations
from
the
hands
of
the
AAA
developments
like
the
quartz,
which
was
quite
a
failure
but
yeah
once
the
Indie
developers
get
their
hands
on
web3
stack.
We
will
see
something
really
amazing
coming
in
the
near
future.
A
C
C
Projects,
but
they
must
use
the
Apple
payments,
which
basically
is
gatekeeping
all
the
payments
and
saying
that
we
take
30
from
everything
which
is
absurd,
like
there
is
no
real
reason.
There
is
no
like
technological
overhaul
that
they
need
this
30
for
to
to
maintain
those
applications
and
I
just
wondering
how
they're
going
to
approach
all
the
mobile
wallets
that
are
out
there,
which
are
web3
applications.
There
are
games
on
their
store,
they
do
transfer
value,
so
they
basically
are
doing
purchases
and
sales,
but
they're
not
using
their
payment
method
at
all.
A
And
they
do
cut
you,
that's
that's,
usually
the
mechanism
that
they
use.
You
know
they're
apps,
fighting
back
against
Apple's
model
by
letting
users
go
into
a
separate,
URL
and,
like
you
know,
sign
up
for
a
subscription
that
way
and
bring
that
to
the
app
to
try
and
get
around
it
got
some
of
these
absolute
platforms
and
things
like
that.
That's
usually
the
the
mechanism
they
use.
B
And
I
think
it's
it's
putting
the
point
that
you
guys
actually
raise
in
terms
of
that
fractionalization
and
that
centralization,
because
to
that
point
in
terms
of
actually
bringing
onboard
users,
you
do
have
to
have
some
sort
of
web
2
to
web
3
hybrid,
at
least
to
allow
some
sort
of
initial
injection.
I
mean
I,
spoke
about
lowering
the
barriers
for
developers,
but
there's
also
lowering
the
ballots
for
users
as
well.
B
B
A
building
on
near
play.
Ember
is
also
an
amazing
example
of
this,
where
they
basically
just
take
the
integration
and
they
everything
happens
on
the
back
end.
So
you,
as
a
user,
pretty
much,
don't
realize
that
you
win
a
web
3
kind
of
world
and
to
the
point
on
Apple,
I
suppose,
for
example,
on
near
on
every
smart
contract
that
executes
30
of
that
reward
goes
straight
to
the
developer,
for
that
smart
contract
execution.
So
it's
things
like
that,
rewarding
and
incentivizing
the
devs
as
well.
B
C
A
A
C
Sure
so
one
of
the
approaches
you
can
go
is
to
keep
everything
on
chain,
of
course,
which
is
a
very
expensive
approach,
usually
even
for
the
users,
and
there
comes
the
problem
of
data.
That's
being
constantly
changed,
so
you
cannot
run
the
entire
backend
on
chain
because
you
have
just
basic
stats,
live
stamina.
Other
parameters
that
have
to
be
running
live
on
the
server
directly
without
checking
with
the
blockchain
itself.
C
But
then
you
have
certain
data
that
you
are
storing
permanently
and
changing
it
like,
let's
say,
experience
points
that
is
reasonable
to
to
hold
on
chain,
but
with
ipfs.
What
you
can
do
is
you
can
hold
on
change
as
the
CID
and
then
you
can
mutate
the
data,
as
goes
in
the
ipfs
stack
and
I.
Think
one
of
the
coolest
Technologies
out
there
for
that.
A
B
A
When
you
read
off
like
the
ipfs
network
or
on
a
blockchain,
you
try
to
like
before
Services
existed,
that
were,
could
tell
you
pretty
quickly
like
what
the
truth
was
like.
You
might
have
some
latency
doing
reads
off
there
and
we
see
Services
stepping
in
to
kind
of
bridge
that
Gap.
So
just
a
quick
anecdote
from
our
experience
at
web3
storage,
we
started
out
by
running
a
bunch
of
ice
ipfs
nodes
in
the
cloud,
a
bunch
of
Kubo
nodes
and
that
worked
for
a
while.
A
Once
we
hit
a
certain
amount
of
scale,
it
really
like
the
performance
degraded
in
a
lot
of
times
across
the
whole
network.
We
were
the
only
ones
who
had
that
data
and
were
broadcasting
it
to
the
network.
So
we
had
to
go
ahead
and
do
a
few
things.
So
one
was.
We
wrote
a
new
implementation
of
ipfs
called
elastic
ipfs
that
was
designed
for
this
sort
of
scale
and
used
actually
a
centralized
Cloud
providers
from
a
physical
storage
standpoint,
because
they're
really
good
at
their
jobs.
A
With
a
snap
of
your
fingers,
you
can
spin
up
more
storage
or
you
know
more
bandwidth
and
things
like
that.
That
are
more
compute,
and
you
know
that
was
that
changed
the
game
for
us
and
instead
of
always
putting
out
fires
mode
like
we
were
serving
the
data
very
reliably,
and
then
there
was
another
Gap
from
going
from
that
and
serving
it
quickly
where
we
did
some
work
with
launching
our
own
Gateway
and
having
users
who
are
storing
their
data
with
us
have
that
data
available
on
edge
servers.
A
So
if
most
folks
are
reading
data
off
of
the
ipfs
network
out
of
a
Gateway,
you
can
use
hours
if
you're,
storing
data
with
us
and
get
it
at
CDN
level
speed.
So
it's
techniques
like
that
that
once
you
bridge
that
initial
Gap,
you
start
to
lean
into
some
of
actually
the
ways
that
a
technology
like
ipfs
wins
content
IDs
on
ipfs
are
insanely
cachable.
It
doesn't
matter
who's
requesting
that
and
from
like
what
origin
or
what
service.
B
That
sounds
that's
a
great
story,
because
I
mean
through
that
came
your
own
Innovation
and
your
own
experimentation.
How
you
could
expand
that
storage
and
how
you
could
make
it
more
accessible
and
readable
and
usable
as
well.
And,
of
course,
that's
kind
of
to
the
point
that
I
just
made
earlier
about
experimentation
within
web3,
and
it
allows
for
this
type
of
ongoing
learning,
which
it
could
be
expanded
by
rapidity,
because
there
are
significantly
lower
barriers
to
hurdles
for
more
people
to
get
in.
So
pretty
cool.
C
Making
games
in
general
is
hard,
and
you
have
to
be
mindful
about
everything
that
you
are
doing
as
an
engineer
when
making
a
game
and
same
goes
with
web
2
and
web3
development.
You
just
have
to
be
mindful
about
what
you
are
loading
when
and
how
much
in
advance
you
have
to
think
about
it,
but
it
doesn't
really
change
the
way
that
you're
building
the
games,
if
you
are
shifting
from
the
web,
to
stack
to
the
web,
3
stack
really.
A
Yeah,
that's
super
interesting
and
I
mean
I
would
also
add,
though,
at
the
same
time,
ideally
it's
as
easy
as
possible.
Right
like
the
fewer
headaches,
the
better
so
I
guess
that's
what
we're
all
working
towards
yeah
and
I
guess,
maybe
from
Adam
I'd
love
to
hear
if
you've
ever
faced
trade-offs
in
thinking
about
using
like
web3
versus
web
2.
Solutions
gateways,
for
instance,
are
a
little
bit
of
a
trade-off
where
you're
using
some
centralized
infrastructure
that
could
go
down.
A
You
could
swap
it
out
with
another
Gateway,
so
you
know
there
is
still
web
threeness
about
it
and
all
the
data
is
still
verifiable
through
the
content
ID.
But
you
know
I
think
that's
an
example
of
a
really
smart
trade-off
that
folks
make
I'm
wondering
if
you've
encountered
anything
else
when
it
comes
to
the
infrastructure.
C
I
mean
yes,
of
course,
we
we
did
find
some
hardships,
When,
developing
web3
applications,
especially
from
for
external
clients,
but
usually
the
actual
drawbacks
that
we
were
facing.
Is
that?
Because
part
of
the
stack
was
a
bit
forced
into
the
solution,
instead
of
being
used,
where
it
actually
makes
sense,
because
it
doesn't
make
sense
everywhere,
it
doesn't
make
sense
for
every
product
and
every
solution.
But
if
you
plan
ahead
of
how
you
are
using
ipfs
in
your
game,
then
you
don't
really
run
into
into
any
blockades.
C
Actually,
it's
actually
even
better
for
multiplayer
games,
because
when
you're
replicating
data
in
between
the
server
and
the
clients,
you
hit
certain
bandwidth
limitations.
When
it
comes
down
to
how
much
bytes
you
can
send
back
and
forth
It's,
not
a
limitation
for
the
clients,
it's
the
limitation
for
the
server,
because
when
you
are
updating
100
clients,
each
client
gets
just
a
position
of
100
other
clients
right
yeah.
But
the
server
needs
to
send
this
100
clients,
location,
200
players,
and
then
we
have
10
000
replications
right.
C
So
if
you
are
replicating
something
that
is
very
complex
and
you
can
instead
of
replicating,
you
know,
the
classes
locations
transforms
and
everything
that
comes
into
this
complicated
structure.
You
can
replicate
just
the
CID
and
the
CID
is
referring
to
a
Json
that
holds
this
structure.
Then
you
have
less
load
on
the
server
itself
and
the
clients
can
update
themselves,
so
they
are
directly
connected
to
the
back
end.
And
this
way
you
can
offload
the
server
a
bit
and
you
can
actually
get
more
clients
into
a
multiplayer
game
with
this.
C
So
there
are
drawbacks
when
it
comes
down
to
how
fast
you
can
fetch
and
load
the
data,
but
usually
in
games,
it's
not
really
important
to
load
data
in
a
fraction
of
a
second.
You
have
the
time
to
load
it
asynchronously
and
load
it
and
show
it
or
as
a
good
game
designer.
You
have
the
time
in
the
lobby,
for
example,
to
to
pre-load
everything
and
just
cache
it
on
the
drive
of
the
client
and
then
flush
it
out.
Once
the
the
session
is
over.
B
I
was
I
was
just
going
to
say
actually
I
think
part
of
the
reason
as
well.
Why
maybe
there's
been
quite
a
bit
of
difficulty
for
web
3
games
in
terms
of
General
traction
as
well
as
not
just
because
of
the
technical
hurdles,
but
also
some
studios
kind
of
maybe
feel
the
need
to
over
utilize
certain
blockchain
elements
into
a
game
when
that
doesn't
necessarily
need
to
be,
which
then
also
can
lead
to
technical
challenges
and
Technical.
C
B
Yeah
and
that
in
itself
obviously
has
points
where
you
hit
difficulties
you
hit
Wars,
but
if
you
maybe
try
to
reinvent
that
process
to
actually
understand
what
elements
of
a
game
necessarily
can
be
used
in
a
blockchain
applicable
way,
so
items
Collectibles
things
that
are
obviously
transferable
in
a
Marketplace
that
you
can
give
ownership
rights
to
pass
it
to,
then
you
can
start
to
maybe
overcome
some
technical
challenges.
Storage
hurdles
Etc,
but
you
shouldn't
have
to
shoehorn
blockchain
into
a
game.
A
Totally
and
I
was
just
gonna
add
that
it
on
the
off
chain
side
of
things.
It's
our
end
of
the
bargain
folks,
like
web3
storage
and
other
service
providers,
or
even
you
know,
people
working
on
the
core
ipfs
implementations
to
make
it
easy
to
easy
enough
to
use
and
perform
it
enough.
So
folks
can
do
a
lot
of
this
stuff
off
chain,
but
still
extend
those
guarantees
that
you
get
on
Chain
by
being
able
to
use
ipfsc
IDs
in
terms
of
linking
the
on-chain
and
off
chain
components.
A
So
yeah
I
mean
it's.
It
is
kind
of
our
goal
to
anytime.
Someone
doesn't
want
to
run
their
own
ipfs
node
and
manage
that
infrastructure.
There
are
tons
of
advantages
to
doing
that
in
some
circumstances,
but
still
just
like
we
see
with
cloud
services
today,
A
lot
of
times,
you
don't
want
to
have
to
run
that
stuff
yourself.
We
tried
to
make
it
really
really
easy
for
folks
to
be
able
to
just
rely
on
us.
B
I
think
the
more
partners
that
you
have
in
the
space
of
obviously
the
better
like
we
live
in
a
decentralized
world.
We
have
many
different
cloud
storage
solutions,
so
I
think
it's
also
providing
the
option,
like
you
said,
to
developers
and
to
users
as
well
when
they're,
initially
building
out
that
stack.
A
A
In
the
token,
the
auth
token
that
it's
actually
this
vote
of
that
person
requesting
that
service
to
be
invoked
and
then
the
other
really
cool
thing
about
it
is
you
can
delegate
permissions
to
other
actors
with
a
key
pair
to
then
interact
with
your
service
directly
on
your
behalf,
so
you
end
up
in
this
situation.
It's
similar
to
the
topic
earlier
about
once
you
have,
you
know
solved
some
of
the
base.
Layers
of
you
know:
fundamental
needs
of
of
developers
like
reliability
and
speed.
A
You
can
get
to
touching
on
other
things
web
through
store
or
web3.
Just
does
better,
so
by
being
able
to
delegate
ucan
tokens
to
other
actors
and
have
them
interact
with
your
service
directly.
You
create
a
much
more
efficient
data
pipeline
to
getting
data
stored
and
retrieved
through
through
the
web.
So
you
no
longer
have
to
stand
up
a
backend
just
to
proxy
data
into
the
service
that
you're
uploading
to
your
user.
C
Oh
yeah,
exactly
that.
That's
what
I
wanted
to
mention
that
currently
we
can
communicate
with
the
ipfs
directly
from
the
clients
and
fetch
any
data,
but
when
it
comes
out
to
uploading
to
ipfs,
we
have
to
relay
through
the
servers
because
of
the
security
issues,
because
if
you
are
shipping
the
game
with
with
the
access
token,
then
someone
can
get
this
access
token
out
of
the
source
code
of
the
game,
and
then
he
can
use
it
in
malicious
way
and
just
you
know,
upload
an
infinite
amount
of
information
into
your
service.
C
So
you
cans
are
the
the
missing
link
that
we
are
waiting
for
to
actually
offload
all
the
communication
with
ipfs
onto
the
clients.
So
it
will
be
a
two-way
communication
directly
with
the
web3
stack,
which
means
that
it
will
even
more
upload
on
the
dedicated
service,
which
means
that
we
can
be
able
to
build
a
lot
easier.
A
lot
bigger
games
in
future.
B
Some
of
the
ogs
might
disagree
with
me
there,
but
definitely
from
our
perspective.
We
also
have
to
be
honest
in
terms
of
the
fact
that
we're
not
yet
at
the
size
of
maybe
an
immutable
or
a
polygon
and
we're
working
towards
that,
but
that's
also
extremely
exciting
for
us,
because
there's
so
much
opportunity
in
this
space
and
really
what
we
want
to
do
from
nears
perspective
is
to
again
remove
those
barriers.
B
Allow
people
to
create
Without
Limits
play
Without
Limits
allow
people
to
engage
in
a
way
that
feels
convenient
and
seamless,
as
well
as
provide
the
tools
necessary
and
the
Gateway
is
also
necessary
for
people
to
come
on
board
and
feel
comfortable
to
do
that.
So
there's
a
huge
amount
of
space
here
for
development,
there's
a
huge
amount
for
growth
as
well.
A
Awesome
well,
I
see
The
Limited
power.
I
had
to
stop
you
all
from
eating.
Lunch
has
been
violated,
so
we
can
wrap
up
here.
Maybe
a
quick
10-15
seconds
from
each
of
us.
If
you
want
to
shout
anything
out
in
terms
of
your
socials
or
where
to
go
to
learn
more
information
for
your
products,
we
can
do
that
really
quickly.
Adam,
starting
with
you
I.
C
Just
wanted
to
say
that
we
are
also
what
people,
tutorials
and
just
educational
materials
for
the
community,
and
this
is
very
important
to
just
educate
people
on
how
easy
it
is
to
use
this
stack
because
they
hear
all
the
rumors
about.
Oh
yeah,
web3
is
bad
for
gaming.
It's
actually
not
it's
not
harder
to
make
Webster
games
it's
easier
and
you
just
have
to
try
it
to
to
find
out
about
it
yourself.
B
B
A
A
U
s
e
slash
just
find
me
around
the
next
few
days,
I'll
be
here
and
would
especially
be
excited
to
to
talk
to
you
if
you're
looking
to
offload
ipf's
infrastructure
that
you're
running
yourself
to
and
just
be
able
to
interact
with
an
API
instead
of
having
to
do
that
and
check
out
our
blog
too,
at
blog.web
through
that
storage,
and
that
represents
the
latest
and
greatest
of
what
we're
up
to
all
right.
Thanks.