►
From YouTube: Saturn, A Web3 CDN - Ansgar Grunseid
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
Good
afternoon,
ladies
and
gentlemen,
hello
from
beautiful
amsterdam,
I'm
onsgard
grunside.
Today
we
are
going
to
take
a
journey
together
into
the
solar
system
and
not
just
any
planet,
but
everyone's
favorite
planet
and
soon
to
be
the
world's
most
favorite
planet
saturn.
So
what
we're
going
to
discuss
today
in
this
talk
is
a
new
effort
to
build
the
world's
best
cdn.
We
call
saturn
and
it's
not
just
any
kind
of
cdn.
A
It's
a
web
3
cdn
and
what
we're
going
to
go
through
is
we're
going
to
talk
about
what
a
cdn
is,
why
they
exist
and
where
there
is
this
momentous
opportunity
to
redesign
and
rebuild
what
a
cdn
is
using
all
the
new
primitives
and
tech
that
we
have.
That
web3
brings
so
with
that.
Please,
if
any
of
you
guys
have
any
questions,
please
jump
in
as
we
jump
into
this
together.
A
A
A
So
the
very
first
cdn
the
progenitor
of
cdns,
was
akamai
built
by
a
bunch
of
mit
grads
in
the
early
2000s,
and
the
problem
they
set
out
to
solve
is
how
do
you
scale
a
service
like
instagram,
where,
if
you
have
a
billion
users
all
around
the
world
and
they're
all
trying
to
access
the
same
content,
they
can't
all
hit
the
same
services
they're
the
same
servers.
Those
servers
can't
scale.
It's
imagine
like
you.
If
you're
the
server
trying
to
hold
a
conversation
with
one
person,
that's
doable
two
or
three
people.
A
That's
doable
all
of
a
sudden
you're
trying
to
hold
a
conversation
with
a
billion
people.
You
can't
do
it
same
thing
with
the
server.
They
can't
scale
that
much
so
the
problem
that
akamai
and
other
early
cdn
set
out
to
solve
is
what
can
we
distribute
in
scale
to
power
a
website?
So
when
you
access
a
website,
that
website
is
comprised
of
many
things.
It's
comprised
of
text.
A
It's
comprised
of
images,
video
elements
like
that
and
the
biggest
and
heaviest
images
and
assets
are
actually
retreat,
the
same
and
retrieved
by
retrieving
the
same
manner
by
everybody.
So
if
you
go
to
instagram
and
you
link
someone
an
instagram
post,
they're
gonna
see
the
same
picture
that
you
see
and
that
same
data
is
going
to
flow
to
them
as
it
flows
as
it
flowed
to
you
and
so
akamai,
and
these
cdns
realized
this
and
they
said,
wait
a
second.
What
if
we
can
take
that
image,
that
everyone
is
loading
and
distribute
it
globally?
A
So,
instead
of
that
one
image
being
stored
on
the
one
instagram
server,
what
if
we
copy
it
and
we
place
it
all
over
the
world
and
thousands
of
servers
close
to
end
users
and
then,
when
I
load
the
instagram
page,
I
will
load
that
image
from
a
server
close
to
me
and
when
someone
else
in
india
or
chile
loads
that
image
they'll
load
the
same
image
but
from
a
much
closer
copy
and
a
server
closer
to
them.
And
that
is
really
the
whole
idea
of
a
cdn.
A
A
A
These
unlock
new
and
larger
networks,
the
network
we're
talking
about
saturn,
has
the
potential
to
be
the
largest
distribution
network
of
content
in
human
history,
and
we
have
and
we're
going
to
get
there
and
we're
going
to
get
through.
Why
now?
Why
are
we
starting
to
build
this?
Where
do
we
see
this
opportunity?
Well,
there's
a
huge
opportunity
with
falcoin's
network,
specifically
where
filecoin
has
assembled
a
massive
storage
network
of
16
exabytes
and
rising,
but
it's
really
locked.
A
What
we
can
do
with
the
retrieval
market
and
saturn
specifically,
is
take
all
the
data
that's
stored
in
filecoin
and
distribute
it
globally,
so
that
it's
fast
and
cheap
to
access
everywhere
and,
most
importantly,
we
can
do
this
and
take
advantage
of
market
dynamics
where
all
cdns
you
know
that
existed
prior
web
2
cdns
they
grow
as
like
a
centrally
managed
organism
grows.
Somebody
in
a
boardroom
in
a
business
room
wearing
a
suit
somewhere,
gets
out
their
checkbook
and
says
we
need
to
build
another
data
center.
A
Here
and
they
go
and
build
that
data
center,
that's
a
completely
old-fashioned
and
different
model
than
the
model,
we're
bringing
with
web3,
where
it's
more
like
a
dynamic
market,
a
marketplace
where
there's
supply
and
demand
where
the
market
can
see,
there's
an
opportunity
for
new
servers
to
be
added,
let's
say
in
mumbai,
and
that
market
opportunity
is
recognized
by
someone
who
has
servers
in
mumbai
and
they
can
connect
them
and
join
the
greater
network.
And
there
was
no
central
coordination
here.
A
A
So
how
are
we
going
to
go
about
building
this?
Let's
zoom
back
in
now,
so
we
kind
of
covered
what
a
cdn
is
and
some
of
the
web
three
primitives
we're
gonna
use
to
build
saturn.
How
are
we
gonna
go
building
it
well.
One
of
the
key
pieces
here
is
when
you
don't
have
a
central
organization
running
everything
and
you
allow
anyone
to
jump
in.
A
You
need
to
build
trustless
permissionless
systems,
and
this
is
a
key
piece,
because
if
anyone
can
plug
their
servers
in
that
server
operator,
mumbai
plugs
the
servers
in
who's
to
say
that
either
through
malice
or
incompetence
he's
going
to
screw
something
up.
You
know
you
request
cat
about
ping,
your
favorite
kitty
picture
and
you
get
back
dog
dog
ping,
someone
else's
favorite
dog
picture.
We
need
to
forfeit
against
that
kind
of
attack
or
or
network
failure,
and
the
other
key
piece
here
is
the
scalability
of
a
decentralized
network.
A
So
again
with
all
the
web
2
cdns
there
is
one
company
and
one
brain
or
one
set
of
brains,
thinking
about
how
and
where
their
network
should
grow.
We
want
to
race
past
that
we
want
to
change
the
game
here,
where,
in
a
decentralized
manner,
anyone
can
see
these
opportunities
and
plug
in
there's
some
other
key
tenets.
Here.
All
of
our
work
is
open
source
and
will
be
open
source.
A
Anyone
can
jump
in
anyone
can
work
on
it
and
we
welcome
that
and
the
whole
network
can
improve,
and
the
last
piece
here
is:
why
should
you
participate?
And
the
answer
is,
you
can
be
rewarded?
If
you
have
spare
capacity,
you
can
add
that
to
the
network
and
reap
the
rewards,
you
can
improve
other
people's
experience
online
and
make
the
internet
faster
and
receive
rewards
in
turn,
and
that's
a
completely
different
model
than
every
other
cdn
has
been
built.
A
Now,
let's
jump
into
some
of
the
technical
pieces
here
of
how
we're
building
saturn.
So,
as
I
mentioned
before,
one
of
the
key
facets
of
saturn
is:
it
is
trustless,
which
means
anyone
can
plug
their
infrastructure
into
the
network
and
start
serving
requests
to
end
users
and
making
their
experience
faster.
A
A
A
You
know
the
content
id
of
that
youtube
video
and
you
know
how
to
verify
that
youtuber
video
as
the
youtube
video
as
you
receive
it
so
as
they
send
it
to
you,
you're,
incrementally
verifying
that
every
frame,
as
is
rendered
in
your
browser,
is
the
right
frame
and
not
necessarily
like
a
random
dog
picture
as
one
of
the
frames-
and
this
is
fundamentally
new-
there
are
no
cdns
that
do
this.
Every
cdn
you
use.
Currently,
you
have
to
trust
the
other
party
to
send
you
the
information
you
request.
A
A
If
you
plug
in
a
smaller
server
connected
to
much
faster
bandwidth,
you
could
be
an
endpoint
that
many
users
connect
to
when
you
serve
as
a
hot
cache
for
those
users
or
if
you're,
just
on
your
local
desktop
machine,
you
could
install
a
saturn
desktop
app
and
bring
your
storage
to
the
network
when
you're
online,
in
a
manner
that
you
want
and
you're
fully
in
control
over,
and
you
also
receive
rewards
and
in
the
future.
We
would
love
to
unite
browsers
together
too,
so,
even
without
having
to
install
anything.
A
You
could
contribute
to
the
network
and
receive
content
from
the
network
and
appear
to
be
your
browser,
peer-to-peer
manner
and
the
reason
we
want
to
do
this
is
imagine
airbnb.
One
of
the
reasons
that
airbnb
has
been
so
successful.
Is
they
didn't
just
focus
on
large
mansions?
They
didn't
focus
on
beachside
resorts.
They
let
the
world
add
all
this
capacity,
we're
from
small
houses
and
kept
all
the
way
up
to
full-scale
castles
and
connected
them
with
an
audience.
A
A
There's
this
dynamic
supply
and
demand
where
the
network
scales
itself
and
again
as
far
as
scaling
and
scaling
speed,
because
it's
permissionless,
you
don't
need
anyone's
permission
to
add
capacity.
No
one
needs
to
sit
there
and
say
yeah,
you
can
add
a
server
in
mumbai
or
you
can't.
You
know
you
can
add
it
to
the
network
and
get
rewarded
justly
so
what
we've
covered
so
far,
some
of
the
high
level
ideas
of
things
that
have
changed
and
new
technology.
We
have
to
build
a
fundamentally
different
cdn
one,
that's
trustless!
A
Where
anyone
can
you
know,
join
the
network
and
send
content.
That's
verifiable
to
end
users,
one
that's
permission
list
where
you
don't
need
anyone's
permission
to
add
your
capacity
to
the
network,
one
where
you
receive
rewards
in
no
matter
the
device.
You
have
no
matter
the
capacity
you
have
when
you
add
it
to
the
network
and
this
open
source
software
that
will
tie
it
all
together
and
the
crypto
incentives
that
unite
everyone
into
the
same
boat.
So
at
a
high
level.
A
B
I
can
start
us
off
sure
badly
from
your
experience
building
arc
and
then
starting
off
building
saturn.
What
would
be
your
takeaways
in
building
something
which
is
essentially
much
more
there's
much
more
power
in
the
hands
of
the
user
and
anyone
can
join
the
network?
What
are
the,
what
are
the
sort
of
things
you've
learned
from
arc
which
will
come
to
assassin?
That's
a
great.
A
Question
the
biggest
takeaway,
the
biggest
lesson
learned
is
one
of
scalability,
which
is
that
when
you
don't
have
to
have
some
sort
of
centrally
planned
or
essentially
managed
team
that
decides
how
and
where
the
network
grows,
the
network
can
grow
much
much
faster,
and
so
we
saw
this
with
arc
a
giant
browser
peer-to-peer
network.
You
know
we
were
able
to
scale
from
zero
to
100
million
nodes
in
our
network
in
a
little
over
a
year
which
is
unprecedented
and
one
of
the
reasons
we're
able
to
do.
A
That
is
because
you
know
like
no
one
was
sitting
there
deciding
who
and
where
to
add
capacity
again,
there's
no
one
with
the
decision,
responsibility
of
getting
out
their
checkbook
and
writing
a
check
to
build
a
data
center
somewhere,
the
owner
of
that
data
center.
The
existing
infrastructure
there
can,
of
their
own
volition,
decide
to
join
to
decide
to
join
the
network.
So
I'd
say
that's
the
biggest
takeaway
one
simply
of
scale
and
speed.
B
B
For
you
mentioned
that,
there's
these
l1
and
l2
nodes
in
the
network.
What
are
the
do?
You
have
off
the
top
of
your
head,
the
hardware
requirements
or
the
system
spec
for
these
two
different
levels
of
the
of
the
network.
A
A
You
know
order
magnitude,
hundreds
of
gigabytes
to
terabytes,
and
we,
you
know,
also
want
to
have
other
hardware,
which
is
more
user-facing,
where
you
need
much
faster
internet.
Let's
say
like
a
gigabit
per
second
connection
and
anywhere
from
you
know
a
few
terabytes
of
cache
attached
all
the
way
up
to
a
full
rack
of
servers.
So
any
really
anything
in
between.
B
Okay,
great
thanks
and
then
another
question
is,
I
think,
well,
I
know
there's
a
test
network
already,
which
has
been
spun
up
privately.
When
will
the
test
network
be
readily
available
more
widely
and
when
will
people
be
able
to
join
the
network?
Another.
A
Wonderful
question
so
right
now,
saturn
is
in
testing
it's
not
in
production
testing
yet,
but
the
tests
are
going
splendiferously
and
I
don't
have
a
specific
time
we
can
announce
publicly
yet
of
when
you
can
get
your
hands
around
a
saturn,
node
and
download
the
software
and
run
it
on
one
of
your
servers
or
run
it
on
your
desktops.
I'm
afraid
I
don't
have
that
yet.
So
keep
your
eyes
out
and
we'll
be
sure
to
let
you
know
be
sure
to
announce
the
up-to-date
timelines
as
we
have
them.
B
Great
and
then
the
final
question
which
came
in
from
the
live
stream,
which
I
think
I
might
answer,
which
is
just
when,
will
the
reward
model
or
reward
structure
for
saturn
miners,
be
released
and
we're
having
a
cryptoeconomic
workshop
after
the
live
stream.
Today,
where
we're
going
to
be
working
on
this
further,
and
it
will
be
most
likely
in
the
second
half
of
this
year
that
we'll
have
something
really
concrete
to
share
with
people.
C
It
baby.
Yes,
you
mentioned
sorry
it's
on
yeah
incentives,
so
I
would
say,
like
probably
for
a
web
2,
the
end
user
has
really
no
incentive
in
terms
of
a
cdn
other
than
the
service
working,
which
is
huge
in
terms
of
web
3.
C
A
Two
pieces
two
answers.
The
first
one
is
in
the
future:
we'd
love
to
have
the
actual
user,
become
a
part
of
the
network
and
start
sharing
content
back
out.
So
in
that
model
which
we're
not
at
yet,
but
in
that
model
the
user
could
also
receive
rewards
for
everything
they
share
back
with
the
network,
so
imagine
kind
of
like
bittorrent
or
like
arc
where
once
you've
downloaded
a
piece
of
content.
A
You
can
begin
seeding
that
and
sharing
that
content
out
with
others
near
you
in
the
network
and
in
that
model,
when
we
cross
that
bridge,
you
will
receive
rewards
for
the
sharing
you've
done
with
the
network.
That's
one
piece
of
it.
The
other
piece
of
it
is
imagine
right
now.
When
you
access
a
website
like
youtube,
you
don't
pay
anything
in
terms
of
like
a
monetary
exchange.
You
don't
open
your
wallet
and
pay
youtube
three
cents
for
every
youtube.
Video.
A
You
watch,
you
know
youtube
monetizes
you
and
then
takes
that
money
to
pay
for
the
cdn
service.
You
know
you,
as
a
user,
don't
pay
for
the
cdn.
There
is
another
model
here
where,
with
web3
and
micro
transactions,
users
could
pay
to
download
the
content,
so
they
themselves.
You
know
the
end
user
pays
and
in
that
manner
they
could
be
incented
and
say
I
will
pay
more
for
loading
faster
or
more
for
better
performance
and
in
that
manner
they
could
be
involved
in
this
kind
of
the
protocol
talking
with
the
cdn.
A
Right
exactly
right,
so
those
are
two
kind
of
different
answers
to
your
question
of
where
things
could
go,
but
the
very
first
version
of
saturn
will
be
built
for
what
the
market,
how
the
market
operates.
Right
now,
which
is
websites
like
youtube
they're,
the
ones
who
are
paying
for
content,
delivery,
they're,
the
ones
buying
cdn
capacity
and
the
end
users,
don't
pay
anything.
So
how
that
model
will
change
in
the
future
remains
to
be
determined,
but
those
are
two
potential
ways
it
could,
where
the
user
would
be
part
of
the
either
earning
rewards
or
paying.
D
I
covered
a
little
bit
in
the
talk
earlier,
but
already
brave
has
bat
and
bat
has
individual
users
earning
for
the
attention
that
they're
giving
to
various
websites.
You
can
imagine
hooking
up
bat
with
automatic,
defy
swaps
to
falcoin
and
paying
for
the
retrieval
live.
So
a
user
is
viewing
a
page
with
brave
part
of
the
bat
that
they
would
be
receiving
for
viewing.
That
page
would
then
go
to
the
cdn
that
is
delivering
the
content,
and
that
would
happen
automatically.
D
But
if
the
browser
does
it
for
them
automatically,
then
it
could
work,
and
so
I
think,
like
thinking
of
integrating
with
things
like
brave
or
pushing
that
the
bat
model
to
many
other
browsers,
I
think
could
be-
could
be
pretty
successful.
There.
E
I
have
a
question
too,
I
feel
like.
I
think
it's
great,
I
think,
there's
lots
of
problems
with
like
web3
cdns,
but
I
feel
like
after
the
presentation.
I
still
don't
have
a
very
concrete
grasp
of
what
exactly
saturn
is
and
how
potentially
differentiate
from
let's
say
like
miles
or
like
the
other,
even
like
titan.
Could
you
share
a
bit
more
light
on
that?
Please,
yes,.
A
I'm
not
sure
so
much
about
titan,
I'm
not
too
knowledgeable
about
titan
and
titan's
effort.
As
far
as
mayell
they'll
be
here
in
two
days,
and
I
don't
want
to
put
too
many
words
in
their
mouth,
but
one
of
the
differences
is
that
mayell
wants
to
pursue.
Is
they
want
to
have
end
users
paying
so
in
initially
for
saturn
again,
the
who's
buying
cdn
capacity?
A
Is
the
service
provider
like
instagram
or
youtube
they're
paying
for
cdn
and
miles
model
is
one
of
the
kind
of
half
of
the
answer
with
craig's
great
question,
which
is
mile,
wants
the
web3
world
where
it
was
building
towards
a
world
where
the
user
ends
up
paying,
and
so
it's
the
same
kind
of
set
of
opportunities
or
changes
where
a
user
now
has
to
kind
of
have
a
wallet
and
be
able
to
pay
for
content.
Yeah
juan
might
be
able
to
answer
some
of
this
better.
B
Yeah,
I
can
give
an
answer
to
that
as
well,
so
in
terms
of
how
the
value
flows
around,
let's
take
the
the
web
2
setup
where
content
publishers,
content
providers,
they
pay
cloud
services
to
host
files
and
host
websites
and
then
accelerate
content
and
end
users
are
essentially
getting
free
retrievals,
as
we
kind
of
spoke
about
in
the
break,
is
not
a
100
free
because
there's
this
idea
that
then
they'll
purchase
something
from
if
they're
browsing
a
marketplace,
they'll
purchase
something
or
they'll
watch
adverts
and
there's
sort
of
a
then
a
circular
economic
flow
back
to
the
content
providers.
B
So
the
what
mile
folks
did-
and
they
said-
okay,
let's
do
what
they
called
worse,
is
better
or
less
is
more.
It's
like.
B
However,
what
that
means
is
that
you
you're
you
do
have
to
kind
of
not
have
free
retrievals
for
loads
and
loads
of
clients,
because
then
you're,
subject
to
these
stable
attacks,
where
you
can
spin
up
loads
of
different
clients
and
try
and
make
all
these
retrievals
and
then
try
and
convince
the
content
provider
that
you've
done
all
this
retrievals
and
they
then
need
to
pay
for
it
all.
B
So
it
does
put
an
onus
on
the
client
to
have
to
then
pay
for
a
transaction
or
just
govern
the
whole
transaction
in
that
local
in
the
local
part
of
the
network.
So
yeah,
it's
it's
in
that
way.
Again,
it's
more
web3
focused
it's
just
between
you
and
me.
If
we
want
to
do
a
transaction
and
no
one
else
is
measuring
it
or
governing
it
or
kind
of
deciding
what
the
payouts
are
based
on.
B
It's
just
that
the
transaction
happened,
and
that's
that
done
and
that's
where
I
think
saturn
is
different,
because
the
premise
there
is
that
people
want
to
retrieve
for
free
into
their
browsers
in
the
same
way
they
do
in
web
2
and
therefore
there
has
to
be
this
economic
flow
from
content
providers
in
the
end
and
therefore
it's
it
kind
of
breaks
this.
The
model
that
mayell
going
for
of
just
focusing
on
that
one
that
one
transfer.
D
Yeah,
that
was
a
great
great
summary.
One
thing
I
might
add
is
the
mild
pop
structure
might
work
between
retrieval
providers
and
themselves.
So
one
if
you
think
back
to
the
internet
grapevine
model,
you
could
have
a
mile
pop,
that's
very
close
to
the
users
paying
for
and
retrieving
data
from
another
mile
pop,
that's
further
up
in
the
grapevine
to
bring
down
data
and
then
eventually
to
a
storage
provider.
D
So
you
could,
you
could
use
the
mild
pops
as
like,
really
good
links
in
between
in
that
grapevine
layer
to
have
the
individual
kind
of
pairwise
transaction
be
accounted
economically
and
so
that
you
can
kind
of
build
up
really
well,
if
you
then
figure
out
a
way
of
getting
falcon
to
the
end
users,
who
are
then
paying
for
the
content,
and
so
there's
probably
a
few
approaches,
one
one
approach
might
be.
D
You
can
build
products
where
you
download
and
install
a
kind
of
a
cache
that
makes
your
browsing
go
faster
and
you
might
potentially
pay
for
that
or
you
can
have
the
browser.
Bat
model
hook
directly
into
my
o,
and
so
you
could
have
like
bat
swapped
for
fat
coin
and
then
going
through
the
mail
network.
I
think
in
the
saturn
case
it's
the
publisher
will
sign
up
to
have
some
content
accelerated
this
maps
closer
to
the
traditional
cdn
model,
where
the
content
publisher
is
trying
to
accelerate
some
subset
of
content.
D
They
set
up
a
relationship
with
saturn,
they
pay
saturn
and
then
saturn
orchestrates
the
network
to
deliver
that
content
quickly
to
the
regions
where
that
content
is
being
retrieved
from
it's
just
two
different
adjacent
tangential
models
that
I
think
both
can
both
can
really
work,
there's
probably
other
content
models
that
are
not
yet
explored.
Well,
I
think
we'll
probably
see
a
lot
of
experimentation
across
these
and
over
time,
we'll
yeah
we'll
see
kind
of
what
happens
in
the
inner
parts
of
the
network.
D
My
sense
is
that
if
we
can
have
good
interfaces,
then
all
these
networks
can
interrupt
and
pay
each
other
and
so
on,
but
I
guess
that
yeah
we're
meant
to
figure
out
the
interfaces
there.
D
Yeah,
I
think
part
of
what's
was
getting
at
is
like
hey.
We,
that
was
a
good
overview.
It
was
kind
of
like
what
patrick
and
I
talked
about,
but
I
think
there's
no,
like
concrete
sense
of
saturn
itself.
It
may
be
useful
to
give
an
example
demo
of
like
how
arc
works
today
and
how
browsers
and
just
talk
through
like
load
up
an
arc
page
and
then
talk
through
like
how
browsers
make
those
requests
now
and
that
might
help
close
some
of
the
some
of
the
questions.
A
I
worry
going
through
arc.
Might
is
the
core
your
question?
Let
me
let
me
not
rephrase
it.
Would
you
mind
expounding
a
bit
on
what
to
be
what
you'd
like
to
see
explained
as
far
as
saturn's
architecture
would
go
diego's
going
to
cover
that
summon
the
next
talk
and
going
to
bid
into
the
kind
of
network
diagram
of
like
multiple
levels
of
the
cache
and
yeah.
E
Sure
we
can
wait
for
the
next
talk
and
then,
like
I
mean
what
was
getting
a
it's
like,
oh
yeah,
I
understand
all
the
high
level
is
pretty
much
similar
across
the
board
and
I
want
to
understand
like
oh
how
how
like
what
like
where's
the
current
status
of
saturn
right
now
and
our
roadmap
go
to
market
and
all
that
kind
of
stuff.
But
if
you're
covering
the
next
talk,
that's
fine
dude.
A
Great
yeah,
so
not
all
that
will
be
covered
in
diego's
next
talk.
So,
let's
see
some
technical
architecture
of
how
saturn
currently
works-
and
this
is
already
built
and
running,
is
so
in
the
browser.
There's
a
new
api
called
the
service
worker,
which
is
like
a
javascript
context
that
runs
beneath
the
parent
page.
It
has
a
superpower
and
the
superpower
of
the
service
worker
is,
it
can
see
and
handle
requests
made
by
the
parent
page.
A
Please
give
me
this
content
behind
this
sid
as
quickly
as
possible,
and
then
what
happens
is
that
request
is
routed
via
dns
kind
of
the
nearest
point
of
presence
of
the
saturn
network,
and
that
would
likely
be
someone
who's
plugged
in
a
bunch
of
servers
in
capacity
near
you
in
the
network
running
saturn
software.
This
will
generally
be
what's
known
as
like
an
l1
or
what
we're
referring
to
is
l1
kind
of
layer
one
in
the
cache.
A
So
these
will
be
machines
that
have
likely
running
in
a
data
center
that
have
a
lot
of
upload
capacity.
Uplink
capacity
and
large
cache
capacity,
so
you'll
be
routed
by
a
dns
to
one
of
these,
and
then
one
of
these
essentially
does
what
a
traditional
cdn,
node
or
point
of
presence
will
do.
It
will
look
for
that
item
that
sid
in
its
cache
and
if
it
has
it
will
return
the
car
file
of
that
file
to
you.
A
So,
instead
of
rather
than
like
the
plain
text
of
the
file,
it
will
return
the
car
file
to
you
and
on
cache
miss
it
will
go
to
like
an
l2
another
layer
in
the
cache.
This
could
be
some
of
the
desktop
nodes
or
some
servers
that
might
not
necessarily
have
as
fast
upload,
but
you
know
a
lot
of
disk
capacity
kind
of
a
second
layer
in
the
cache
and
on
onwards
that
you
know
cache
lookup
will
cascade
all
the
way
to
the
service
providers
who,
at
the
again
kind
of
the
nexus.
A
A
A
Let
you
know
let
anyone
around
the
world
participate,
but
that's
kind
of
the
first
strategy
is
there's
already
all
these
requests
flowing
into
the
ipfs
network,
they're,
currently
landing
at
the
ipfs
gateway
and
they're
not
being
served
as
well
as
they
can
be,
and
saturn
is
going
to
be
there
to
serve
them
better
than
they
currently
are.
So
that's
kind
of
the
first
step
does
that
help
sagacious
questions?
Thank
you.