►
Description
Dirk will demonstrate an incredible new technique for bridging Filecoin content directly into IPFS.
A
So
yeah
I'm
talking
about
bit
swap
in
boost
so
can
I
have
a
show
of
hands.
People
who
know
what
Lotus
is
yeah
got
a
few
people.
Okay,
so
I'm
going
to
talk
a
little
bit
about
that
as
well.
So
basically,
what
we're
trying
to
do
is
Bridge
the
ipfs
network
with
the
filecoin
network,
which
is
something
we've
kind
of
been
waiting
to
do
for
about
two
years.
So
we're
very
excited
about
this,
so
I'm
going
to
talk
about
what
is
Boost.
But
first,
let
me
talk
about
Lotus.
A
A
So
last
year
we
broke
off
that
last
piece.
The
data
transfer
piece
and
we
put
it
in
a
new
binary,
called
boost,
so
this
makes
it
a
bit
easier
to
manage
one
of
the
most
complicated
pieces
of
the
filecoin
network
today,
so
yeah
I
guess
I've
put
here
that
it's
a
storage
and
retrieval
front
end
to
Lotus.
A
So
we
have
this
thing
called
an
auto
retrieve
node,
so
ipfs
connects
to
the
auto
retrieve
nodes
over
ipfs's
native
transport
protocol
called
bitswab,
and
then
Auto
retrieve
connects
to
filecoin
over
filecoin's
native
retrieval
protocol
called
graph.
Sync.
So
obviously
there's
some
inefficiencies
there
from
translating
between
the
protocols
and
it's
also
the
discovery-
is
a
little
tricky
because
the
ipfs
node
has
to
be
connected
to
Auto,
retrieve
node,
in
order
for
them
to
be
able
to
talk
to
each
other
and
there's
no
Discovery
mechanism.
A
A
A
So
when
a
filecoin
storage
provider
receives
a
storage
deal,
it
receives
data,
it
creates
an
index
of
all
of
the
blocks
in
that
data
and
it
tells
an
indexing
node
Hey
I've
I've
got
all
of
these
blocks
and
then,
when
clients
want
to
know
where
a
block
is
in
the
network,
they
just
query
the
indexer.
They
say
hey,
who
has
this
collection
of
blocks
and
the
indexer
says
these
storage
providers
so
talking
specifically
about
Kubo
in
Kubo
V
0.14
they've
added
a
new
protocol
called
reframe.
A
So
traditionally,
ipf
has
found
stuff
on
the
network
by
asking
the
DHD
the
distributed
hash
table.
So
with
the
reframe
protocol
we
can
configure
Kubo
to
instead
talk
to
our
indexer,
so
Kubo
queries
indexer.
It
asks
who
has
my
blocks.
Kubo
says
these
storage
providers
have
the
blocks,
or
these
endpoints
and
Kubo
downloads
over
bit,
swap
in
parallel
from
the
storage
providers.
A
Okay,
so
this
is
a
video
that
I
uploaded
a
few
days
ago.
It's
about
a
big
bunny
I
recommend
you
watch
it.
It's
pretty
fun
and
I'm
going
to
give
you
the
Cid
in
a
second,
so
you
can
download
it
over
ipfs.
A
This
is
the
interface
that
people
using
boost
C,
that's
the
web
interface.
So
you
can
see.
This
is
a
list
of
downloads
that
I've
done
over
the
last
week
in
our
production,
Miner
in
Bulgaria
and
sorry,
oh
yeah,
sure.
A
I
mean
the
text
is
the
same
size
but
yeah
yeah,
it's
just
it's
just
a
web
interface
showing
like
a
list
of
deals.
Basically
so
at
the
moment
we're
having
some
issues
with
Estuary,
so
not
so
many
successful,
but
once
they,
it
sort
of
shows
you
the
stage
that
a
deal
goes
through
once
it
gets
into
filecoin
like
it
goes
through
a
data
transfer,
it
gets
verified
and
then
goes
into
the
ceiling
state.
A
A
So
the
first
question
is:
how
do
I
find,
where
my
see
ideas
and
the
answer
is
I-
go
to
the
indexer
and
I
paste
into
this
box
over
here,
the
CID
that
I
want
click
on
contacts,
and
it
tells
me
okay.
This
prid
at
this
address
with
the
bit
swap
protocol,
has
that
piece
of
data,
and
also
it's
also
exposed
over
the
graph
sync
protocol.
A
So
this
is
automatically
done
now
by
the
reframe
protocol,
but
essentially
what
the
reframe
protocol
is
doing
behind
the
scenes
is,
it
does
I,
don't
know
if
you
can
read
that
it
does
an
ipfs
swarm
connect,
meaning
connects
my
my
node
on
my
laptop,
my
apfs
node
to
the
ipfs
nodes
on
our
production,
Miner
and
then
ipfs
get
so
what
you're
looking
at
there
is
ipfs
talking
through
bit
swap
to
a
storage
provider
in
production.
For
the
first
time
in
history,
can
we
make
some
noise
foreign.
A
Yeah,
so
you
can
see
this
is
the
same
video
that
I
uploaded
a
few
days
ago,
cool.
A
Okay,
so
how
do
we
motivate
storage
providers
to
actually
run
booster
bit
swap
so
there
is
a
cost
for
them.
They
have
to
it's
a
resource
cost
in
terms
of
actually
running
the
server,
and
they
also
have
to
pay
egress
costs.
So
we're
currently
working
on
incentivization
for
storage
providers,
and
that
was
what
the
the
summit
was
about
yesterday,
the
retrieval
incentive
Summit,
but
for
the
time
being,
we're
going
to
say
that
anyone
who's
part
of
the
Evergreen
program
needs
to
run
booster
bit
swap
as
a
requirement
and
story.
A
So
next
steps
so
today,
file.
Coin
storage
providers
can
announce
a
single
address
on
chain,
and
so
any
client
that
wants
to
talk
to
a
storage
provider
just
has
to
go
and
get
that
one
address,
and
it
means
that
all
of
our
requests
have
to
be
proxied
through
a
single
node.
So
we're
going
to
update
the
protocol,
so
you
can
add
different
addresses
for
different
transport
protocols.
So
you
can
have
one
for
graph
sync,
one
for
bit
swap
Etc
and
we're
going
to
scale.
A
So
what
I've
shown
you
today
is
mostly
proof
of
concept
it
still.
It
works
quite
well,
but
we
need
to
scale
out
to
you
know
the
kind
of
load
that
the
internet
is
going
to
put
on
it,
and
next
is
for
you
guys
to
show
us
what
you
got
so
we're
super
excited
now
that
we've
connected
the
ipfs
network
to
the
filecoin
network
to
see
what
all
of
you
come
up
with
and
the
exciting
things
that
you're
going
to
be
building
on
top
of
it.
Thank
you.