►
From YouTube: Building IPFS, Filecoin, and the future of web3
Description
Join us for Filecoin Liftoff Week, an action-packed series of talks, workshops, and panels curated by the web3 community to celebrate the Filecoin mainnet launch and chart the network’s future. https://liftoff.filecoin.io/
Events take place all week, October 19-23, 2020.
For more information on Filecoin
- visit the project website: https://filecoin.io/
- or follow Filecoin on Twitter: https://twitter.com/Filecoin
Get Filecoin community news and announcements in your inbox, monthly: http://eepurl.com/gbfn1n
A
I
wanted
to
start
kind
of
back
at
the
at
the
beginning
when
I
joined
the
ipfs
team
and
everything
that
the
the
ips
team
has
been
doing
kind
of
over
the
past.
I
mean
years
at
this
point
but
past
a
while
that
has
laid
the
foundation
for
foulcoin
and
that
folkland
has
now
built
on
top
of,
and
so
really
thinking
about
how
a
lot
of
the
tools
and
technologies
in
ipva,
slippy2b
and
ipld
contributed
to
the
building
of
filecoin.
A
A
I
mean
you
know,
probably
honestly
long
before
fall
coin
was
it
was
even
a
a
major
project
with
the
the
ipfs
community
evolving
to
identify
that
sharing
content
and
being
able
to
find
it
was
like
a
really
critical
core
component
to
a
resilient
storage
network
and
so
the
the
existence
of
ipfs
and
then
the
nucleation
of
ipld
and
lid
p2p.
Out
of
it.
A
Just
straight
up
is
like
a
very
core
building
block
that
was
involved
in
the
the
growth
and
evolution
of
falcoin,
but
even
once
filecoin
kind
of
became
its
own
project
after
2017
and
started
having
its
own
development
team
and
building
its
own
open
source
community.
A
There
were
a
lot
of
ways
in
which
those
the
the
ties
of
what
was
being
built
in
ipfs
in
la
p2p
fl
flowed
into
what
falcone
could
utilize
and
benefit
from
as
a
technology
building
on
top
of
the
same
stack,
and
so
I
wanted
to
highlight
a
couple
of
the
different
things
that
have
happened
over
the
past.
A
You
know
year
and
a
bit
that
have
all
gone
in
and
been
utilized
by
the
falcoid
network
now
that
it's
kind
of
flowing
into
the
world,
and
so
one
one
that
kind
of
was
was
in
the
early
days
when
I
first
started
out
on
the
iqvs
team
was
grassing
and
so
back
in.
I
think
this
was
early
2019
ipfs.
A
The
the
main
data
transfer
algorithm
that
we
use
when
connecting
two
nodes
with
each
other
is
called
bitswap
and
bitswap
is
really
good
for
kind
of
session
management
being
able
to
have
a
one-to-many
or
many-to-many
data
exchange
kind
of
algorithm,
so
that
you
can
quickly
assemble
from
many
different
nodes.
A
The
the
data
that
you
want,
however,
not
super
great
for
deep
graphs,
because
every
request
every
time
that
you
go
down
a
layer
of
your
merkle
tree,
it
requires
another
round
trip
and
another
round
trip
to
get
the
children
and
the
children
and
the
children.
Children
now
think
about
a
really
deep
graph.
A
Like
a
blockchain,
you
end
up
in
a
pretty
rough
situation
where
just
the
the
round
trips
in
order
to
load
that
data
and
to
traverse
your
way
to
the
parents
and
the
parents
and
the
parents
of
the
current
block
would
become
prohibitive,
and
so
that's
where
the
concept
of
graphsync
and
ipld
selectors
came
in.
This
was
a
solution
that
would
benefit
all
blockchains.
That
would
utilize
things
like
ipfs
and
ipld,
but
was
was
also
motivated
around
other
use
cases
of
kind
of
deep
graphs
of
data.
A
A
That's
that's
what
graphsync
was
created
for,
and
it
was
actually
hannah
gave
an
awesome
talk
on
this
yesterday,
and
so,
if
you
want
to
understand
the
the
deeper
workings
of
that,
but
that
was
one
of
the
first
problems
and
challenges
taken
on
specifically
with
with
the
thought
in
mind
of
all
right.
How
could
blockchain
clients
utilize
these
core
modular
components
and
and
then
it
can
come
back
into
ipfs
as
a
more
more
general
use
case,
of
empowering
the
the
fetching
and
transfer
of
deep
graphs?
A
And
I
think
that
that
highlights
some
of
the
the
principles
that
work
in
play
here
as
well
around
having
very
flexible
modular
kind
of
composable
solution.
So
a
stack,
that's
interoperable!
That
allows
you
to
select
the
the
transport
or
the
the
the
algorithm
that's
most
most
valuable
for
your
use
case
as
a
developer.
That's
you
know
one
of
the
core
principles
and
kind
of
how
this
interplanetary
stack
combines
together
to
enable
a
whole,
very
diverse
ecosystem
of
parties.
A
Who
are
you
know,
maybe
not
even
using
go
ipfs
itself,
but
are
using
the
the
core
building
blocks
and
components
like
how
falcoin
is
using
many
of
the
core
components
of
ipfs
but
doesn't
wrap
ipvs
itself.
So
graph
sync
was
the
start
of
that,
but
that
definitely
continued
on
as
well.
Falcoin
also
uses
bitswap,
which
is
the
precursor
to
graphsync,
for
for
the
data
transfer
between
many
different
nodes,
and
it
also
uses
the
dht
and
it
instead
of
using
it
for
content
discovery.
A
It
uses
it
for
pure
discovery
so
when,
in
addition
to
the
bootstrap
nodes
and
anyone
you
connect
to
kind
of
directly
by
by
dialing
them
through,
like
gossip
sub,
it
also
uses
the
dht
to
find
peers
in
the
network
and
connect
you
to
people.
You
want
to
sync
them
watching
from,
and
so
ipfs
did
a
num
a
number
of
projects
kind
of
before
understanding
that
these
are
going
to
be
massive
problems,
but
that
flowed
downstream
into
into
five
coin.
A
As
a
user,
and
so
the
there
was
a
chunk
of
work
done
around
bitswap
and
improving
bitswap
performance
performance,
this
was
in
collaboration
with
edgar
from
netflix
working
on
the
kind
of
package
area
package
manager,
space
where
you
want
to
very
quickly
be
be
syncing
graphs
of
data
to
each
other
nodes.
Who
are
you
know
many
nodes
who
are
all
looking
to
to
synchronize
content
quickly.
A
We
made
a
ton
of
improvements
to
bitswap,
especially
around
understanding
whether
you
do
or
don't
already
have
a
message
and
then
not
getting
duplicate
messages
from
many
different
different
nodes
in
the
network,
which
is
super
valuable
for
falcon
as
well,
where
you
don't
want
to
be
sending
duplicate
data
all
across
the
network.
Similarly,
for
the
dht,
we
made
some
massive
performance
improvements
that
landed
in
ipfs
0.5
back
in
it
was
may,
which
just
you
know
was
a
was
a
huge
update
to
the
dht
added
some.
A
You
know
significant
improvements
to
how
how
nodes
advertise
themselves
and
node
hole
punching,
not
hole
punching
and
a
ton
of
the
just
the
core
selection
logic
of
how
you
query
down
down
the
dht,
and
so
that's
also
very
useful
for
allcoin
to
make
sure
that
you
know
when
we
have
many
myers
in
china
and
around
the
world
that
they
can
connect
to
each
other.
They
can
find
each
other
and
they
can
exchange
peer
records
so
that
everyone's
kind
of
resiliently
connected
another.
Even
more.
A
Recent
chunk
of
work
happened
also
in
lapeda
p,
on
gossip
sub,
v
1.1,
and
so
this
was
a
huge
chunk
of
work
that
the
the
lupita
b
team
took
on
specifically
looking
at
use
cases
around
ethereum,
2
and
filecoin,
both
of
whom
were
really
making
the
p2p
the
core
networking
library
of
blockchains,
but
there's
a
significant
financial
incentive.
A
If
you
can
disrupt
or
manipulate
them,
and
so
it
was
a
very
thorough
analysis
using
testground,
which
is
the
the
peer-to-peer
simulation
and
and
analysis
tools
that
we
built
up
in
order
to
improve
the
ipf
sdhd
and
be
able
to
simulate
and
measure
that
effectively.
We
then
use
those
on
gossip
sub
in
order
to
look
at
the
and
model
and
simulate
the
different
attack
dynamics
around
these.
A
These
various
kind
of
abuse,
vectors
that
we
were
researching
and
looking
at
around
blockchains
and
how
lib
p2p
could
mitigate
those,
and
there
was
a
ton
of
work
done
there.
David
has
a
great
talk
on
it.
I
think
giannis
presented
on
the
the
paper
that
was
written
on
gossip
sub
v1.1
yesterday,
and
so
that
that
kind
of
understanding
and
and
seeing
those
downstream
use
cases
also
flowed
back
into
improving
gossip
sub
and
really,
you
know
helping
it
come
into
its
own.
A
It's
been
kind
of
an
optional
or
experimental,
with
an
ipvs
for
a
little
while,
but
this
really
helped
us
like
bring
it
more
fully
into
a
a
kind
of
full
full
use
case
and
and
with
major
downstream
applications
depending
on
it,
which
you
know
really
helped
improve
it
and
solidify
it.
So
that
was
huge,
I
guess
to
mention
test
ground
as
well.
A
That
was
another
major
area
that
the
ipves
and
lip2p
teams
who
came
together,
recognized
that
you
know
neither
of
these
protocols
and
projects
could
improve
effectively
without
better
testing
and
modeling
tools,
without
being
able
to
see
the
impact
of
a
change
and
an
optimization
ahead
of
time
in
a
simulated
network
and
simulated
conditions
before
rolling
them
out.
A
To
you
know
millions
of
weekly
active
users,
whose
you
know
applications
and
tools
depend
on
the
effective
functioning
of
the
ipfs
network
or
lip
pdp
protocols,
and
so
identifying
the
that
need
for
test
ground
and
going
and
building
it
into
its
own
kind
of
standalone
project
and
tool.
Then
also
was
massively
value
for
valuable
for
falcoin
kind
of
in
a
in
a
couple
of
different
ways.
We
were
able
to
bring
that
that
technology
to
bear
on
the
falcon
network
itself.
A
You
know
spinning
up
many
different
lotus
nodes,
running
them
in
a
coordinated
network
and
being
able
to
simulate
all
sorts
of
message,
passing
dynamics
and
edge
cases
that
you
know
are
very
difficult
to
simulate
in
a
in
a
test
net
or
a
definite,
and
we
speed
up
time
so
that
you
can
actually
much
much
more
quickly
like
nci
see
what
whether
or
not
your
changes
have
caused
new
issues
created.
A
Regressions
created,
you
know,
slowed
down
performance
instead
of
speeding
it
up
and
that's
exactly
the
the
sort
of
toolbox
that
we
need
as
protocol
developers
in
order
to
be
constantly
improving
and
growing
the
protocol,
because
if
you
don't
have
a
fast
feedback
loop,
then
your
your
cycle,
time
of
making
improvements
out
into
the
wild
and
your
error
rate
goes
way
up.
A
And
so
tools
like
that
that
were
developed
initially
to
help
us
prove
improve
the
ipvs
dht
flowed
downstream
to
gossip
sub
and
then
to
then
to
falcon
as
well
and
now,
we've
actually
seen
kind
of
some
of
those
same
practices
and
tools
flow
into
the
test,
vector
cases
developed
kind
of
again
by
the
the
test.
Ground
team.
That's
helping
with
cross
implementation,
compatibility
and
conformance
testing,
which
I
think
we'll
talked
about
yesterday.
A
So
that's
a
number
of
the
many
ways
in
which
ipfs,
libya,
b
and
ipld,
of
course,
contributed
to
building
falcon.
I
guess
one
more
I'd
highlight
that
maybe
almost
falcon
contributed
back
to
ipfs's
as
well
was
the
car
file
format.
This
was
something
that
had
had
been
around
and
been
in
design
for
a
really
long
time,
but
recognizing
that
you
know
logical
chunks,
large
logical
chunks
of
merkle
data.
We
want
to
ship
it
around
and
really
quickly
transfer
it
between
between
nodes
and
realizing.
A
How
effective
that
could
be
as
a
means
of
of
like
plugging
and
playing
data
in
ipfs
network,
the
ipld
team
went
and
built
up
the
specification
and
and
kind
of
created
the
the
low-level
tooling
around
it
added
it
into
falcoin
and
into
ipfs.
You
can
now
import
and
export
car
files
of
data
and
it's
magical
it's.
It
adds
a
a
really
improved
capability,
around
kind
of
just
logically
working
with
with
merkle
data
and
that
hadn't
existed
before.
A
But
there
was
like
a
need
and
we
hadn't
quite
solved
the
problem,
and
so
I
guess
that's
an
opportunity
as
well,
where
we
were
involved
in
that
creation,
but
hadn't
quite
crossed
that
gap
until
we
had
a
really
clear
use
case.
Like
the
you
know,
the
neces
necessity
of
car
file
formats
for
shipping
around
data
in
the
falcon
blockchain
case
cool,
so
to
switch
gears
a
little
bit.
The
ipv's
ecosystem
has
been
growing
exponentially
over
the
past
year.
A
A
A
lot
of
these
kind
of
lower
level
tools
empower
the
next
layer
of
developers,
which
kind
of
came
in
the
the
march
time
frame
when
sleek
and
unstoppable
domains
and
ens
and
a
number
of
other
groups
like
really
came
into
their
own.
There
was
just
this
amazing
flourishing
of
d
web
content,
where
many
kind
of
building
on
top
of
those
tools
that
had
you
know
kind
of
come
into
existence
around
this
time,
a
little
bit
before
this
time.
A
Last
year,
six
months
later,
there
was
the
next
layer,
the
next
level
of
tooling,
that
could
build
on
those
foundations
and
then
deliver
something
even
closer
to
the
user,
even
more
seamless,
even
more
directly
connected
into
the
the
expectations
or
defaults
as
a
developer.
A
You'd
be
coming
from
web
2.,
so
like
literally
a
netlife
on
ipfs
building
on
top
of
the
foundations
of
textile
and
and
others,
and
so
that's
the
that's
the
same
thing
that
ipv6
ecosystem
growth
is
doing
with
filecoin
as
well,
one
bringing
over
all
of
these
amazing
like
developer,
tooling
ecosystem
partners,
so
textile
fleek,
pinata
infuria.
All
these
groups
are
also
now
excited
and
investing
in
the
filecoin
ecosystem.
A
So
textile
powergate
is,
like
you
know
the
go-to
solution
for
building
dapps,
that
back
up
data
to
file
coin
and
do
file
point
retrieval
requests
and,
and
things
like
that,
fleek
has
built
the
space
demon,
which
kind
of
similarly
is
just
really
leap:
frogging,
the
the
sort
of
developer
experience
and
so
having
these
groups
already
builds
up
onto
ipfs
and
laptop
already.
A
Building
tooling,
that's
like
really
accessible
to
large
developer
groups
is
kind
of
a
superpower
that
kind
of
is
coming
is
flowing
into
the
falcon
network,
and
we've
already
seen
seen
the
way
that
that's
impacting
things
with
like
falcon
ignite.
How
quickly
kind
of
hundreds
of
projects
and
teams
were
able
to
spin
up
on
using
this
tooling.
A
I
think
the
the
number
of
projects
on
ipfs
alone
through
file
point
ignite,
was
like
over
a
hundred
projects
in
the
first
month,
and
so
that's
you
know
you
can
imagine
if
that's
what
you
get
within
a
month
six
months
from
now
think
of
all
of
the
projects
that
will
have
evolved
from
there
into
new
use
cases
into
the
next
level
of
infrastructure,
which
can
then
support
a
whole
new
flowering
of
projects,
and
so
I
think,
that's
a
very
close
way
in
which
you
know
the
ips
ecosystem
and
the
falcon
ecosystem
are
like
interlinked.
A
They
are,
they
are
in
so
many
ways
the
same
community
and
ecosystem
or
at
least
very
highly
overlapping
and
so
and
they
incentivize
and
grow
each
other.
So
adding
in
the
incentive
layer
of
file
coin
to
people
who
are
already
operating
in
the
ipf
space
opens
up
new
business
models.
It
opens
up
new
features
and
new
efficiencies.
You
know
talking
to
the
pinata
team.
They've
been
you
know,
building
an
ipvs
pinning
solution
for
a
long
time
working
on
it,
but
one
of
the
challenges
was
their
aws.
A
Costs
were
just
really
really
high,
they're,
really
struggling
to
offer
a
service
at
a
price
point
that
helped
them
survive,
help
them
be
a
good,
solid,
sustainable
participant
in
the
ecosystem,
but
that
also
supported
all
of
the
downstream
users,
and
it's
really
cool
to
see
falcoin
even
on
day,
one
being
able
to
like
massively
reduce
that
that
cost
pattern
and-
and
you
know,
give
them
a
new,
a
new
feature
to
offer
to
their
users,
while
meeting
some
of
their
users
biggest
biggest
needs
and
complaints,
so
very
much
win-win
and
bringing
more
people
to
these
ecosystems
together
to
solve
solve
problems.
A
Thinking
a
little
bit
about
kind
of
what
these,
what
these
technologies
unlock
for
web
three,
the
definite
the
criticism
for
for
web3
is,
you
know
hard
to
use
under
focus
on
things
like
developer
ux
and
on
reaching
populations
that
are
not
already
in
the
you
know,
the
early
adopter
space
that
caught
on
caught
the
fire
of
of
web
3
and
blockchain
in
the
early
days
and
actually
reaching
out
to
more
communities-
and
I
think,
ibfs
and
falcone-
have
some
really
good
fodder
and
an
opportunity
to
lead
the
charge
in
that
space.
A
You
know,
of
course,
we're
always
learning
so
much
from
the
way
that
the
ethereum
ecosystem
has
reached
out
to
like
an
immense
number
of
communities
and
really
engaged
everyone's.
You
know
creativity
and
passion
and
formed
like
a
really
phenomenal
ecosystem,
but
I
think
there's
a
lot
of
opportunities
for
falcon
and
ipfs
in
kind
of
the
more
enterprise
space
blockchain
is
is
something
that
is
kind
of
a
buzzword
and
isn't
always
like
clear
how
it
adds
value
or
where
it
adds
value.
A
Whereas
cloud
storage
is
something
that's
like
near
and
dear
to
every
large,
enterprise's
heart
and
the
performance
gains
around
being
able
to
kind
of
ship
data
and
disseminate
data
in
a
cloud-native
way
is
also
like
cutting
edge
from
from
what
most
most
large
businesses
are
thinking
about.
So
I
think
the
domain
area
of
filecoin
and
ipfs
has
the
potential
to
open
a
lot
of
doors
for
for
projects
that
hadn't
previously
considered
that
blockchain
could
be
for
them
and
saying
that's.
A
You
know
a
really
massive
opportunity,
especially
coupled
with
you
know,
the
the
incentive
structure
around
truly
driving
down
storage
costs,
especially
for
very
large
amounts
of
data.
Where
you
know,
petabytes
of
data
gets
prohibitive
on
any
cloud
service
and
having
the
incentive
structure
of
falcoin
really
pull.
Not
only
any
data
but
especially
like
the
the
follow
coin
plus
program
data
could
make
it.
A
You
know
almost
ryan
zero
was
saying
this
morning:
negative
negative
interest
rate
style,
a
little
pay
to
store
your
data,
sorts
of
use,
cases
for
humanity's
most
important
information.
That's
pretty
exciting,
pretty
impressive
and
unlocks
some
some
businesses
who,
I
think,
wouldn't
have
considered
the
web3
space
in
you
know
from
a
you
know:
that's
that's
new!
A
That's
cutting
edge,
that's
something
that
maybe
is
untried
untested,
but
you're
much
more
willing
to
to
try
something
if
you,
if
you
think,
there's
an
opportunity
for
a
really
large
discount
on
a
cost
you're
already
paying,
and
so
I
think
that
that
has
an
opportunity
to
really
speak
to
the
business
aspect
of
decision
making
from
many
many
organizations
that
are
excited
to
continue
investing
and
finally
thinking
a
little
bit
about
the
future
along
those
lines.
A
One
of
those
areas
when
we
think
about
groups
bringing
on
big
data
to
ipfs
and
filecoin
super
exciting,
to
think
about
the
sorts
of
use
cases
that
are
going
to
be
unlocked
and
the
sort
of
businesses
that
never
could
have
existed
before,
and
so
I
think,
there's
some
really
interesting
opportunities
there
around
data
that
supports
itself
that
grows
itself
and
that
kind
of
expands
over
time.
A
A
Like
it's
gonna
blow
our
minds,
there's
gonna
be
new
things
that
we've
never
even
thought
about
that
six
months,
twelve
months
from
now
will
be
coming
online
in
the
network
and
will
take
people
by
storm
the
way
that
you
know,
dot,
eth
addresses
and
using
fleek
to
get
up
a
web3
site
that,
just
like
was
a
fire
that
swept
the
the
web
through
community
and
brought
others
into
the
web3
community.
A
A
I
think
that
the
combination
of
pressures
within
these
three
is
going
to
be
really
exciting
for
development
and
just
like
core
technical
velocity
with
ipfs,
like
we're
going
to
see
through
file
point
through
this
incentive
structure,
larger
and
larger
amounts
of
data
coming
onto
the
ipfs
network.
You
know
to
to
upload
and
to
share
your
data
on
filecoin.
Most
people
start
by
adding
it
to
ipfs.
A
It's
the
easiest
thing
to
do,
and
so
I
think,
we'll
we'll
see
a
lot
of
improvements
there
required
around
how
the
dht
supports
large-scale
publishing,
maybe
alternate
content,
routing
models,
kind
of
borrowing,
some
of
the
incentive
understanding
from
filecoin
to
think
about
the
the
retrieval
market
and
how
that
will
intermesh
with
ipfs
and-
and
I
think
you
know
when
you
think
about
how
the
dht
works
right
now
for
publishing
content.
That's
an
area
in
particular
that
I
think
will
be
very
focused
on
in
the
first
couple
months
of
the
year.
A
Is
you
know
when
you
have
internet
archives,
scale
data
sets
that
are
being
brought
into
the
the
falcon
and
ipvs
ecosystem?
How
do
you
publish
that
content
in
a
way
that
scales?
How
do
you
route
around
that
content?
You
know
what
what
efficiencies
can
we
get
right
now?
You
know
every
time
we
want
to
add
a
data
set,
we're
republishing
every
block
of
content
every
24
hours
and
that's
just
not
going
to
be
scalable
when
you
have
pedophytes
and
petabytes
of
data.
So
there's
going
to
be
some
really
interesting
innovations
there
around.
A
You
know
changes
into
how
we
publish
content
and
how
that
affects
the
structure
of
the
network
and
how
you
find
data
we're
also
talking
earlier
about
indexes
and
and
other
new
niches
forming
in
the
ecosystem,
for
others
to
help
route,
others
to
content
such
that
you're
you're
not
doing
a
random,
walk
and
query,
but
that
you're
able
to
go
to
to
nodes
and
parties
who
spend
time
forming
a
niche
as
a
router
of
information
and
that'll
be
super
interesting.
A
We
also,
of
course,
have
continued
improvements
to
like
last
mile
connections
like
again
lots
of
people
use
ipvas
and
lip
pdp
in
all
sorts
of
internet
environments,
and
so
thinking
about
not
traversal
as
it
affects
your
home,
wi-fi
and
whatever
node
you
might
be
running
inside
your
house
or
things
like
the
great
firewall.
A
How
do
we
make
sure
that
nodes
can
stay
connected
in
all
sorts
of
dynamic
situations
we
spent?
I
kid
you
not
like
a
week
and
a
half
during
the
beginning
of
space
race,
working
with
all
of
the
the
nodes
in
the
falcon
network
to
make
sure
that
folks
could
get
connected,
stay
connected,
make
deals
with
each
other,
send
data
back
and
forth,
and
it
you
know,
was
a
difficult
process.
So
how
do
we
make
the
defaults?
A
There
work
for
the
99.99
of
cases
so
that
everyone
can
be
really
resiliently
connected
to
each
other
and
exchange
data
and
participate
on
the
la
p2p
side.
I
think
we
should
be
super
excited
for
some
of
the
kind
of
classical
components
of
lip22p
coming
with
new
opportunities
for
kind
of
incentives
or
payment
channels
or
ways
to
kind
of
propagate
content,
with
an
eye
towards
marketplaces
and
incentive
systems.
A
So
I'm
super
interested
this
may
or
may
not
be
a
next
year
might
be
a
year
after
next
thing,
but
things
like
incentivize
pub,
sub,
incentivize
dhts
doing
things
like
offline
republishing.
So
you
know
I
is
a
node.
You
know
if
we're
thinking
back
to
that
ipfs
case,
and
I
want
to
publish
all
of
the
internet
archive.
How
can
I
incentivize
other
people
to
help
republish
subtrees
of
the
content
that
I
care
about
and
shard
it
across
many
nodes?
A
You
know:
can
I
have
more
more
flexible
structures
here?
Can
I
you
know,
pay
and
select
for
people
on
the
network
to
continue
propagating
content,
even
after
I,
as
the
as
the
host
or
router
go
offline,
and
then
finally,
filecoin,
like
you,
know,
we're
still
we're
barely
a
week
into
the
falcon
main
net.
So
it's
really
crazy
to
think
about
where
we'll
be
even
just
a
week
from
now
again
so
much
progress
is
happening
here,
falcons
already
innovated
massively.
A
You
know
we,
I
believe,
we're
the
first
implementation
of
eip
1559,
which
had
a
new
proposed
gas
model
for
ethereum,
which
we
innovated
on
iterated
on
a
little
bit
implemented
and
and
are
kind
of
evaluating
and
analyzing
within
falcoin's
network,
and
so
I
think,
we'll
continue
to
see
great
ideas
from
other
networks
as
a
thing
that
we
can
learn
from
like
be
agile,
to
implement
and
utilize
quickly
and
then
have
a
feedback.
Loop
of
you
know
surfacing
back
to
other
groups
how
they've
worked
and
vice
versa.
A
You
know
the
the
solutions
that
we
come
up
with
to
help
scale
and
and
improve.
Fallpoint
may
also
be
things
that
we
can
then
share
and
broadcast
to
other
blockchains,
of
course,
like
every
like
every
blockchain
will
be
making
lots
of
improvements
on
those
on
those
angles
around
scalability
and
throughput.
I
think
there's
a
lot
of
low
hanging
fruit
here
that
we
can
continue
working
on
we're
working
on
things
like
chain
check,
pointing
we
already
have.
A
You
know
a
very
sizable
chain
because
we
have
you
know
so
much
data
being
onboarded
to
the
network,
and
so
I'm
making
sure
that
we
have
a
pretty
flexible
checkpointing
model
so
that
we
don't
have
to
sync
nodes
from
scratch.
A
Every
time
folks
want
to
join
the
network
is
definitely
an
important
area
and
then
also
thinking
about
things
like
light
nodes,
browser
nodes
and
groups
who
want
to
be
able
to
make
storage
and
retrieval
deals
from
the
browser
and
and
interact
with
falcon
without
having
to
sync
the
entire
chain.
That's
also
a
work
in
progress
that
I
think
yeah
I
mean
we
may
even
see
it
before
2021,
but
that
will
unlock
a
lot
of
the
retrieval
use
cases
and
make
them.
You
know,
there's
already
a
ton
of
progress.
A
That's
happened
on
the
retrieval
side,
but
it
will
make
it
much
more
lightweight
for
many
many
people
to
participate
from
their
phone
from
you
know
their
iot
device
from
wherever
they
are
without
having
to
to
run
like
a
full
heavy
lotus
node,
and
so
I
think,
that'll
be
a
really
interesting
area
and
then,
of
course,
you
know
the
the
other
falcon
implementations.
We
have
four
implementations
of
file
coin
two
and
go
one
in
c,
plus
plus
and
one
in
rust
and
they're
all
getting
very
close
to
full
interoperability.
A
So
I
think
we'll
be
starting
up
a
defnet
in
just
you
know,
probably
the
next
couple
of
weeks
with
all
four
interoperating
implementations
on
them
and
then
bringing
them
onto
the
main
net,
and
then
those
implementations
also
specializing
and
focusing
on
different
sub
areas,
whether
that's
kind
of
performance
around
chain-
sync,
whether
that's
you
know,
mining
pools
or
whether
it's
kind
of
usability
and
an
interconnectedness
with
other
ecosystems
or
working
on
various
different
device
types
or
things
like
that.
A
So
I
think,
there's
a
lot
of
a
lot
of
things
that
we'll
see
there
as
each
of
those
implementations
kind
of
innovates
forward
and
brings
back
new
suggestions,
new
fips
to
the
falcon
community,
for
how
we
can
keep
improving
so
super
excited
about
that.
A
Well,
I
am
officially
a
time
should
wrap
people
over
to
juan's
keynote
talk
to
wrap
up
the
day
if
you
do
have
other
other
questions
after
that,
I
believe
we're
going
to
be
hosting
an
ama
this
evening
and
maybe
like
two
hours
from
now,
and
thank
you
all
so
much.
I
don't
see
any
questions
so
I'll,
just
wrap
up
and
hopefully
see
you
all
on
the
internets
have
a
wonderful
rest
of
filecoin
liftoff.