►
Description
Filecoin Meetup Paris 2022
A
Thank
you.
Thank
you.
So,
yes,
I'm
ali
talking
about
the
falcon
virtual,
a
machine
today,
and
actually
we
have
one
of
the
early
builders
here,
thomas
who
is
building
on
the
falcon
virtual
machine,
so
knows
a
lot
about
the
technicals
as
well.
A
So
for
those
of
you
that
I
mean
we've
had
a
lot
of
information
today.
So
currently
we
all
know
farcoin
network
as
a
storage
and
retrieval
centric
network
for
storing
our
valuable
data
over
17
extra
bytes
or
something
crazy
of
data
capacity
and
thousands
of
awesome
storage
providers
across
the
globe,
some
of
whom
also
might
be
here
in
the
room
today.
A
So
we
can
think
of
this
storage
and
retrieval
functionality
as
being
file
coins,
layer,
zero
architecture
and
even
with
just
this,
we've
achieved
a
lot
and
we
have
so
many
things
being
built
off
this
network,
and
we've
heard
some
of
those
here
in
the
room
today,
falcone
green,
the
social
contributions
that
are
happening
at
falcon
foundation
for
the
decentralized
web.
All
of
this
stuff
has
been
made
possible
by
the
file
coin
network.
A
But
really
storage
is
just
the
start.
We
want
to
be
able
to
do
things
with
this
data
in
our
network,
and
this
is
the
goal
of
the
falcon
virtual
machine.
So
fvm
is
aiming
to
bring
smart
contract
program
programmability
to
the
filecoin
storage
network
and
enable
users
to
deploy
their
own
smart
contracts
or
actors,
as
they're
called
in
the
file
coin
ecosystem
on
filecoin,
and
this
really
introduces
the
ability
to
perform
on-chain,
computation
or
computation
over
data
which
we
can
think
of
as
bringing
layer.
A
One
capabilities
to
the
file
coin
stack,
which
is
super
exciting
and,
as
molly
said
earlier,
this
is
not
something
that
any
blockchain
has
attempted
before.
So
once
we
have
our
storage,
the
layer,
zero
and
we
add
on
chain
computation
the
layer,
one
then
you're,
really
enabling
so
many
many
downstream
use
cases
and
applications
to
emerge
both
on
and
off
chain,
which
is
a
layer
two.
A
So
before
I
get
to
some
of
the
use
cases
that
we're
envisioning
here
and
there's
so
many
more
than
we
can
possibly
mention.
I
want
to
quickly
dive
into
the
technicals
that
are
behind
the
fvm
and
how
we're
making
this
happen
so
user-defined
smart
contracts
have
always
been
part
of
the
filecoin
development
roadmap.
The
priority,
though,
has
first
been
to
develop
the
layer
zero.
This
robust,
secure
and
permissionless
decentralized
storage
network.
That
is
falcoin
as
we
know
it.
A
A
So
when
architecturing,
the
fbm,
the
team
were
initially
inspired
by
the
virtual
machine
hypervisor
model
and
its
ability
to
host
multiple
virtual
run
times
over
a
common
base
layer
and
for
those
of
you
that
have
done
anything
in
cloud.
You'd
likely
be
familiar
with
this
model,
so
the
team
was
also
inspired
by
the
erlang
or
otp
actor
model,
which
is
why
we're
calling
them
actors,
and
particularly
the
self-healing
mechanisms,
the
location,
transparency
and
the
concurrency
programming
paradigm
of
the
messaging
model.
A
So
both
of
these
models
have
influenced
the
long-term
vision
on
how
programming
on
blockchains,
we
think
will
look
in
the
future.
So
the
current
implementation
of
fvm
is
based
on
wasm
or
webassembly,
which
is
a
low-level
assembly-like
language
with
a
compact
binary
format.
So
it
runs
with
near
native
performance,
so
many
different
languages
can
be
compiled
down
to
webassembly,
which
is
another
advantage,
and
it's
become
popular
in
the
blockchain
space,
because
the
wasn't
byte
code
doesn't
require
any
further
compilation
or
transpilation
to
be
executable,
and
it's
really
suitable
for
secure
and
content
addressable
code.
A
Therefore,
so
the
fbm
reference
sdk
is
built
on
rust,
and
this
is
because
rust
ships
with
a
really
minimal
runtime,
so
the
size
of
the
package
that
needs
to
be
included
in
the
webassembly
binary
stays
small.
But
we
do
have
early
builders
in
our
program
that
are
building
an
assembly
script
sdk
for
the
fbm
reference
and
also
experimenting
with
a
tiny
go
sdk
and
actually
thomas
and
his
team
are
building
a
high
level
rust
sdk
to
make
it
easier
for
developers
to
use
as
well
so
with
the
help
of
the
falcon
community.
A
As
I
just
mentioned,
we're
working
on
supporting
our
next
evm
solidity
and
compatibility
out
of
the
box-
and
this
is
the
current
milestone
that
we're
working
on.
So
we
are
currently
scoping
and
coding
out
our
evm
compatible
fvm
straight
out
of
the
box,
which
means
you
can
bring
all
your
tooling
from
the
avm
ecosystem,
hard
hat,
open,
zeppelin,
all
your
contracts
onto
the
fbm
chain
and
that
will
be
available
in
q4
of
this
year.
A
So
it's
super
important
for
us
to
meet
developers
in
the
ecosystem
where
they're
at
and
the
evm
community
has
built
up
a
massive,
really
useful,
tooling
and
applications
and
programs
to
help
develop
and
test
with.
So
we
want
to
make
sure
that
we
are
enabling
that
as
well.
If
you're
looking
for
more
details
on
the
implementation,
though
check
out
the
reference
spec
on
the
github
for
more,
I
think
that's
just
there
at
the
bottom.
Yep
awesome.
A
So
what's
all
this
tech?
Enabling,
though-
and
why
are
we
so
excited
about
it?
So,
as
I
mentioned
a
little
earlier,
there's
so
many
opportunities
for
innovation
on
this
layer,
one
programmability
kind
of
that
it's
going
to
get
unveiled.
So,
and
we
know
the
falcon
community
is
super
creative,
so
we
know
there's
not
going
to
be
any
short
shortfall
of
ideas
there.
So
the
above
map
here
illustrates
some
of
the
use
cases
that
the
fvm
can
enable
and
that
we're
personally
really
filtastically-
sorry
not
sorry
excited
about.
A
So
these
are
things
like
data
downs.
The
database
tokenized
data
sets
nfts
getting
minted,
are
exchanged
and
stored
under
one
like
roof,
so
to
speak,
are
replication
workers
so
imagine
anyone's
able
to
write
a
new
smart
contract
that
makes
new
deals
to
maintain
a
specific
level
of
replication
of
the
data
that
you're
putting
onto
the
network.
A
That
means
you
don't
have
to
like,
distribute
it
to
multiple
data
systems
that
will
happen
automatically
for
you,
trustless
reputation
systems
and
I'm
gonna
double
click
on
some
of
these.
I
don't
know
if
they're
still
in
the
room,
but
anyway,
storage,
bounties
and
actor
auction
mechanisms,
sorry
futures
and
derivatives
on
storage,
so
opening
up
defy
even
you
know
or
composing,
in
a
d5
like
fashion
conditional
loans,
time
locked,
retrieval,
perpetual
storage
and
if
you
really
think
about
it,
some
of
these
use
cases
are
really
hard,
if
not
impossible.
A
Thanks
for
turning
up
for
to
your
talk
dragon
to
build
on
centralized
and
cloud
storage
systems-
and
this
is
where
farquaad
with
the
fpm
is
really
going
to
thrive
anyway.
Sorry
dragon
is
also
on
the
team
and
is
going
to
help
us
out,
but
I'll
just
get
through
the
use
cases,
and
you
can
you
know
chime
in
where
you
want
to
I'm
just
at
the
use
cases
and
possibilities.
A
B
A
I'm
kidding
he's
amazing
and
doing
a
lot
with
our
ecosystem
and
helping
our
early
builders
and
building.
B
A
Program
as
well
so
anyway,
let's
get
on
to
the
first
use
case
that
we're
excited
about
dataverse
and
datadowels,
so
imagine
kickstarting
the
dataset
economy
by
tokenizing
datasets
right,
so
you
could
capture
and
represent
the
value
of
these
data
sets
to
society
with
tokens.
A
So
imagine
if
those
tokens
are
then
harvested
and
exchanged
between
peers,
you
could
request
computation
services
to
be
from
performed
on
that
data.
Things
like
validation,
transformation,
feature,
detection
and
extraction,
which
leads
into
machine
learning,
which
is
particularly
exciting
for
me
and
analysis
and
like
there's
just
like
as
soon
as
you
can
do
this
like
there's
so
many
things
you
can
do
with
this
data
and
you
own
it
and
you
can
decide
on
the
value
or
the
market
can
decide
on
the
value
as
well.
A
Possibly,
and
these
open
data
sets,
I
think,
as
we
were
mentioning
before,
like
climate
data
sets
being
able
to
machine,
learn
over
that,
and
you
know
come
up
with
new
solutions
for
what's
going
on
in
you
know
some
really
complex
problems
so
anyway,
you
could
deploy
also
self-governing
entities
on
chain
to
steward
and
maintain
and
curate
and
preserve
those
data
sets
as
well,
and
that
will
be
particularly
important
for
these
open
data
sets
and
scientific
data
sets
and
for
data
that's
valuable
to
humanity
in
general,
you
could
even
start
crowdfunding
that,
so
you
can
see
we're
excited
about
dataverse
and
data
dows
here.
B
I
tried
to
understand
it
when
I
first
read
about
it,
but
it's
basically
a
way
how
you
can
automatically
replicate
workers
on
the
on
the
net
and
then
you
can
automatically
set
them
up
for
like
have
easier
use
cases
in
terms
of
file
coin,
how
it
actually
works.
So
you
don't
need
to
involve
your
client
or.
A
B
A
Then
so,
enabling
specification
of
things
like
the
region
or
the
location
or
the
latency,
which
really
brings
up
another
important
use
case
here,
actually
not
that
one,
maybe
smarter,
storage
markets
anyway.
So
another
important
use
case
is
you'll,
be
able
to
like
basically
run
an
actor
across
this
system.
A
To
make
pretend
or
like
you
know,
to
create
storage
deals,
small
storage
deals
and
that
in
that
way
you
can
kind
of
get
the
reputation
of
those
storage
providers
which
then
you
have
a
whole
network
with
reputations,
with
latency
with
the
geography
with
price
things
like
that.
So
then
you
can
decide
where
you
want
to
store
it.
You
can
make
a
user-defined
policy
for
your
data
to
do
that
as
well.
A
That's
part
of
replication
workers,
I
guess
and
smarter
storage
systems
anyway.
We
also-
I
it's
also
going
to
enable
these
l2
commitments.
So
it's
going
to
currently
we
have
some
tooling
in
the
ecosystem,
and
that
includes
like
textile.
A
Some
people
are
making
some
experimental
bridges
to
things
like
polygon
and
to
nia
and
a
few
others,
but
it's
really
hard
to
do
these
so
with
l1
and
with
actors
we're
going
to
be
able
to
provide
these
l2
bridges
as
well.
So
it's
going
to
make
it
a
lot
easier
to
do
that.
Do
you
have
something
to
add,
or
no
no
okay
just
know
great.
A
Take
that
as
a
thumbs
up
so
smarter
storage
markets
as
well.
This
is
another
use
case.
We're
excited
about
it's
adding
richer
functionality
to
storage
markets,
so
auto
renewing
of
deals,
which
is
what
we
really
want:
an
ability
to
renew
deals
without
repeating
those
data
transfers,
because
at
the
moment
this
is
a
real
pain
point
like.
A
Also
self-repairing
deals
are
available
then
so
even
node
goes
down,
and
this
is
you
know,
cloud
is
is
up
to
this
point.
Evernote
goes
down
or
sectors
fail
to
prove
providers
can
fetch
a
copy
of
the
original
deal
from
another
provider,
potentially
with
like
the
mediation
of
that
replication
worker
and
restore
this
data.
There's
also,
you
know,
retrying
of
deals,
so
clients
can
dispatch
their
deal
data
to
a
depot
and
maybe
pay
a
small
fee
to
the
depot
to
ensure
the
satisfactory
delivery
of
that
data.
A
So
a
lot
of
a
lot
of
use
cases
around
how
we
interact
with
data
and
like
making
it
really
more
up
to
you
how
you
how
you
want
to
do
that
and
making
it
really
easy
to
do
as
well
and
another.
One
of
these
smarter
deals
is
like
time
locked
retrieval.
Potentially,
you
don't
want
to
allow
that
data
to
be
retrieved
by
anyone
for
a
certain
amount
of
time,
so
you
can
make
storage
deals
with
data
that
can't
be
retrieved
by
anyone
until
that
time
elapses
or
maybe
a
specific
event
triggers.
A
So
this
could
be
a
really
cool
use
case
as
well
in
certain
situations,
yeah.
B
A
B
A
Another
exciting
use
case
is
decentralized,
verifiable,
computation
so
because
you're,
adding
computation
onto
the
falcon
blockchain,
I'm
ready
for
a
drink.
You
have
both
computation
and
storage
under
the
same
roof,
and
this
is
going
to
be
one
of
the
big
strengths
and
differentiators
of
the
farcoid
network,
so
like
being
able
to
define
computation
to
be
applied
to
data.
Incentivizing
is
execution
all
the
way
through
and
being
able
to
certify
that
the
result
and
output
is
verifiably.
A
Correct
link
all
that
with
like
the
dataverse
and
data
dowse
and
the
tokenization
of
datasets
and
you've
got
like
so
much
computation.
That's
pushed
to
the
edges
and
pushed
to
the
providers
that
have
the
data
and
with
the
incentivization
improveability
circuit
inbuilt
to
verify
results
so
super
exciting.
What's
the
roadmap
looking
like,
when
is
this
going
to
be
out?
When
can
you
build
with
it?
So
our
roadmap
at
the
moment
is
available
to
track
across
both
our
github
and
the
website.
So
that
is
our
website
up
there.
A
You
can
go
and
check
that
out
at
the
moment.
What
was
super
exciting?
Is
we
just
released
a
v16,
the
skier
update.
B
Yeah,
so
it's
just
live,
I
think
from
last
week
and
it's
basically
what
what
molly
told
us
like
introducing
wasm
to
filecoin
and
now
we're
on
our
way
to
m2,
which
is
basically
introducing
programmability,
so
yeah.
So.
B
A
Yeah
so
kind
of,
like
I
mentioned
earlier,
we
want
to
enable
these
evm
compatible
fvm
fevm
as
we're
calling
it
on
the
network
first,
which
is
milesto
2.1
and,
in
parallel,
we'll
also
be
working
on
the
infrastructure
and
tooling
for
the
native
version
of
fvm.
So
we
have
a
lot
of
builders
in
our
ecosystem.
A
Doing
so
many
amazing
things
and
helping
us
with
this,
and
if
you
are
interested
in
being
part
of
that
ecosystem-
or
you
know,
teams
that
are
on
either
deploying
evm
use
cases
to
fvm
or
on
creating
some
of
the
native
tooling
or
creating
businesses
around
some
of
that,
then
there's
heaps
of
ways
to
get
involved.
The
first
way,
if
you're,
just
interested
in
general
or
curious.
The
first
way
is
check
out
our
fifth
discussion.
A
A
Applications
for
our
early
builders
program,
like
I
mentioned
thomas
and
a
few
others-
are
part
of
our
early
builders
program,
building
out
the
native
rails
of
the
native
fem,
but
we're
opening
up
to
people
that
want
to
deploy,
maybe
solidity
contracts
they
already
have,
or
programs
or
applications
they
already
have
onto
the
filecoin
virtual
machine.
Just
opened
them
today
so
check
that
out
on
our
website.
A
The
other
way
that
you
can
keep
in
touch
is
just
follow
the
youtube
and
there's
a
few
demos
on
there
of
what
our
early
builders
are
doing
as
well
so
assembly
script,
sdk
written
by
zondax
is
up
there
as
well
as
there's
a
playground
and
active
playground
at
the
moment.
So
you
can
play
around
with
how
actors
might
work
and
that's
done
by
a
community
member
called
gympic
as
well,
and
that's
very
useful
if
you're
just
trying
to
kind
of
learn
a
little
bit
as
well.