►
From YouTube: Developer Community Call #9
Description
00:00 Intro
00:33 Updates
01:21 About Fluence offline hackathon
06:02 Bounties for ETHDenver 2022
07:23 Development updates
08:08 Projects on Fluence are heterogeneous
09:09 Fluence has many techs. Harmonized ("Aquamarined")
13:15 What has changed over the last 1,5 month
13:24 New libp2p. Node updates. Metrics
14:22 Trust-graph service. Two purposes. Sign service explanation
17:33 Updated libp2p on the Java Scrip side. The new version of fluence-js and aqua
18:21 Changes in aquavm. New node and fluence-js version. Aqua follows all the changes on aquavm.
20:06 Aqua deeply refactored. Two purposes of changes
22:40 fldist is dead. Aqua dist
25:03 What's going to change
25:20 Deliver Marine-Web (milestone 1)
26:12. Registry. Difference between Aqua DHT and Registry
27:42 Working hard with aquadist.
28:15 Language changes. Simplification of the development process and making sintax more lightweight.
29:06 Q&A.Discussion
44:43 Event list on Denver: online and offline
Useful Links
https://fluence.network/
https://twitter.com/fluence_project
https://doc.fluence.dev/docs/
https://doc.fluence.dev/aqua-book/
https://github.com/fluencelabs/aqua
https://github.com/fluencelabs/aquavm
https://fluencenetwork.notion.site/fluencenetwork/Fluence-Developer-Community-Calls-2733beddb3ff4e4986c4fcc830fc796d
https://github.com/fluencelabs/marine-rs-sdk
https://t.me/fluencedev
https://fluence.chat
A
A
We
have
just
updates
q
a
and
euros
sharing
his
work,
so
it's
very
unofficial,
non
formal
call.
So
yeah
should
we
should
we
start
from
someone?
Maybe
eugene.
Do
you
want
to
share
the
latest
about
the
hackathon?
We
we've
just
done
the
in-person
hackathon
in
moscow
that
last
weekend,
do
you
want
to
share
maybe
some
insights
or
how
did
it
go
or
what
do
you
think
of
the
whole
thing.
B
Yeah
yeah
sure
we
actually,
it
was
a
lot
of
fun
because,
like
it
was
the
first
fluency
in
person,
offline
hackathon
in
maybe
two
years,
something
we
did
only
online
last
year,
so
brought
some
additional
insights
so
which
is
which
is
nice,
and
it
was
kind
of
real,
quick
stress
test
for
the
fluency
infrastructure
and
latest
updates
to
aqua,
because
we
just
really
like
hackers,
found
few
bugs
and
we
were
trying
to
fix
them
right
away
during
the
weekend.
So
it
was
really
really
dense
and
hard.
B
I
guess
what's
what's
most
interesting
is
the
projects
that
people
are
trying
to
build,
we'll
try
to
have
some
of
them
presenting
on
on
these
calls,
but
basically
one
example
is
the
project
that
implemented
the
off
chain
governance
protocol
on
fluence,
which
was
inspired
by
snapshot
design.
But
you
know
the
problem
of
snapshot.
B
Is
it's
centralized
and
all
the
volts
and
proposals
go
through
a
single
server,
so
they
created
the
basically
the
fluent
service
that
and
the
network
of
fluent
services
that
had
consensus
about
the
voting
data
and
the
on-chain
verification
about
the
data.
We
are
more
merkle
trees
and
merkle
miracle
proofs,
and
also
they
have.
They
have
really
interesting
and
well
thought
mechanism
of
dispute
resolutions
with
users.
B
We
are
merkle
trees,
so
this
is.
This
is
pretty
cool.
The
other
example
and
a
few
teams
actually
wanted
to
do
it
is,
was
cross
chain
nfc
transfers
when
you
basically
use
fluence
as
also
the
off-chain
worker,
that
looks
at
one
chain
and
sees
if
the
nft
was
burned
on
this
chain
and
then
means
this
the
same
nft
on
another
chain.
So
it's
basically
like
a
transfers
and
then
so,
like
guys
created
just
just
one
worker.
B
B
So
it's
pretty
cool,
and
also
there
was
you
know,
ideas
in
order
to
build
integrations,
fluence
integration
to
something
else
like
to
the
graph,
for
example,
which
sort
of
would
be
would
be
really
cool.
For
the
I
mean
for
the
future,
where
we
want
to
connect
many
blockchains
to
fluence
and
read
data
from
from
blockchains
easier,
we
maybe
even
have
we
will
maybe
publish
the
bounty
specifically
for
integration
of
the
graph,
and
I
had
this
funny
funny
background
that
appeared
for
a
second.
B
A
Great,
thank
you,
jenny.
If
anyone
has
questions,
please
go
ahead,
just
to
say
that
those
bounties
will
be
divided,
50
50.,
so
seven
and
a
half
thousand
are
for
online
bounties
and
tomorrow
we
should
be
presenting
it
during
the
evening
evening,
time
europe
and
morning
time
us,
and
we
also
will
publish
it
in
the
discord.
So
everybody
knows
cool.
So,
let's
move
on
to
the
next
tech
updates
dmitri.
Do
you
want
to
share.
C
C
C
With
the
help
of
a
marine,
so
you
can
think
about
protocols
which
you
deploy
and
influence
the
spiritual
protocols
as
homogeneous
as
they
are
just
expressed
with
aqua.
But
actually
applications
are
polyglot
and
for
every
part
of
the
application
of
fluence
has
some
technology
which
is
backed
by
another
technology
and
maybe
by
another.
So
that
forms
a
fluency
stack,
which
is
a
harmonized
or
argument
aquamarine,
so
that
everything
works
together
and
helps
developer.
C
D
C
Itself
is
affected
by
aqua,
vm
and
fluence.js
is
affected
by
aqua
vm,
the
virtual
machine
that
handles
aqua
and
in
the
near
future.
All
this
setup
will
be
affected
by
marine
web
when
we
will
run
web
assembly
modules
in
the
web
browser
and
in
the
javascript
environment,
with
the
help
of
marine
and
web
assembly
interface
types.
C
C
C
C
So
that's
that's
one
line
for
sure
rust.
Node
is
based
on
lipit2p
on
rastly
p2p
flux.js,
based
on
javascript,
like
p2p,
ras
node,
based
on
rustly
different
projects
serving
the
same
purpose,
and
they
they
must
connect
to
each
other.
When
you
interact
with
the
node
from
fluence.js,
actually
you're
using
fluence
des
le
p2p
and
then
rustly
p2p,
and
then
you
go
to
the
node
and
also
with
the
node.
We
have
a
set
of
default
web
assembly
services,
which
are
the
part
of
the
fluence
protocol.
C
C
C
C
Like
to
to
unlock
more
complex
scenarios
of
trust
graph
usage,
they
sign
service
was
implemented
on
the
fluence
jazz
site,
so
science
service
is
a
an
abstraction
over
a
key
pair
in
influence.js
that
lets
you
sign
different
payloads
from
aqua.
So
you
you
can
ask
the
javascript
peer
to
sign
your
payload,
but
it
will
not
work
for
everything.
It's
protected,
it's
protected
with
a
tetraplates,
so
it
it
plays
well
with
all
the
all
the
setup
of
the
fluence
stack.
C
C
C
C
C
C
C
So
currently
you
need
to
have
a
temporary
temporary
variable
to
hold
the
value
to
pass
it
to
some
other
function,
and
now
we
have
a
place
where
we
can
just
just
remove
this
unnecessary
variable
from
from
the
code,
yeah
and
finally,
apple
dist
is
dead,
so
we
had
a
tool
that
was
implemented
more
than
a
year
ago,
probably,
and
it
was
implemented
when
we
had
no
aqua
at
all.
It
was
implemented
at
either
the
low
level
language
with
er
and
we
haven't
had
a
fluency
s
as
well.
C
So
it
cannot
support
a
newly
b2p,
so
it
doesn't
work,
no
functions
in
flgs
works
and
actually
we
we
even
forgot
about
it.
We
don't
like
a
full
disk
and
we
were
so
waiting
so
long
for
it
to
to
to
do
it
today
that
that
we
before
the
hackathon,
we
updated
everything.
We
checked
everything
and
we
forgot
to
check
apple
dist,
and
that
was
what
we
fixed
that
evgeny
mentioned
during
the
hackathon
that
there
should
be
some
means
to
deploy
webassembly
modules
to
the
network.
C
It
took
a
few
hours,
and
now
you
have
aqua
dist
collection
of
commands
in
aqua
cli,
so
you
have
aqua
dist,
deploy.
Probably
you
have
aqua,
dist
remove
and
some
other
stuff
that
calls
aqua
under
the
hood
and
it's
a
single
bundle.
So
it's
it's
not
a
distinct
tool
anymore.
It's
part
of
the
aquacompiler
which
is
nice,
but
it
has
some
new
bugs
and
it
doesn't
have
all
the
features
that
we
had
and
for
this,
so
it
changed
very
rapidly.
C
C
So
the
goal
is
to
have
the
same
web
assembly
modules,
running
in
the
browser
and
influence.js
and
in
the
rust
node
to
have
the
same
codebase
and
the
same
behavior
for
all
the
built-in
services,
and
it
will
help
with
stability
testability
and
eventually
it
will
let
us
to
make
a
full
feature.
The
javascript
node
currently
fluence.js
can
only
work
with
a
with
the
connectivity
help
of
the
rust
node,
but
in
the
future
with
the
marine
weapon,
it
will
work
on
its
own
as
well.
So
we'll
have
two
implementations.
C
So
the
main
difference
between
aqua,
dht
and
registry
is
that
awkward
ht
wasn't
secure.
It
wasn't
protected
with
the
signatures,
while
registry
is
secure,
it's
protected
with
signatures
and
that's
what
was
actually
expressed
in
in
the
rfc
and
also
registries
integrated
with
plus
graph.
So
they
form
a
synergy
between
them.
Where
trust
graph
is
the
grant
part
of
a
take
grant
model
and
registry
is
the
take
part
of
the
take
grant
model.
C
C
C
E
C
C
So
how
can
I
advertise
my
web
browser
so
that
you
can
deliver
the
message
to
me
with
a
registry?
You
can
do
that,
so
you
can
register
the
like.
You
can
call
it
topic
or
a
topic
in
terms
of
pub
sub
or
you
can
call
it
the
root
id
and
you
can
register
many
pathways
and
many
peer
ids
and
service
ids
to
this
root
id
to
be
resolved
on
runtime,
so
that
you
can
use
these
parts
of
topology
of
the
network
topology
when
you
call
the
code,
something
like
that.
C
C
F
D
B
C
E
C
Something
like
that,
it's
deployed
to
all
the
fluent
spheres.
It's
a
part
of
the
fluency
protocol
and
it's
replicated
following
the
academia
dht
algorithm.
So
you
have
some
redundancy
like
eight
copies
of
every
record
and
so
on,
and
it's
stored
in
in
the
local
database
so
probably
should
outlive
the
node
restart
police.
Will
it?
C
E
Sorry
in
the
node
restart
it
doesn't
get
stored,
you're
saying
or
does
it
get
stored?
Yes,
it
is
stored
in
the
node
okay.
What
happens
if,
like
a
bunch
of
notes,
go
down?
It
just
gets
replicated
if
they
pop
back
up
and
is
there
like
a
threshold
or
something
like
that
that
you
look
at
that's
like
stating
the
true
fact
of
the
registry.
C
Every
every
key
and
every
value
is
replicated
by
a
factor
of
20.
Currently,
so
it's
a
lot
and
if
part
of
the
network
goes
down
eventually,
if
it
goes
back
or
if
any
other
kind
of
a
network
structure
happens,
it
will
be
replicated
in
a
in
the
neighborhood
of
the
hash
of
the
id
of
the
root
id,
and
eventually
you
will
be
able
to
find
it.
C
B
C
E
C
B
You
it
doesn't
work,
no,
we
cannot
hear
you
and
speaking
about
the
registry
data.
Actually,
I
guess
the
most
organic
solution
should
be
that
when
we
have
the
economic
model,
the
ability
to
pay
to
nodes
for
hosting
your
services,
there
should
be
also
commitment
from
nodes
to
keep
the
registry
data
that
related
related
to
your
application.
B
So
it's
kind
of
should
be
for
the
application
developer
to
decide
whether
or
not
it's
it
worth
paying
for
gas,
because,
like
every
time,
every
time
there
is
an
update,
you
need
to
send
a
new
transaction
right
to
unchain
pay
for
gas,
and,
if
you
use
not
really
popular
chain,
then
you
know
you
have
problems
that,
like
it's
always
the
balance
between
the
cost
and
availability
right,
so
ethereum
probably
will
give
you
availability
like
permanent.
It
will
be
permanent,
it
will
be
forever
there,
but
it
costs
a
lot.
B
B
Yeah
yeah,
for
sure
I
mean
like
the
problem
is
polygon,
is
that
they
hold
their
own.
They
have
their
own
issues
with
growing
state
and-
and
you
know
who
knows
where
polygon
will
be
in
a
few
years
right
so,
but
but
in
general,
the
ability
to
to
give
the
application
developer
the
choice
the
right
to
anchor
or
to
deploy
registry
data
to
some
of
the
chains.
That
would
be
nice
yeah.
I
agree
with
you.
C
C
So
if
you
can
find
right
now,
if
you
can
find
this
node
with
a
spot,
it
should
appear
on
the
red
string.
But
if,
if
it's
not
the
case
because
of
the
loss
of
connectivity,
if
the
node
is
dead
or
something
else
happened,
it
should
disappear
from
from
the
right
strain.
Otherwise,
you
will
have
a
lower
level
of
services
of
the
network.