►
From YouTube: vCPU, Parallel Computation, DAPP Hackathon, LiquidX, AI and more with LiquidApps CTO - Tal Muskal
Description
Tal Muskal, CTO at LiquidApps, joins Everything EOS to discuss the upcoming DAPP Network Hackathon and more.
** REGISTER FOR THE DAPP HACKATHON! It's remote, so no travel required, and features up to $50,000 in prizes. Running from November 17 to 24. Find out more and sign up at dappnetwork.hackerearth.com **
1:00 Tal as Lead Hackathon Mentor
4:00 vCPU to Offload Computation
9:00 Immortal Applications / LiquidChess
11:00 End-to-End Decentralization
13:00 LiquidStorage
15:00 LiquidAccounts
18:00 Artificial Intelligence on a Blockchain
25:30 Horizontal Scalability with vCPU
27:00 vGPU Acceleration
29:00 Import Any C++ Library Onto EOSIO
30:00 Training a Deep Learning Model on EOSIO
35:00 Hackathon
37:30 Javascript Compilers to WASM
39:30 LiquidX
42:00 Multi-Chain Vision
45:00 Open Invite to Block.One Employees
49:00 Deferred Transactions and LiquidScheduler
53:00 Zeus IDE
54:30 Hackathon Tips and Advice
55:00 Tal Hacks Nintendo Wii
***
Series Sponsor:
Cypherglass - https://www.cypherglass.com
https://www.eosnameservice.io
For more information on sponsorship opportunities or content partnerships, please contact zack@everythingeos.io
https://www.EverythingEOS.io
https://youtube.com/c/Everything_EOS
http://t.me/Everything_EOS
https://anchor.fm/everythingeos
https://itunes.apple.com/us/podcast/everything-eos/id1434560651?mt=2bi
A
A
A
You
could
have
been
a
judge,
but
I
think
it's
really
cool
that
you're
going
to
basically
be
the
lead
mentor
you're,
going
to
basically
get
away
from
your
terminal
for
a
week
to
help
with
other
people's
terminals
and
just
be
fully
available
to
everyone
participating
in
this
hackathon.
Like
you
said
it,
the
best
part
about
it,
for
me
is
that
it's
glue
it
will
loco
in
Tel.
A
Aviv
I
will
be
in
Tel
Aviv
I'm,
looking
forward
to
that,
but
the
global
virtual
aspect
of
it
is
something
that's
never
been
done
before
with
it
within
the
EOS
io
community,
and
it's
just
a
great
opportunity
for
the
people
who
weren't
able
to
make
it
to
any
of
the
hackathons
last
year.
What
was
like
your
motivation
for
wanting
to
organize
the
hackathon,
the.
B
A
What
has
me
excited
is
like
our
technical
team.
Has
you
your
whole
team
I've
just
worked
so
hard
to
get
these
services
released
and
get
them
to
it
to
a
state,
that's
usable,
and
now
it's
like
they're
ready
for
Showtime.
We
want
to
see
what
they
could
do
because,
like
liquid
I
adapt
Network
liquid
apps,
like
the
tools
were
building
for
the
top
Network,
it's
like
an
extension
of
ESA.
Oh
it
makes
it
more.
A
Powerful,
I
explain
it
to
people,
it's
like
putting
EOS
IO
on
steroids,
and
we
have
all
of
the
these
talented
developers
that,
especially
lately
with
the
resource
constrictions.
Let's
just
jump
into
that
I
know
it's
later
in
my
plan
of
topics.
I
wanted
to
go
over
with
you,
but
let's
just
jump
into
it.
The
V
CPU,
because
I
think
this
is
super
relevant.
The
network
has
basically
been
congested
to
almost
a
standstill
over
the
last
two
days
and
because
of
the
mechanism,
that's
causing
this
network
spam
and
and
I
call
it
spam.
B
Very,
very
scarce
resource.
It's
the
single
thread
of
computation.
It's
live
illan
of
one
CPU,
that's
everyone's
sharing.
We
should
expect
it
to
be
fully
utilized.
It's
that's
supposed
to
be
the
steady
state.
If
this
thing
is
successful,
which
UNIX
we
should
expect
no
idle
time
and
these
cpu
research
should
be
fully
utilized.
B
Having
said
that,
we
can
we
can
now
think
of.
We
cannot
think
of
utilizing
things
like
the
CPU
to
offload
most
of
the
intensive
computation
to
to
other
means
and
other
more
horizontally.
Scalable
means
as
I
see
it
the
CPU
of
the
main
and
should
mostly
be
used
to
for
provisioning
purposes
and
for
for
thinking
other
types
of
resources,
because
it
requires
a
complete
consensus
and
and
replication
of
the
entire
state
of
replication
of
everything.
A
A
A
We
are
now
in
a
state
on
the
main
net,
where
there's
not
going
to
be
any
more
free
lunch,
but
that
doesn't
mean
that
applications
decentralized
applications
on
ESI,
o
or
o
es
in
particular
in
this
case,
need
to
need
to
be
limited
in
what
you
can
and
can't
do
on
chain,
and
you
want
to
kind
of
get
into
since
V.
Cpu
is
ready
to
be
used
and
we
have
this
hackathon
coming
up.
A
B
B
Even
one
DSP,
without
really
interesting,
without
really
trusting
it
another
example
can
be
building
machine
learning
model.
You
can
actually
evaluate
this
score
of
the
machine
learning
model
and
each
it's
much
much
more
lightweight
than
building
in
the
building
in
the
model.
We
think
the
model
is
much
much
heavier
than
actually
validating
were
evaluating
the
scores
or
model,
so
I
think
those
kind
of
a
symmetrical
computation
opportunities
are
or
turn
of
a
lot
of
data
intensive
applications
and
those
things
are.
B
We
didn't
really
begin
even
to
imagine
them
up
until
now
that
intensive
on
blockchain,
something
that
less
than
a
year
ago
and
when
I
discussed
it
like
people,
people
kind
of
raised
their
eyebrows
when
I
wanna
interact
like
thing
when
I
discussed
things
like
unchain,
artificial
intelligence
and
the
economics
around
it,
so
people
were
quite
skeptical.
That's
one
of
the
reasons
that
chess
was
important,
the
showcase
for
me
personally
so.
A
Let's,
let's
just
get
into
the
the
the
first
use
case,
that
we
we
have
a
blog
article,
where
the
link
will
be
in
the
description
below
but
Tao
Tao
built
what
is
called
liquid
chess
and
it
is
the
first
application
of
V
CPU
that
that
we
released
on
our
github
in
our
blog.
So
why
don't
you
let
everyone
know
what
liquid
chess
is
and
how
it
has
something
which
is
the
AI
that
nothing
has
ever
had
before
on
a
blockchain.
So.
A
B
Up
until
now,
the
the
selling
points
of
technologies
and
cryptocurrencies
are
the
immutability
and
decentralization
and
those
two
words.
Those
two
terms
say
say
nothing
to
most
people.
It's
like
completely
gibberish
to
most
people
like
what
does
that
even
mean
what'swhat's,
the
centralization
mean,
isn't
the
Internet
decentralized
in
the
first
place
and
like
it's
very
far
from
people
and
and
I,
think
mortal
resonates
better
in
the
sense
that
that
that
something
is
there
forever
and
nobody
can
touch
it
and
that's
kind
of
other
another
way
to
describe
this
perfectly
central
I
have
any
beautiful
system.
B
A
Front
end:
that's
decentralized
like
when
liquid
chest
got
released.
A
lot
of
people
heard
the
term
immortal
application
for
the
first
time
that
people
are
coming
into
the
telegram
channel
and
they're.
Like
tell
me
more
about
these
immortal
applications.
What
is
like,
what
are
more
details
that
this
has
never
been
done
before,
like
your
smart
contract,
could
always
have
been
on
a
blockchain,
but
you
still
need
a
user
interface
to
work
with
to
interact
with
that
smart
contract
and
that
could
always
get
taken
down.
You
could
have
man-in-the-middle
attacks
and
DNS
attacks.
B
B
A
B
A
B
B
You
can
say
that
theorem
already
has
all
the
parts
you
need
to
provide
something
like
five
point
or
decentralized
storage
and
with
liquid
storage,
we
we
make
it
more
accessible
and
easier
to
use,
is
as
a
storage
solution
and
not
something
that
the
states
that
the
contract
can
need
as
a
stage.
Having
said
that,
if
you
store
something
of
liquid
storage
or
an
IPSS,
your
smart
contract
can
still
read
it
in
a
trust.
This
way
it
has
the
hash
of
the
content.
B
A
A
So
there's
an
IP
FS
front-end
that
that's
what's
displaying
what
you
see
on
this
screen
here.
So
you
see
like
these
images
of
the
the
water
drips,
like
that's
all,
being
hosted
on
ipfs.
So
the
very
first
thing
you
do
in
this
application
is
you
create
an
account?
I,
don't
need
a
wallet.
I
don't
need
scatter
open-eye.
All
I
need
is
a
username
and
a
password.
A
B
A
Assign
a
transaction,
though
you
need
a
private
key,
so
without
a
private
key
being
visible
to
everyone
on
a
blockchain
like
how
could
you
have
the
application
know
what
the
private
key
is
to
interact
with
the
contract
without
the
DSP
themselves,
knowing
your
private,
key
or
I?
Guess
the
blockchain
itself,
knowing
your
private
key
right.
B
So
the
way
we
do
it
in
the
demo
is
the
liquid
test
application.
Is
we
generate
the
private
key
from
the
username,
the
password
assault
assault
for
the
application?
That
way,
the
keys
can
be
reconstructed
from
from
those
things,
but
the
implementation
itself
requires
a
privately
the
liquid
accounts
on
the
native
level
it
uses
private.
Is
it
and
as
the
building.
A
B
A
Keep
those
theoretical
is
coming
because
the
whole
point
of
this
conversation
is
to
drive
attention
and
ideas
for
the
hackathon
coming
up
and
all
of
these
theoretical
it
could
actually
happen
and
what
better
motivation
to
do
it
than
fifty
fifty
thousand
dollars
and
prizes
from
being
able
to
work
from
home.
So
back
back
to
the
liquid
chess
I
keep
going
all
over
the
place,
there's
just
so
much
to
talk
about,
so
we
got
to
the
liquid
accounts.
So
you
go
down,
you
create
your
username
and
password.
A
Then
you
check
the
box
play
against
the
computer
AI.
So
what
if
you
see
this
like
we're
used
to
like
regular
applications
like
if
I'm
playing
a
regular
web
app
playing
against
the
AI
is
nothing
new,
but
this
AI
is
different
because
it's
actually
on
the
blockchain.
It's
not
off
chain,
like
everyone
thinks
that
something
like
this
needs
to
be
off
chain
and
it
did
up
until
now.
So
how
did
you
get
an
AI
on
a
blockchain.
A
B
B
A
B
A
B
So
in
the
first
place,
it's
very
hard
to
even
do
the
validation
in
such
limited
constraints
in
the
simple
plus
the
original
optimization.
The
original
implementation
is
it's
already
quite
optimized.
It
uses
things
like
these
boards
to
do
tricks
like
big
masking
numbers
that
represent
ports.
Think
about
it.
You
have
you.
A
B
8
by
8
boards
in
chess
and
you're
64
bits
on
numbers,
so
you
can
do
arithmetic
operations
that
help
you
represent
numbers
as
states
of
the
board
and
if
you
do
it
efficiently
enough,
it's
it's
mostly
arithmetic
and
in
shifting
on
64-bit
numbers.
So
it's
already
quite
optimized
but
thin.
The
constraints
of
blockchain
is
is
something
that's
not
not
scalable
enough
to
do.
Move
eliminations
for
everyone.
B
B
I
would
probably
needs
weather
check
the
computation
somehow
and
then
it
would
and
and
they
need
to
optimize,
not
just
optimize-
maybe
maybe
just
even
just
the
chunking
part
taking
existing
code
and
trying
to
change
it
to
30
millisecond
chunks
of
computation
is
not
trivial.
You
need
to
understand
the
code
that
you're
taking
in
this
think
about
it.
Just
throwing
it.
B
A
Because
of
V
CPU,
you
didn't
have
to
modify
the
code.
Much
you
took
physical
droid
fish
is
that
the
engine
stop
fish,
sorry
yeah!
So
you
you
took
an
existing
C++
written
chess
engine
called
stock
fish
which
it's
very
pod
I.
Think
droid
fish
is
a
different
one
and
you
what
kind
of
modification
so
you
just
went
to
github.
You
fork
the
code
downloaded
the
repo
and
you're
like
what
modifications
needed
to
be
made
to
that
before
you
could
use
it
in
your
contract.
There.
B
A
A
Very
small,
small
things
so
I
think
this
is
such
an
important
thing
to
highlight
about
what
you
can
now
do,
because
you
were
like
the
whole
point
of
that
network
is
that
things
that
are
redundant
and
being
built
over
and
over
again
by
different
teams
like
they
shouldn't
have
to
do
that,
build
whatever
your
try
to
do
with
your
application.
You
shouldn't
have
to
build
these
foundation
layers
to
your
application,
and
the
chess
engine
is
a
very
foundation
layer
to
your
application.
A
A
A
B
B
A
A
board
so
it
it.
That
means,
if
you
have
to
do
anything,
that's
more
computation
intensive.
Then
you
can
accomplish
in
30
milliseconds
the
transactions
gonna
fail
and
get
aborted
sure.
So
how
did
you
hack
that
out
of
the
system
like
that
is
like
built
into
the
system?
You
shouldn't
be
able
to
break
that,
but
you
got
around
it
right
and
when
I
say
break
I
mean
you
met
made
it
better.
Yeah.
B
A
B
B
B
There
is
no
to
allow
simulation
of
things
that
are
bigger
than
30
milliseconds,
but
one
of
the
things
that
taking
the
hosting
environment
from
the
nodes
for
this
relation
part
can
actually
gives
us
some
very
interesting
potential
capabilities
like
exposing
GPU
interface
to
that
contract
execution.
Whatever
is
running
on
the
CPU
and
can
actually
utilize
GPU
acceleration
that
we
can,
we
can
expose
things
that
were
not
originally
exposed.
A
A
A
The
description
obviously,
and
at
the
hacker
EarthLink
also
earlier,
you
brought
up
how
used
an
open
source,
c++
chess
library,
there's
plenty
of
machine
learning.
Libraries
in
AI
libraries
that
already
exists
so
you're
saying
that
George
yeah
these
libraries
could
be
actually
put
onto
Lake
ESI,
oh
through
the
DAP
Network,
and
you
could
be
running
with
very
little
modifications
here
and
there.
You
could
be
running
these
battle-tested
machine
learning,
engines.
B
B
B
A
A
B
A
A
B
B
A
A
A
B
B
And
for
years
we've
been
working
and
labeling.
Those
data
sets
for
those
companies
for
those
corporates
can
we
move
those
applications
to
people
and
then
decentralize
we
won't
have.
We
won't
have
the
data
even
even
on
the
block,
so
that's
a
critical
part
for
it,
and
actually
it's
another
reason
for
us
as
users
to
care
right
now.
We
we
don't
see
the
implications
of
that.
We
just
see
the
inconveniences
of
you,
know:
big
corporations
controlling
air
or
data
like
that
they're
able
to
delete
it
or
monetize
on
it.
B
A
B
There
are
two
types
of
data
that
I've
seen
the
future:
there's
private
data
and
public
data,
there's
transparent
data
and
and
private
data.
There
are
many
ways
to
do
decentralized,
logics
logic
on
private
data.
Most
of
them
involve
things
like
zero
knowledge
and
there's
public
data.
Public
data
shouldn't
can
and
should
be
transparent.
It
shouldn't
be
accessible
in
the
same
way
for
everyone
right
now.
We
even
know
the
public
data
support.
Facebook
is
not
something
that
you
can
easily
access
and
processes
as
a
researcher.
B
A
B
A
The
machine,
learning,
algorithms
and
and
the
a
is
how
how
advanced
could
you
get
with
that
before
someone
decides
to
build
something
that
extends
the
G
the
into
a
V
GP
or
whatever?
They
might
call
it
if
they
want
to
build
it
like.
If
you
don't
have
that
computation
power
today,
because
the
hackathons
in
two
weeks,
I,
don't
I,
don't
think
that'll
be
built
yet
by
anyone.
Maybe
unless
some
DSPs
been
working
behind
the
scenes,
we
don't
know.
B
B
B
A
B
B
B
B
In
ported
to
technology,
to
make
it
decentralized
so
I
think
JavaScript
too
wasn't
compiling.
We
will
open
the
potential
audience
of
people
writing
decentralized
applications.
This
will
be
CPU
and
removing
the
limitations
can
actually
allow
any
kind
of
developer
to
to
write
blockchain
decentralized
applications
with
without
really
knowing
anything
about
cryptography
or
without
knowing
anything
about
they
are
underlying
magic.
A
B
A
We
had
some
really
big
news
come
out
today,
which
I
hope
is
today,
because
we're
gonna
do
some
magic
here.
If
it's
not
today,
this
video
will
come
out
whatever
day.
That
today
was
whatever
interoperability
has
been
a
theme
of
the
message.
We've
been
trying
to
convey
as
liquid
apps
a
multi-chain
vision.
So
to
speak,
and
as
of
a
couple
hours
ago,
we
now
have
the
ability
to
use
liquid
x.
Why
don't?
Why?
Don't
you
just
introduce
it
here
in
case
anyone
hasn't
read
the
blog
order,
who
yet
sure.
A
A
A
B
Yes,
yes,
I
own
community,
we
have
even
we
have
some
tribalism
even
inside
the
community
simulation
evening's.
Are
there
on
small
community
and
I
hope
this
means
we.
We
have
to
really
bridge
that
to
amplify
all
the
great
advantages
that
there
are
all
those
different
sister
chains,
so
there
are
no
longer
experiments
for
for
public
chain
or
vacation
specific
chain.
I
think
think
that
we
we
had
those
those
systems
to
take
those
those
amazing
innovations
that
they
did
as
developers
and
as
communities
and
and
actually
bring
back.
That.
A
A
A
Let
me
backtrack
because
a
lot
of
times
when
we
discuss
these
services
when
we're
talking
about
it
on
the
main
that
we're
talking
about
saving
on
on
resource
costs,
but
on
other
chains,
the
resource
costs
aren't
nearly
the
same.
So
there
still
is
a
savings
on
resource
costs.
If
you
use
something
like
VRAM,
but
it's
maybe
not
as
critical
but
there's
other
aspects
of
VRAM.
Besides
cost
that
is
going
to
be
game-changing
for
it.
For
these
sister
chains,
could
you
kind
of
just
highlight
maybe
some
of
the
things
that
vram
allows
for
an
application?
B
A
B
You
can
aggregate
cue
of
messages,
one
chain
and
dispatch
it
once
a
minute
with
liquid
scheduler
and
then
process
it
on
the
other
chain
and
kind
of
an
infinite
bandwidth
between
the
chains,
which
you
can
do
almost
anything
you
can.
You
can
speed
up
your
own
chain
for
things
that
you
trust
only
yourself,
and
you
know
that
you're
gonna
trust
you.
You
basically
get
the
full
spectrum
of
centralization
and
validation
and
efficiency
to
build
your
own
postal
needs
for
I.
A
Want
to
get
into
one
more
thing
before
we
end
this,
but
let's
close
out
the
hackathon
here,
so
the
hackathons
open
anybody
in
the
world
that
included
and
any
any
blockchain
really,
but
especially
with
liquitex
any
ears
I/o
chain
and
then
aetherium
more
than
welcome
to
join
I
thought.
We
talked
about
this
the
other
day
and
we
were
half
joking,
but
this
hackathon
is
different
than
the
EOS
hackathons.
Last
year
the
EOS
hackathons
last
year
were
sponsored
managed
organized
by
block
one.
B
Not
one
I
think
it's
a
nice
and
a
great
opportunity
for
for
the
people
from
block
one
to
actually
participate.
I
know
that
if
I
were
then
I
would
love
to
participate
in
I
mean
our
own.
A
Cotton's
and
I
know
that
their
opportunity,
until
now,
not
just
to
you,
know
to
not
just
to
show
how
good
they
are,
but
to
actually
build
things
and
in
public
that
aren't
related
to
their
name
there.
There
may
need.
A
A
Block
one
will
just
give
their
employees
their
blessing,
I
guess
I,
don't
know
if
they'll
outright
say
it,
but
I
would
love
for
block
one
employees
to
participate
in
this
hackathon
I'm,
not
just
talking
about
Dan
Larimer,
obviously
Dan
wanted
to.
We
would
let
him
participate
in
whatever
way
he
wanted
to
well,
but
I'm,
just
thinking
like
anyone
there,
anyone
that's
working
with
it.
This.
A
A
Hope
it
happens
because
I
think
once
they
dig
into
that
network
tools,
more
they're,
going
to
understand
the
power
that
it
unlocks
and
I
hope
that
they
end
up
having
the
same
respect
for
the
tools
that
you've
built
as
the
respect
that
you
have
for
the
tools
that
they
built
and
are
currently
still
building
and
that
drives
us
into
like
the
final
topic.
I
want
to
hit
on,
and
that's
just
es
IO
in
general,
so
back
in
at
b1
June.
A
Yes,
I
o
2.0
was
announced
along
with
ESB
m
web
off
and
a
lot
of
other
tools
and
improvements,
but
it
was
just
very
recently
a
couple
weeks
ago
that
they
actually
released
the
code
base
for
OCIO
2.0.
It's
still
a
release
candidate,
but
I
just
wanted
to
kind
of
get
your
thoughts
about
some
of
these
improvements
that
are
coming
with
it.
There
are
a
few
features
that.
A
B
A
B
A
That
that
was
one
of
the
promises.
Wasn't
it
during
the
ESI
CEO.
Was
this
horizontal
scalability,
parallel
processing,
so
I
guess
what
you're
saying
is
that
es
VM
increases
the
throughput
of
that
single
thread.
It's
like
a
one
lane
highway.
That
just
is
good.
The
speed
limit
is
increased
on
the
highway
yeah.
A
Processing
is
it's
very
important
yeah,
but
what
parallel
processing
does
is
it
turns
that
one
lane
highway
into
an
infinite
lane
highway
as
wide
as
you
want
to
go,
and
that's
that's.
What
parallel
processing
isn't
the
car
is
being
able
to
go
faster
is
just
that's
cool
too,
but
you're
still
limited
to
that
one
lane
and
what
you're
saying
is
with
vcp.
You
can
go
parallel.
Yeah,
the
part
of
the
see
SiO,
2
blog
release
that
block
one
put
out.
A
A
B
Aside,
like
setting
up
a
devout
mediator,
if
you
want
to
combine
it,
it
takes
hours,
takes
hours
to
get
an
environment
environment
running
before
they
organize
their
releases.
Now,
it's
a
bit
better,
but
but
still
there,
the
entire
environment
and
the
testing
environment
and
having
a
local
chain
running
I.
Think
that's!
It's
really.
B
A
Going
to
be
invaluable,
I
think
for
the
hackathon,
because
I
remembered
the
iou's
hackathons
like
the
first
couple
hours
for
some
people
were
spent
just
getting
set
up
if
they
didn't
prepare
ahead
of
time
were
already
had
been
a
developer
on
yes,
prior
to
it.
So
if
this
is
a
tool,
that's
available
for
people
in
adapt,
hackathon
that
that's
going
to
be
huge
and
it
actually
piete
piete
piete
k
didn't
know
that
net
was
working
on
this.
A
B
A
B
It's
actually
a
good
point
in
general,
I
participated
in
many
hackathons,
not
just
as
mentor,
and
one
of
the
most
important
thing
is
to
come
recur
it's
to
reign
in
advance.
What
you're
not
not
to
try
to
learn
both
the
technology
stack
and
what
you
need
to
do
for
your
application
at
the
same
time,
because
the
hours
are
precious,
so
we
want
to
come,
prepare
for
every
habitat
and
then,
especially
in
terms
of
the
actual
technical
environment.
A
We
should
close
this
out
with
a
legendary
towel
story.
I
know
there's
so
many
legends
and
you
just
said
you've
participated
in
many
hackathons,
not
just
as
a
mentor.
What's
the
what
give?
Let's
close
this
out
on
a
towel
story
and
it
doesn't
even
have
to
be
like
a
competitive
hackathon,
it
could've
been
a
weekend
hackathon
at
your
house
like
what's
something
you've
done
in
a
hackathon
style,
where
it's
like
that,
when
that
short
window
of
time,
we
just
get
something
done
like
what's
something
that
you
want
to
share
with
everyone.
B
B
But-
and
someone
already
has
had
hacked
it,
but
the
hacked
wasn't
wasn't
completely
enough.
You
couldn't
use
to
the
two
states,
so
a
big
portion
of
the
hackathon
was
trying
to
hack
it
in
a
different
way
that
you
could
actually
reuse
the
remote
without
the
way
itself,
just
with
the
laptop,
and
so
that
was
the
end
of
the
first
day
and
the
second
way
we
just
you
know,
build
drama
shading
you
using
using.
B
B
A
Awesome
so
I
guess
sign
up
for
the
DAP
Network
hackathon.
You
could
sign
up
at
DAP
network
dot,
hacker
earth
comm
the
sign
up
informations
in
the
description
below,
along
with
some
other
resources,
to
learn
about
DSP
services.
There's
also
information
on
the
hackathon
page
I
encourage
everyone
to
sign
up
early.
We
want
to
kind
of
be
able
to
plan
for
this.
We
want
to
start
the
mentorship
early,
we're
gonna,
start
rolling
out
emails
and
education
material.
A
It's
like
tau
said
you
want
to
be
prepared
for
this
and
we're
gonna
communicate
to
our
general
channels
as
well,
but
we'll
also
be
doing
like
direct
messaging.
We
have
a
slack
group
set
up
for
the
hackathon.
We
want
everyone
to
get
started
as
soon
as
possible.
So
if
you're
watching
this,
you
haven't
signed
up
yet
head
on
over
to
DAP
network
dr.
earth
comm.
So
that's
my
call
to
action.
You
have
anything
you
want
to
tell
anyone.