►
From YouTube: CasperLabs Community Call
Description
Development status update, What is CasperLabs' mission & why we are here and deep dive into the work on the Chainlink integration.
A
A
Been
having
problems
assume
for
a
bit
I'm
just
drop
this
shot
there.
You
go
perfect
all
right,
so
we've
entered
our
fourth
and
final
weekly
sprint
to
this
release
cycle
for
those
of
you
that
are
joining
you.
The
engineering
team
swapped
out
their
engineering
process
to
go
to
shorter
sprint,
so
we're
doing
one-week
sprints
and
we
have
a
features
backlog
of
everything
we
want
to
achieve
between
now
and
May
meant
we
operate
on
four
week
cycles
for
releases
and
so
we're
the
fourth
and
final
final
weekly
sprint
to
this
release.
A
Cycle
2.4
at
the
end
of
this
release
cycle
will
be
releasing
node
0.20
our
20th
release
and
updating
the
test
net
and
then
kicking
off
the
beta
test
net
milestone
and
so
the
beta
test
at
milestone.
We
have
publicly
said
we're
going
to
go,
live
on
July
15th,
that's
the
date
we're
targeting
it
might
go
sooner
than
that,
but
the
official
launch
date
is
the
15th.
C
A
A
Between
those
two
goalposts,
that's
right,
yep
with
the
beta
test
net
launch,
it
will
be
a
fresh
new
Genesis.
So
we'll
talk
a
little
bit
about
testing
it,
real
quick.
So
for
the
beta
test
net
timeframe
we
won't
have
bonding
and
unbonded
in
places.
Yet
we're
going
to
be
focused
fully
on
smart
contract
authoring,
we'll
be
deploying
contract,
headers,
multi,
key
support
and
then
the
host
side,
token
transfer,
contract,
yeah.
D
A
A
So
that's
the
goal
for
our
beta
test
net,
so
we're
also
going
to
be
opening
it
up
in
providing
incentives
for
adapt
developers
on
the
alpha
test
at
time
frame.
Our
incentives
are
product
primarily
on
validators,
but
we're
also
offering
incentives
for
DAP
developers
during
the
beta
test
net
and
o'nora
I'd,
like
for
you
and
I,
to
work
on
that
and
figure
out
what
those
incentives
look
like.
So
we
can
talk
publicly
about
them.
We
probably
need
to
get
that
done
this
week,
so
we
can
talk
about
it.
A
I'm
talking
it
no
I'm
talking
about
DAP
developer
incentives
to
participate
in
the
test,
all
right,
so
the
beta
timeframe,
we're
gonna
start
opening
it
up
for
developers
to
start
building
on
the
work,
and
we
want
to
provide
incentives
to
them
to
do
so.
Yeah.
So
we've
got
validator
incentives.
We
had
an
elephant
will
continue
to
have
validator
incentives
for
beta,
but
we
also
have
a
pool
that
we
want
to
allocate
to
deaf
developers
as
well
to
encourage
them
to
start
building
yeah.
A
Okay,
so
during
the
testing
phase,
we've
discovered
18
bugs
in
total,
primarily
on
the
consensus
layer.
That's
really
our
focus
Ben
and
we've
fixed
13
of
them,
and
we've
got
five
more
that
are
being
triaged.
It's
our
intention
that,
with
the
next
release,
we'll
be
deploying
all
13
of
these
bug
fixes
at
a
minimum.
Each
bug
fix
has
gone
through
a
minimum
of
one
to
two
weeks
of
hardening
and
testing
and
what
we
call
our
LR
tees.
Our
long-running
test
beds,
long
running
tests,
test
beds,
where
we
validate
that
the
network
is
stable.
A
Hot
off
the
presses,
there's
Alex
great,
welcome
Kurt
focus
on
highways,
so
for
those
of
you
that
are
familiar,
we
are
doing
a
cleanup
and
reimplementation
of
the
node
in
rust
in
highway
or
supporting
the
serialization
and
deserialization
of
messages
containing
highway
vertices.
These
are
votes
and
blocks
all
the
data
structures
that
are
part
of
highway
and
then
signing
up
votes
and
validating
signatures,
and
then
communicating
detected
equivocations
for
slashing
on
the
node
rust
side.
A
We
are
you
know
the
node
rust
piece,
the
sort
of
we
look
at
the
node
container
as
basically
deploy
management,
storage
management,
communications,
core
communications
and
then,
of
course,
packaging,
right,
packaging
and
release
and
installation.
That's
all
like
that's
what
the
node
pieces
and
Consensus
snaps
on
to
that
and,
of
course,
the
global
state.
The
contract
run.
Time
runs
within
the
node
right.
So
it's
also
like
how
the
node
and
the
GRP
see
the
end
points
for
the
note,
so
how
they
accept
deploys
how
deploys
are
processed
and
sent
to
the
execution
engine.
A
The
execution
engine
then
bundles
that
up
into
blocks
and
then
passes
it
to
consensus
and
then
consensus
and
then
sends
those
messages
across
the
communications
there.
So
that's
a
node
container
that
we
are
working
on
reimplemented,
we
started.
We
went
to
the
reactor
model,
which
is
basically
like
an
enterprise,
bus
enterprise,
service
bus
model
and
this
architecture.
I
just
got
off
the
call
with
Marc
this
morning
and
he
said
he's
been
looking
at
a
lot
of
other.
You
know.
A
Big
projects
in
the
space,
mainly
Libre,
was
the
one
he
mentioned
today
and
he
says
that
they're
react.
Their
model
for
Libra
is
very
similar
to
ours
and
they
implemented
one
one
tiny
feature
he
didn't
specify
what
it
was
that
was.
That
was
something
that
we
explicitly
chose
not
to
do
in
our
system.
A
So
we
feel
we're
very
much
on
the
right
track
here
in
terms
of
what
we
built
for
our
enterprise
bus
model
for
another
node
and
so
we're
implementing
the
initial
client
API
one
to
receive
a
rudimentary
deploy,
stub
and
another
one
to
query
the
existence
of
a
deploy
by
ID
and
node
storage.
This
is,
and
then
we've
got
the
unified
storage
model.
A
So
can
you
send
me
the
pull
request
that
has
the
initial
client
API
I
want
to
take
a
look
at
that
and
I
want
to
make
sure
I
understand
what
that
API
looks
like,
because
I
want
to
make
sure
we
have
a
rest
alike
API.
That
is
easy
for
developers
to
develop
against
natively
for
web
and
mobile
apps.
Yes,.
A
I'd
like
very
much
to
see
it
and
get
a
demo
of
it
on
the
Scalla
side,
we're
fully
focused
on
just
bug,
fixes
and
hardening
of
the
protocol
before
we
get
implemented
on
test
test
net
and
then
adding
support
for
secure
Enclave
key
types.
Have
we
done
the
necessary
business
analysis
for
this
feature
Ashok?
Is
it
clearly
understood
this.
C
A
Okay,
so
it
is
well
understood.
We
know
what
type
of
encryption
curve
we
need
to
implement
and
somebody's
going
to
have
to
prototype
against
a
secure,
Enclave.
Api
is
probably
just
to
make
sure
that
they
work
so,
for
those
of
you
that
are
not
familiar,
secure,
Enclave
is
the
hardware
wallet
that
exists
on
all
iPhones
and
I.
Believe
it's
also
been
adopted
by
Android,
and
this
will
basically
enable
a
very
nice
mobile
wallet.
Workflow,
where
you
create
your
keys
in
secure
Enclave.
Secure
Enclave
does
not
allow
keys
to
be
ported
in.
A
So
if
you
want
to
use
the
biometrics
that
come
with
secure
Enclave,
you
basically
have
to
create
the
keys
natively
via
the
secure
Enclave
interface.
The
keys
never
leave
the
Enclave,
and
so
to
do
that,
we
would
need
to
support
secure
Enclave
keys
on
Casper
lives.
So,
basically
you
would
sign
your
transaction
via
the
Enclave
right,
so
you
would
select
the
account
you
want
to
use.
You
would
retrieve
the
keys
from
the
Enclave
and
then
sign
the
transaction,
and
this
turns
your
basically
turns
your
phone
into
a
treasure
if
you
wanted
it
to
do
that.
A
A
So
this
is
Ho
side
token
transfer,
so
we
figure
that
the
biggest
use
case
for
crypto
is
the
transfer
of
tokens
and
so
we're
bringing
this
in
host-side
on
the
execution
engine
which
will
allow
us
to
get
quite
a
big
performance
boost
final
cleanup
for
contract
headers
on
contract
run
time,
and
then
we're
also
going
to
we're
also
going
to
do
a
bunch
of
work
on
ecosystem
around
documentation.
So
we've
got
to
do.
A
A
A
A
That's
great
okay,
one
of
the
things
we've
been
talking
a
lot
about
internally
is
vulnerability.
We
discovered
in
the
gossiping
layer
of
the
protocol,
and
so
we
were
doing
a
little
research
on
gossiping
incentive
schemes
for
those
of
you
that
are
not
aware,
there's
limited
or
no
incentive
for
nodes
to
gossip
right,
so
particularly
read-only
nose.
So
we're
just
doing
a
little
bit
of
research
to
figure
out
what
are
the
incentive
mechanisms
around
we'd,
only
node
gossiping?
A
A
A
D
B
B
C
B
Network
I'm
excited
to
see
how
much
of
an
optic
we
can
get
when
we
correctly
price
your
codes.
We
currently
use
the
values
we
got
from
Waze
me.
We
don't
even
you
don't
even
meet
her
like
host
functions,
so
it's
a
potential
DDoS
attack,
vector
once
we
once
we
get
this
done,
I'm
expecting
and
like
much
much
higher
TPS
yeah.
A
B
Up
on
this,
the
one
on
Thursday
is
about
the
first
item,
but
I
I'm
I
want
to
have
a
meeting
with
that
team
to
finally
like
make
the
changes
in
its
source
code,
so
I
have
values
or
we
should
update
the
like
Genesis.
Will
you
change,
spec
and
stuff
Angelo,
so
Peter,
so
yeah
host
function
site
might
require
some
work,
but
already
on
the
wasp
site,
we
will
have
some
new
values
and
new.
We
changed.
B
A
Scope
it
all
out
and
get
a
plan,
sounds
great,
let's
see
so,
and
that
will
probably
go
into
the
node
rust
branch.
Isis
yeah
sounds
good
and
the
the
community
wanted
to
know.
The
question
was
so
we
run,
we
run
a
poll
each
week
and
the
community
really
wanted
to
know.
You
know
what
is
our
mission.
Why
are
we
here
and
so
I'll
talk
a
little
bit
about
about
our
mission
and
why
we're
here
so
we
set
out
so
you
know
kind
of
two
prongs
right.
So
one
of
the
prong
was
as
we
believe.
A
We
believe
that
the
CBC
Casper
type
protocol
will
provide
a
mechanism
for
blockchain
to
scale
right.
So
we
believe
that
proof
of
stake
is
the
future.
We
believe
that
the
current
proof
of
stake
protocols
are
not
decentralized
enough
and
we
believe
that
there's
a
there's,
a
path
forward
using
the
CBC
Casper
research,
and
so
we
set
out
to
build
a
blockchain
that
will
demonstrate
that
this
is
possible
and
the
only
way
you
really
know
that
a
protocol
works
is,
if
you
build
a
blockchain
around
it,
and
you
build
some
value
into
that
blockchain.
A
You
take
a
risk
and
see
if
it's
gonna
hold
up
over
the
test
of
time
and
we
started
with
mathematical
principles-
we're
not
shooting
from
the
hip
here
right.
So
we
know
that
the
protocol
is
provably
alive
and
safe.
We've
been
spending
a
lot
of
time,
doing
work
around
attack
vectors
and
during
us
literally
I'm,
pretty
sure
he
sits
at
his
desk
and
just
brews
up
new
ideas.
Every
day
he's
a
very
conservative
individual
as
he
should
be.
A
You
want
your
consensus
researcher
to
really
be
looking
with
a
hairy
eyeball
at
attacks,
particularly
if
you
want
to
build
a
decentralized
permissionless
protocol.
It's
important!
So
that's
the
first
thing,
and
we
know
that
this
is
not.
This
is
not
a
small
task
right.
We
completed
our
research
in
October
of
2019.
We
found
it
took
us
12
months
to
get
the
protocol
base
research
done
and
then
we
went
into
implementation
and.
A
A
A
A
Ultimately,
consensus
protocols
will
be,
there
will
be
maybe
a
single
one
or
two
consensus
protocols
that
will
provide
the
security
and
decentralization
that
we
need
in
the
market.
But
then,
ultimately,
adoption
is
going
to
come
down
to.
What
can
you
do
with
the
contract
runtime?
How
easy
is
it
to
deploy?
How
easy
is
it
to
build
defect,
free
code
and
smart
contracts?
A
It's
simply
too
risky
to
build
a
business
on
top
of
to
go
out
again
investor
capital
and
then
build
upon
a
technology
that
you
have
absolutely
no
control
over
businesses
simply
will
not
do
that.
I
certainly
wouldn't
do
that
and
I
think
any
any
know.
Cto
you
talk
to
will
tell
you
the
exact
same
thing
that
if
I
can't
control
it
I'm
not
going
to
build
on
top
of
it,
and
so
that's
why
we're
here
we're
here
to
bring
blockchain
to
the
masses?
A
We
want
to
scale
opportunity
for
everyone
that
includes
individuals
that
are
looking
for
financial
instruments,
people
to
change.
You
know
money
to
make
it
more
sound
and
hard
so
that
people
are
incentivized
to
save
instead
of
spend
and
can
can
can
save
their
wealth
for
developers
that
want
to
learn
about
blockchain
technology.
A
We
provide
a
workflow
and
tools
that
enable
them
to
integrate
blockchain
into
their
workflow
instead
of
having
for
them
to
cross
the
chasm
we're
going
to
meet
them
on
their
ground,
with
using
existing
IDs
and
existing
CIC
D
systems
and
existing
development
operations
processes
that
make
it
really
easy
for
you
know
and
familiar
for
them
to
work
with
our
technology.
All
they
need
to
learn
about
is
how
to
storage
work.
How
do
I
put
data
in
how
do
I
get
data
out?
What
are
the
semantics
of
the
blockchain?
What
pieces
belong
on
the
blockchain?
A
What
pieces
do
not
belong
on
the
blockchain?
How
do
I
optimize
for
costs?
These
are
the
only
things
I
need
to
learn.
Everything
else
is
as
native
for
them
as
they
can
deal
with
right.
You
can
use
C
Line
and
to
run
rust
contracts
or
any
any
IDE
of
your
choice
and
then
run
your
contracts
in
a
local
runtime
environment.
So
that's
our
mission,
that's
why
we're
here
and
if
there's
questions
about
this
as
you
listen
to
it
feel
free
to
hit
me
on
telegram.
Hit
me
on
desk
chord.
D
You
know
in
order
to
provide
us
with,
you
know,
use
the
sea
LAX
price
pairs
right
so
and
just
to
refresh
your
memory
right.
So
what
happen?
Those
who
initially
have
this
on
top
sort
of
this
entry-
and
you
know,
read
mechanism
where
an
Oracle
node
in
some
way
verifies
its
ability
to
selection
contracts.
You
knows-
and
you
know,
basically,
the
platform
publishes
requests
to
work
for
contract
and
the
Oracle
nodes.
Read
from
that,
so
what
they
procure.
D
You
know
more
detailed
terms
so
because
you
know
I
mean
okay,
so
you
know
like,
for
example,
you
know
like
what
are
these
contracts
right?
What
do
they
actually
do?
So?
Something
like
a
selection
contract
is
basically
a
registry
of
these
perspective.
Oracle
notes
right
so
you
know
somehow
they
you
know,
verifies
their
eligibility
right.
So
we
may
have
some
kind
of
a
you
know.
Existing
list
of
you
know
potential
nodes
that
you
might
want,
and
they
just
you
know
basically
registers
themselves
using
some
kind
of
a
secret
message.
D
You
know,
or
you
know,
consumables,
there
could
be
other
means
of
entry
right
I
mean
you
know
you
could
implement.
You
know
bidding
systems,
you
know
fee
systems
just
free
entries.
There
are
many
ways
to
do
it,
but
you
know
in
the
end
it
is
easiest
to
handle
in
this.
You
know
discrete
piece
of
logic
is
that
we
call
the
selection
contract
all
right.
The
output
of
the
sake
is
just
here.
Basically,
it
keeps
a
list
of
for
current
Oracle
nodes
in
the
global
state
right.
So
what
say?
Oh
these
guys
register
right.
C
D
Every
hundred
blocks,
you
could
say:
okay,
it's
been
long
enough.
I
want
a
new,
updated
price
right.
So
that's
the
simplest
way
to
do
it,
of
course,
or
you
could
do
something
more
sophisticated
because
you
know
obviously
have
to
pay
the
Oracle
notes.
That
course
link-
and
you
know
you
would
like,
of
course
to
save
on
that.
So
then
you
know
then
we
do
is
a
basic
volatility
right.
D
So
if
you're
getting
you
know
block
after
block
after
block
you're
getting
you
know
pretty
much
the
same
prices,
we
can
say:
okay,
you
know
what
you
know.
We
will
try
updating
in
like
50
blocks
and
if
it's
the
same
it'll
say:
okay,
well,
maybe
in
a
hundred
blocks
wobbly
DM
right
so
again
there
are
many
options
here
and
we'll
the
when
she
was
settle
on
something
but
right
now
I'm
kind
of
right
now
towards
this
might
look
like
alright.
So.
D
And
let's
see
right
so
and
then
there
you
know
we
get
to
the
response.
Part
of
this
and
the
segregation
conditions,
reward
contract
segregation
contract
pretty
much
says
what
you
know.
That
does
what
it
says
in
the
name
right.
It
takes
in
some
number
of
reports
right.
So
these
reports
are
basically
just
numbers,
but
it
may
utilize.
Other
information
such
as
you
know
some
quality.
You
know
properties
or
Oracle.
D
You
know
how
you
know,
and
you
know
whatever
information
you
might
want
to
get
as
it's
stored
in
the
selection
corner,
basically
right
so
enters,
and
once
it
acquired
enough
reports
either
you
know
by
some
weight
or
by
just
you
know.
Absolute
number
of
you
know,
reporting
notes.
You
know
it
basically
takes
all
these
in
the
calculate
something
right
so
most
likely
it
will
be
just
calculating
something
like
media
right.
D
Then
it
reports
a
basically
restores
this
median
in
the
global
state
and
then
the
you
know
the
node
on
our
end
starts
reading
from
that
and
executes
in
all
transactions
as
any
price
right-
and
you
know
eventually,
sometime
after
you
know,
a
reward
contract
is
executed.
Is
it
has?
You
know
recorded
all
the
reports
from
you
know
the
aggregation
contract
previously
and
this
reward
also
somehow
determines
who
gets
what
right.
D
So
this
is
actually
the
fuzziest
aspect
of
all
this,
because
it's
you
know
it's
very
and
clear
to
me
how
exactly
it
gets
incentives
right
here,
but
I
believe
that
something
may
be
found
in
either
conventional
mechanism,
design,
fear
in
economics
or
in
what's
called
elicitation
theory.
Each
year
lies
kind
of
what,
in
the
middle
between
you
know,
computer
computer,
science
and
statistics,
but
so
again
this
is
a
slightly
more
detailed
summary
of
what
we
intend
to
do
with
this.
D
And
you
know,
after
our
meetings
there
is
chain
link,
it
appears
that
they
should
be
feasible,
but
we
discovered
that
there
are
some
potential
complications
didn't
quite
anticipate
right.
So
basically
you
know
is
essentially
you
must
write.
You
know,
middleware
for
the
chain.
Look
knows
to
actually
do
all
these
things.
That
I'm
describing
here
is
the
region's
the
right
and
all
of
this
right,
so
working
understanding
how
exactly
that
works.
D
And
finally,
the
last
you
know
implication
of
verses
being
like
is
that
we
must
effectively
implement
a
link,
or
you
know
some
linkage
to
link
pardons
upon
on
our
system
right
and,
of
course,
link
is
just
an
ear,
see
six
seven
seven.
If
you
ain't
talked-
and
you
know
it-
you
know
it's
a
you
know:
it's
a
myriad
things.
D
You
know
you
have
to
read
from
that
gene
and
so
and
so
forth,
so
that
your
May
or
something
like
that
may
require
significant
modification
or
note
software,
even
though,
logically,
it
sounds
pretty
simple
right.
So
you
know
if
I
was
you
know,
I
guess.
If
I
was
writing
the
contract
on
the
theorem,
I
would
probably
choose
that
option,
but
for
our
purposes
it
seems
the
chain
link
is
in
fact
superior,
but
so
so
far
again
they
only
been
summarizing.
D
Just
you
know,
a
more
detailed
version
would
have
said
before
it
results
that
we
have
to
deal
with
a
bunch
of
other
things
that
I
haven't
mentioned
in
the
previous
presentation
right
so,
and
these
are
very
fuzzy
for
me
right
now.
But
basically
you
know,
since
you
have
to
have
a
bunch
of
fallback
mechanisms,
because
this
is
such
a
critical
part
over
the
platform
right,
you
cannot
be
cut
off
from
you
know
from
a
volatile.
D
You
know
crypto
market
right,
so
you
know
it's
very,
very
important
that
you
have
some
sort
of
fullback
and
there
are
all
sorts
of
folds
that
can
happen
right.
So
you
know
you
may
have
just
outright.
You
know
outages
right,
and
you
know
one
of
the
you
know.
Maybe
you
know
chain-link
rolled
out
an
update
and
it's
somehow
propos
notes,
or
at
least
the
probes
them.
As
far
as
communication
is
goes
right,
so
this
has
to
be
somehow
detected
right.
So
similar
situation
arise.
If
I
don't
know
some,
you
know
bug
or
some
you
know.
D
D
Well,
if
you
encounter
one
of
these
folds
well,
if
you
encounter
is
insufficient
link
for
presumably
you
know,
you
would
go
and
you
would
refill
you
know
you
would
refuse
that
contract,
but
know
that
I'm
being
deliberately
fuzzy
about
who
is
doing
the
refilling
right
I
mean
I
deal
in
the
long
term.
You
know
it
doesn't
require,
like
somebody
at
the
Gasper
laps
to
go
and
actually
center
transactions
because
sort
of
a
major.
You
know
you
know
major
centralization
issue.
D
So
if
you,
if
you
should
do
that
so
you'd
like
to
avoid
that
at
least
in
the
long
term,
but
let's
say
what
happens
if
you
encounter,
you
know
mode
false
right,
I
mean,
like
know,
start
going
to
flying.
What
do
you
do
then?
Well
one
thing
you
could
possibly
do,
which
is
kind
of
our
you
know
our
default
for
now
is
that
you
initiate
the
stuff
in
a
crowd
mode.
So
what
happens?
Is
that,
instead
of
authorised,
you
know,
gene
lick
notes?
D
D
So
you
know
you
don't
have
24-hour
outages
and
stuff
like
that.
But
of
course
you
know
these
fallback
modes,
especially
right
special.
You
know
this.
Is
you
know
what
M
cones
emergency
crowd
mode
where
essentially
anyone
to
join?
You
really
start
hitting
this
problem
with
incentives
for
truthful
reporting,
and
you
know
the
there
there's
some
discussion
of
this
issue
in
a
chain
link
sold
white
paper,
but
nothing
really
conclusive.
D
So
it's
kind
of
based
on
you
know,
informal
reputation,
you
know
not
I
think
it
would
necessarily
like
formal
reputation
great
and
so,
as
a
question
you
know
becomes
okay.
Well,
you
know
this
platform
becomes.
You
know
big
right.
It's
a
really
really
big
platforms.
There's
a
lot
of
money
floating
around.
You
really
have
to
make
this
stuff
secure
and
then
it
becomes
really
complicated.
D
Thankfully,
you
know
you
see
here
mentioning
of
mentions
of
you
know:
work
installation
of
Maine,
secondary
parent
data
on
say
in
synchronicity,
I,
believe
that
our
latest
redesign
actually
solves
this
problem,
because
I
don't
think
we
have
secondary
pairs
anymore
in
the
chain
so
and
well.
That
is
basically
it
for
today,
just
a
small
update.
A
A
Terrific
great
so
this
will
be
posted
on
YouTube.
Well
is
streaming
live.
You
can
always
go
check
it
out
later
on,
and
we
do
these
calls
every
Tuesday
morning,
I'm
pleased
to
report
I've
been
doing
them.
The
entire
time
I've
been
here
so
since
we
found
other
company
will
continue
to
do
them
for
the
foreseeable
future.
Cheers
everyone
stay
safe
out
there
and
hope
to
see
you
next
week.
Bye
thanks.