►
From YouTube: CasperLabs Community Call
Description
Rewards Distribution presentation & status update.
A
B
B
B
Well,
we
promised
to
deliver
a
test
net
by
the
end
of
the
month
and
I'm
happy
to
report
that
that
is
well
underway.
We
will
be
launching
the
network
this
afternoon
at
1
p.m.
Pacific.
We
have
our
wonderful
validators
to
thank
for
their
diligence
in
provisioning,
the
hardware
and
spending
the
time
to
get
set
up.
Thankfully,
this
the
setting
said
getting
set
up
was
not
that
much
work.
B
We're
happy
to
report
that
we
started
working
with
validators
a
couple
of
days
ago,
and
a
majority
of
them
are
configured
and
set
up
and
we
are
just
preparing
the
final
touches
on
our
nodes
and
some
of
the
infrastructure
before
we
are
ready
to
launch.
So
we
will
be,
you
know,
sharing
out
the
chain
specification
here
shortly
with
the
validators
via
discord.
This
is
where
the
validators
have
chosen
to
coordinate
the
majority
of
them.
You
know
preferred
discourse,
so
we're
coordinating
there.
B
So
if
you
want
to
observe
the
launch
of
the
network
join
us
on
our
discord
channel,
you
can
get
to
it
from
our
website.
So
with
that
I'll
dive
in
we
cut
node
17
last
night.
That
is
the
release
that
will
power
the
test
net
and
we
have
21
validators
participating
in
the
test.
Net
team
is
in
the
second
week
of
sprint
33.
B
B
We
are
launching
our
test
net
with
an
alpha
version
of
the
highway
consensus
protocol.
This
assumes
honest
validators,
but
it
does
support
the
era's
a
fixed
round,
exponent
and
other
highway
parameters.
We
don't
yet
support
bonding
and
unbonded,
so
the
validators
needed
to
coordinate
a
little
bit
with
us
to
give
us
their
public
keys,
so
we
could
put
them
in
the
Genesis
block,
so
bonding
unbonded
flashing
rewards.
These
are
all
economic
security
parameters.
These
are
not
part
yet
of
the
protocol.
We
do
have
the
Genesis
process,
of
course,
graph
QL
structured
logging.
B
This
is
really
our
first
chance
to
get
feedback
on
the
node
software
and
really
see
how
the
network
the
protocol
performs
with
a
distributed
set
of
validators
right.
How
easy
is
it
to
use
how
easy
it
to
monitor
and
manage
so
we're
gonna
learn
a
lot
to
this
process.
We
have
a
clarity
instance
of
the
test
net.
B
Of
course,
this
is
not
up
yet
because
the
network
hasn't
yet
launched
will
maintain
definite
for
a
while,
but
we're
gonna
be
decommissioning
it,
and
then
we
will
redirect
both
the
deploy,
endpoint
and
the
clarity
endpoint
towards
testing
it.
But
that'll
be
a
few
days
from
now.
It's
not
quite
ready.
Yet
our
current
focus,
you
know
we're
focused
on
testing
and
debunking
the
protocol
and
making
sure
that
test
net
is
stable
and
reliable.
B
We're
working
on
the
s
tests
harness
to
make
it
scale
it
out.
Actually,
we
Durant
a
small
test
where
we
needed
actually
more
s
tests,
instances
in
order
to
load
the
network
up,
which
is
a
great
problem,
have
we're
also
working
on
the
vulnerability
we
found
in
the
protocol
for
spam
protection.
This
is
the
equivocation
bomb
and
so
we're
updating,
integrating
and
updating
those
proofs
and
then
we're
also
updating
the
technical
specifications
and
we're
doing
research
on
our
pricing
model.
B
Formalizing,
the
approach
by
using
we're
gonna
adapt
reliable
broadcast.
It's
not
going
to
be
specifically
like
the
traditional
notion
of
reliable
broadcast,
but
a
flavor
of
it.
Andreas
found
actually
an
innovative
way
to
implement
this
and
make
the
protocol
simpler,
which
is
great.
Having
a
simpler
protocol
is
always
excellent.
It's
a
great
direction
to
go
into
and
we're
implementing
deploy
gossipping.
This
is
really
important.
Deploy
gossipping
is
not
part
of
the
alpha
test
myth,
so
we've
had
to
do
a
little
bit
of
you
know.
B
Workarounds
on
the
the
the
god
deployed,
gossipping
portion,
the
you
know
not
the
least
of
which
is
the
deploy
endpoint,
but
we're
going
to
be
implementing
deploy
gossipping
in
our
first
update
of
the
network
for
sure,
and
then
we've
also
already
received
some
feedback
from
validators
about
some
more
metrics
from
monitoring
the
protocol.
We've
got
pull
requests
open
for
that,
so
that'll
be
another.
Another
update
will
do
the
protocol.
B
We
have
some
documentation,
we
have
to
do
on
assembly
script
and
tutorials.
So
we,
as
you
may
remember,
we
launched
support
for
assembly
script
and
smart
contracts,
but
we
need
to
now
document
how
to
use
that.
So
that's
that's
coming
in
and,
of
course,
the
execution
engine
team
is
supporting
the
launch
of
the
Testament.
B
B
We
had
to
make
some
clarity.
You
know.
Clarity
had
to
be
enhanced
to
work
with
highway.
Highway
has
a
lot
of
ballots
in
there
right,
so
we
want
to
be
able
to
filter
for
ballots.
Do
some
more
analytics
do
deploys
from
the
browser?
So
there's
some
good
interesting
things
that
we
wanted
to
do.
There's
a
typo
in
here
and
it's
annoying
at
me,
so
I'm
going
to
go
fix
that
where's.
That.
B
B
And
then,
yes,
we
gotta
get
the
deaf
developer
guide
updated.
We
have
quite
a
few
updates.
You
need
to
make
there
too,
because
we've
also
implemented
the
type
system
at
the
client
which
allows
you
to
pass
in
full
CL
types
into
your
deployments
as
parameters
and
do
things
like
cool
things
like
creating
a
batch
Trent's,
a
batch
token
transfer
that
transfers
token
to
multiple
accounts
at
the
same
time,
in
a
single
transaction.
B
So
there's
some
interesting
things
you
can
do
with
CL
types
at
the
at
the
client
interface
right,
so
we
gotta
get
some
documentation
in
the
DAP
developer
guide.
We
also
started
building
our
node
operator
guide
our
validator
guide,
so
we
have
an
alpha
version
of
that
for
the
test
net
validators
that
we've
been
sharing
with
them
and
we
will
continue
to
iterate
on
this
with
feedback
from
them,
and
we've
got
some
researcher
redoing
on
rewards
calculations
and
we're
thinking
a
lot
about
transaction
pricing.
B
B
I,
like
I
love
the
ideas
you've
got
so
today.
Alexander's
gonna
talk
a
little
bit
about
a
potential
Oracle
implementation,
which
or
Oracle
strategy,
which
many
of
you
have
may
have
heard
in
the
blockchain
space.
But
then
next
week
on
or
is
gonna
talk
about
his
idea,
which
is
super
interesting
and
as
well.
So
we're
looking
at
several
avenues
here
and
how
we're
gonna
fix
the
transaction
transaction
pricing
problem.
This.
A
B
That's
exactly
right,
so
we
have
some
exciting
announcements
coming
this
week,
so
I'm
a
lot
of
these
announcements
are
out.
So
if
you
want
to
learn
more
about
them,
go
and
hit
or
a
telegram
channel
you'll
find
some
information
about
that
out
there.
It's
it
pertains
to
everybody
except
us.
Persons
so
know
us
folks
can
participate
in
this
exciting
announcement,
but
if
you're
not
in
the
United
States,
you
certainly
can
go
check
it
out
on
telegram.
D
D
D
D
Host
functions,
in
fact,
and
finally,
in
my
end,
I've
been
evaluating
some
of
the
options
we
have.
You
know,
for
you
know,
nailing
down
the
structure
of
the
link
between
our
node
software.
You
know
it's
a
production
process
and
external
sources
of
truths
that
would
actually
provide
us
CL
Xing
is
the
price
fair,
so
today,
I'll
be
talking
in
something
about
this
last
point,
and
so,
first
of
all,
very
you
know
evaluating
multiple
options.
Instead
of
you
know
what
say
taking
something
off
the
shelf.
Well,
there
are
a
few
reasons.
D
The
most
important
reasons
is
that
there
is
a
pretty
sharp
trade-off
between
some
of
the
criteria.
If
you're
considering
you
know
as
our
success
benchmarks.
For
this
next
slide
reason
number
two
is
at
this
paradigm
is
relatively
novel.
There
does
not
seem
to
be
a
blockchain
that
is
running
right
now
that
actually
implements
a
link
between
you
know.
Let's
say
real
world,
the
price
was
a
token
and
their
work,
its
transaction
schools
right,
you
know,
there's
it's
a
basically
price
and
it's
transactions
in
dollars.
D
So,
of
course,
you
know,
I
mean
we
need
to
be
careful
here,
because
we
want
to
get
this
right
for
the
users.
Reason
number
three
is
it
you
know
there
are.
You
know
canonical
implementations
of
these
Oracle
links
for
other
purposes,
but
they
don't.
You
know
the
disapprobation
super
cell
email
down,
one
of
the
free
options,
considerate,
they're,
actually
combative,
competitive,
more
or
less,
is
all
three
depending
on
how
you
structures
or
contracts
that
are
involved
here.
D
So
you
know
so
there's
these
traders,
if
you're
talking
about
the
trade-offs
are
between
you
know
our
ability
to
succeed
on
the
three
criteria
that
and
they
deal
solutions
that
doesn't
exist,
satisfied
right.
This
reliability,
decentralization
and
accuracy.
Reliability
means
that
you
know
the
exchange
rate
that
was
used
by
the
platform
is
updated.
You
know
fairly
frequently
right
so,
ideally
under
ten
minutes,
you
know,
maybe
you
know
even
under
five
minutes,
you
know
so
tracks
very
close
to
expand
and
cryptocurrencies
mark
ii.
D
This
every
single
element
of
our
design.
We
want
fully
decentralized
system
and
they're.
Finally,
well,
you
know.
We
also
want
this
thing
to
be
accurate,
of
course,
so
unfortunately,
there's
not
going
to
be
a
solutions
that
will,
you
know,
be
best
on
all
three
of
these.
So
that's
one
of
the
reasons
why
we
going
to
be
considered
multiple
options,
and
so
now
we
will
go
to
our
internal
document.
D
Is
that
actually
has
diagrams
and
describe
what's
happening
so,
first
of
all,
so
we
are
going
to
go
through
four
well,
actually
free
of
these
diagrams
and
the
first
from
going
to
go
through
this.
You
know
diagram
of
how
the
platform
or
the
steel
node
sees
this
process
right,
so
the
so
here's
the
Oracle
link
is
a
sort
of
a
completely
abstract
entity.
We
don't
really
know
how
it
works.
D
C
D
Right
and
the
parameter
key
from
this
block
and
the
parameter
K
is
basically
blocks
in
this
scene
of
main
parents.
Since
the
last
price
update,
as
once
it
occurs,
Reds
it
it.
You
know,
executes
the
transactions
from
transaction
buffers
at
you
knows
a
given
price
and
definitely
makes
a
request
for
a
new
price
if
ki
is
greater
than
this
parameter,
capital
n,
which
designates
you
know
how
many
blocks
you
wait
before
you
request
another
update,
of
course
right
so
well,
if
that
actually
happens
in
the
execution
engine.
D
On
top
of
you
know,
processes
as
regular
transactions
also
execute
request
to
this
Oracle
contract
assemblage,
which
we
will
analyze
detail,
and
you
know
basically
large,
is
a
request
for
a
new
price.
So
all
of
this
being
done
is
the
node
weights
new.
This
is
this
is
just
how
it's
just
a
description
of
very
high-level
description,
how
it
works
your
perspective
of
the
cash
prolapse.
Note
all
right
so
well.
What
does
it
look
like
from
the
perspective
of
the
Oracle
räikkönen?
D
D
You
know
I
took
about
these
as
being
like
regular
contracts,
almost
right
with
special
features,
but
you
know
for
performance,
as
they
might
end
up
being
embedded
in
the
execution
engine
itself.
So
our
first
option
is
decentralized
oracle,
and
this
is
going
to
very
much
like
you
know
what,
for
example,
chain-link
describe
since
they're
white
paper,
and
so
what
happens
here
is
a
prospective
oracle
node,
which
is
you
know,
conceptually
machines,
the
transpose
some
proprietary
software.
It
generates
CPI
your
quests
and
you
know,
checks
up
on
these
contracts
in
the
oh
is
also
a
CL.
D
D
So
the
selection
contract
here
is
abstract.
You
know
so
very
fine
eligibility
may
need
many
different
things
right.
It
may
mean
paying
a
fee.
You
know
submitting
some
kind
of
a
deposits.
I
don't
know,
provide
the
proof
of
work,
it
could
mean
just
you
know,
just
verifying
the
cryptographic
who's
it.
You
actually
have
access
to
the
you
know.
Api
is
that
you're
supposed
to
cover
access
to,
but
whatever
it
is
you
some
Co
verify
your
eligibility
and,
let's
suppose
it
you
know,
you
pass
the
checks,
and
now
you
are
an
errand.
D
D
Price
right
so
well
update
actually
goes
into
this
work.
Flow
contract
of
the
work
flow
contract
now
has
a
flags
that
you
know
we
need.
You
know
we
need
a
price
update
and
essentially
continuously
reading
from
contract,
to
see.
If
you
know
a
price
update
is
requested,
so,
let's
suppose
the
price
the
price
update
is
requested.
So
then
the
Oracle
notes
goes
to
its
data
source.
D
Reads
from
that
API,
then
it
goes
back
to
the
platform
and
writes
the
new
price,
possibly
some
other
information
such
as
you
know
when
it
was
retrieved
where
it
was
retrieved
from
to
would
equals
the
aggregation
contract
and
again.
This
is
at
this
point
a
fairly
abstract
entity.
It
can
do
many
things,
but
as
an
example,
you
may
think
of
aggregation
contract.
Simply
you
know
taking
the
average
of
you
know
whatever
you
know,
whatever
prices
are
submitted
to
it
and
then
the
platform.
D
Now,
after
this
happened,
we
have
the
platform
runs
the
reward
contract
which,
based
on
Nino's
data
contained
in
the
selection,
contract
and
segregation
contract,
distributes
rewards
from
a
transaction
fees
to
the
Oracle
nodes.
Right
and
again,
this
is
abstract.
There
may
be
many
ways
to
implement
it,
and
you
know
we
are
not
yet
settled
on
any
particular.
D
Most
basic
one
is
the
Casper
labs
Oracle.
The
gas
problems
Oracle,
as
you
can
see,
is
going
to
cut
out
some
of
these
contracts
because
they're
no
longer
necessary
right
I
mean
you
don't
need
to
reward
our
own
Oracle.
We
don't
need
to
verify
its
suitability,
so
it
really
does
is
reads
from
the
work
flow
contract
and
then
goes
to
the
data
retrieval
procedure,
you're
still
quarreling,
there's
still
Nagre
Gatien
contract
and
they're
still
:
its
aggregation
contract,
because
potential
is
node,
might
submit
in
multiple
prices
and
them
they're,
aggregated.
D
The
third
option
we
are
considering
are
is
sort
of
a
variation
on
the
decentralized
oracle
is
a
particular
eligibility
criterion.
So
legibility
criterion
is
that
if
you
know
who
you
are
basically
right,
so
the
most
basic
case,
the
Oracle
knows,
would
simply
be
operated
by
the
exchanges
themselves,
and
so
these
are
the
three
options:
we're
considering
and
being
a
well.
How
do
they?
You
know
possibly
compare
on
our
reliability.
Decentralization
accuracy
criteria.
D
Note
that
this
is
very
early
stage,
of
course
you
know
this
assessment
is
going
to
change
once
we
come
up.
You
know
actual
designs
for
these
and
so
well.
You
know
it's
very
simple.
Of
course,
decentralized
oracle
is
most
decentralized
them
all.
The
problem
is
that,
well,
it
is
actually
difficult
to
design.
You
know,
incentivization
schemes,
you
know
to
have
these
up
pseudonymous
participants
report,
you
know
correct
prices
right
and
this
problem
also
plagues
as
accuracy.
D
The
main
process
fundamental
problem
here
is
that
Oracle's
are
not
really
resilient
against
the
simple,
so
you
know
you
can
really
do
something.
Basically,
just
you
know
penalize
people
from
deviation,
because
one
honest
node
and
this
report
is
the
truth
and
every
other
node
is
operated
by.
You
know
a
malicious
attacker
who
pretends
that
they're
all
separate
nodes,
but
you
know
reports
surprises
that
are
too
high
or
too
low,
and
you
know
there's
no
way
to
detect
that
you
know
using
some
very
basic
technique.
D
It's
in
fact,
probably
in
order
to
deter
that
you
would
have
to
do.
Some
kind
of
you
know
know
your
customer
die
process
and
at
that
point
you're
already
in
the
no
in
the
land
of
proprietary
Oracle's,
so
you
know
says
they
said:
the
trade
off
is
pretty
sharp
here,
so
the
proprietary
Oracle's
are
probably
the
most
attractive
ones
here,
because
you
know
well,
you
know
they
must
be
reliable.
Essentially,
because
you
know
responsibility
lies
with
the
exchanges
here
and
the
exchanges.
Of
course
you
know,
have
reputation
to
uphold
and
the
legal
obligations
as
well.
D
We
also
talked
sure
since
I
don't
their
accuracy
is
it
have
to
be
accurate
because
you
know
otherwise
they
should
become
a
legal
problem.
The
problem
is
the
solutions
that
you
actually
have
to
get
in
those
exchanges
to
go
along
with
this
plan
and
actually
create
these
workers.
And
finally,
we
have
the
unitary
Oracle,
which
is,
of
course
not
the
centralized
in
the
tool.
But
you
know
it
is
you
know
as
reliable
as
you
can
make
it
and
there
you
know.
D
Of
course,
you
know
VR
and
since
in
two
ways
they
simply
by
our
need
to
present
a
viable
alternative
to
cerium,
and
this
is
a
you
know-
major
feature
that
the
differential
differentiates
us
from
with
you.
So
of
course
you
know
it
is
a
you
know.
It's
a
big
problems
that
you
know
it's.
You
know
it's
centralized
the
Beaufort.
You
know
just
PR
reasons
know
so
because
in
a
well
you
know
what,
if
something
happens
to
our
infrastructure.
Oracle
goes
down.
You
know
it's,
it's
nothing!
D
So
right
now,
if
you're
still
at
the
stage
where
we're
considering
these,
you
know
free
options
and
once
we
settle
down-
and
one
of
them
will
start
coming
up
with,
you
know,
actual
designs
for
the
contracts
themselves
and
that's
like
at
this
point.
This
is
still
essentially
you
know
software
design
work.
This
is
not
really
economics
work,
so
the
real
economics
work
is
actually
designed
in
things
like
you
know,
ability,
contracts
and
the
aggregation
contracts.
D
B
B
C
B
The
validators
are
very
excited,
they're
ready
to
go
so
this
is
I'm.
Super
excited
we're
just
focusing
now
on
the
last
last
mile
small
pieces
that
we
need
to
just
get
everything
buttoned
up,
but
they
are
ready
to
go
as
soon
as
we
are
so
looking
forward
to
launching
the
network
today.
So
the
test
net
is
going
live
this
afternoon,
so
lots
of
hard
work.
Coming
to
this
point,
I
want
to
thank
the
development
team
who
has
worked.