►
From YouTube: CasperLabs Community Call
Description
Rewards Distribution presentation & status update.
A
A
A
You
know
we
don't
have
as
many
seasons
in
San
Diego,
we
like
to
say
we
have
seasons
in
San
Diego,
but
it's
mostly
just
you
know,
winter
and
summer.
For
us
it
is
kind
of
spring-like,
though
there's
a
lot
of
birds
and
whatnot
out
there.
Alright,
let's
get
started,
let's
get
started
so
welcome
everyone
going
to
do
our
engineering
update
on
the
project
and
then
owner
is
going
to
talk
about
rewards
distribution
for
the
blockchain,
so
I'll
get
started.
We
cut
our
February
release.
A
We
were
previously
calling
this
no
12,
but
we
had
a
versioning
change,
so
the
version
numbers
actually
node
0.14,
because
we
released
a
few
big
things:
our
contracts,
SDK
and
our
Rus
stock,
and
we
ran
into
a
bit
of
a
technical
issue
with
Rus
stock.
So
we
had
to
break
from
versioning
so
way
to
do
a
couple
of
iterations
on
the
release,
and
so
we
had
to
up
the
version
with
each
time
because
that's
cargos
forcing
us
to
do
that.
A
So
our
February
release
is
version,
0
T
and
it
was
released
on
Friday
and
the
version
numbers
will
fluctuate
a
lot
more.
So
we'll
start
talking
about
the
releases
on
a
monthly
cadence
right,
so
our
March
release
will
be
going
out
in
the
middle
of
March
with
our
test
net
going
out
two
weeks
after
that,
and
let's
see
current
focus
so
we're
building
honest
highway
and
what
that
means
is
we're
planning
for
all
the
validators
to
be
honest,
validators
in
in
this
first
milestone,
and
this
will
get
us,
get
us
highway
up
and
running.
A
Let
us
do
some
performance
testing
hardening.
While
we
go
ahead
and
build
out
the
economics
and
security
and
all
of
the
fault
detection
that
we
will
need
for.
Prime
time,
honest
highway
is
still
Byzantine
fault.
Tolerant
is
not
attack
resistant
around
equivocation
and
those
like
of
that,
because
they
won't
actually
be
much
in
the
way
of
economic,
slashing
we're
setting
up
a
test
environment
to
get
ready
for
that
test
net
and
we're
still
doing
some
research
on
spam
protection.
This
is
one
of
the
dos
vectors
for
a
proof
of
stake
network
actually
for
all
networks.
A
You
have
to
worry
about
what
your
dose
vector
is
for
being
dosed
out
my
deploys
so
there's
a
variety
of
different
mechanisms
that
help
with
that,
and
so
we're
doing
some
research
on
that
on
consensus.
We've
mostly
got
our
arrows
with
fixed
length
length
rounds
to
complete
and
now
we're
going
to
do
an
update
around
equivocation
detection
will
be
able
to
detect
equivocation
in
in
honest
Highway.
A
We
just
won't
slash
for
it
and
then
we're
doing
the
log,
an
optimization
for
the
fourth
choice,
and
this
basically
reduces
the
number
of
messages
is
needed
in
order
to
reach
a
summit
or
agreement
on
a
block,
and
the
team
is
starting
to
test
the
consensus
protocol
locally
and
we're
building
out
big
test
beds
that
will
demo
next
week,
which
is
our
s
test
framework.
We're
gonna,
have
mark
Greenslade
come
in
here
and
talk
about
it.
I
have
a
demo
either
recorded
that
we
can
share
and
show
or
I'll
have
mark
on
here.
A
Live
will
also
be
publishing
the
S
test
specification,
which
I'll
lead
to
in
a
release.
Announcement
we'll
be
producing
that
and
publishing
that
to
github
next
week,
and
this
will
be
a
large-scale
testing
system
that
will
help
us
get
ready
for
testing
it
and,
of
course,
maintenance
till
building
the
sequoia
simulator.
We
created
a
small
simulator
to
provide
us
estimates
of
consensus
overhead.
This
is
what
really
drove
the
realization
that
we've
got
to
get
the
log
n
optimization
in
place.
A
All
the
way
out
to
the
to
the
edge
of
the
system
at
the
client
level
of
the
execution
engine
side,
we've
made
some
great
inroads
into
performance.
We're
gonna
continue
working
on
that
by
making
the
system
contracts
made
up
in
EE.
For
those
of
you,
that
aren't
aware
is
our
system
contracts,
the
proof
of
stake
and
mint
contract
are
very
generic
right
now,
so
their
contracts
that
just
get
deployed
at
Genesis.
A
You
know
one
layer
lower,
and
this
makes
the
performance
a
lot
faster
in
terms
of
you
know,
contract
execution,
if
you
think
about
it,
the
the
proof
of
State
contract
and
the
token
contract
are
involved
in
every
single
transaction,
so
optimizing.
These
gives
us
a
lot
of
performance,
we're
also
contemplating
bringing
in
the
standard
payment
code
as
a
quote/unquote
system
contract.
It
would
be
something
we
would
deploy
Genesis
and
again.
A
A
We
you'll
get
a
chance
to
see
the
testbed
environment
next
week,
so
dial
in
next
week
to
see
what
the
S
test
specification
looks
like
and
what
we're
building
this
is
super
important
to
building
a
huge
network
that
is
going
to
be
available
in
a
hostile
environment
right.
So
we
fully
expect
that
we'll
have
a
good
set
of
malicious
validators
in
the
network
and
so
testing
for
that
is
critical.
A
We're
talking
about
a
few
small
test
cases
we're
also
working
with
here,
as
well
as
optimizing
CI.
That's
work
has
been
underway
for
a
few
weeks
ecosystem.
We
had
a
request
to
put
a
lot
more
inline
documentation
in
the
smart
contracts,
with
the
release
of
the
smart
contract
development
kit.
We
are
also
building
adapt,
developer
guide
and
we're
building
a
deploy.
A
Interface
within
clarity,
I'm
also
going
to
be
working
on
a
broader
vision
for
the
clarity
block
is
for
and
coming
up
with
some
business
requirements
in
terms
of
what
we
want
to
achieve
with
it
clarity.
We
release
support
for
typescript
last
week
we
want
to
build
an
in-browser
IDE,
and
so
this
deployer
face
is
the
first
step
towards
that.
A
On
the
economic
side,
we're
figuring
out
our
pricing
model
and
we're
looking
at
on
the
research
side,
span
protection
and
we're
going
to
be
talking
and
going
through
the
reward
distribution
today
and
we've
got
a
mock-up
or
simulator
for
that
I.
Don't
know
if
we'll
be
talking
about
the
simulator,
but
we'll
be
explaining
the
rewards
distribution,
how
it
works
and
we're
looking
at
probably
doing
off-site
end
of
March
early
April
I'd
actually
like
to
have
a
team.
A
B
B
I've
been
referencing,
the
formulation
dog
I
think
I
also
talked
a
couple
of
times
about
censorship
in
the
past,
but
it
was
always
like
I
always
found.
I
could
explain
it
better.
So
today,
using
very
basic
game
theory.
I'll
explain
why
the
reward
distribution
we
desire
we
designed
for
highway
is,
you
know,
deters
oligopolies
and
actually
supports
a
small
guy
and
also.
B
Messages
so,
let's
just
explain
our
setting
violators
have
stakes:
I
will
I
will
go
through
with
a
very,
very
basic
game,
theoretic
example,
and
in
this
in
the
example
validator
a
has
state
200
tokens
B
has
taken
other
tokens
and
in
in
the
saying
violators
are
rewarded
for
proposing
and
finalizing
blocks.
We
simplify
this
to
sending
messages,
so
each
validator
has
to
sign
one
message:
if
they
send
a
message
they
receive
rewards
for
Chanel
to
their
steak.
A
B
B
B
You
also
have
to
take
costs
into
account,
while
leaders
think
our
operational
costs.
It's
not
free
to
be
a
validator.
Do
you
have
to
run
servers
so
for
the
for
the
setting
we
will
assume
be
the
minority
incurs
and
operational
costs
between
zero
and
one
so
that
their
their
payout
is
always
greater
than
one.
So
if
B
doesn't
these
sends
a
message
and
isn't
censored?
He
ends
up
with
1
minus
C.
B
B
B
If
B
decides
not
to
enter
the
game
ends
so
because
there's
no
one
to
censor
and
then
you
have
like
three
different
possibilities
with
three
three
payouts
Beach
with
for
each
validator
in
the
naive
reward
distribution.
I
talked
about
this
now.
If
someone
appears
to
have
sent
a
message
they
just
receive
their
allocated
rewards,
so
nothing
nothing
complicated.
So
if
the
simplest
one,
if
he
doesn't
enter,
we
have
total
three
rewards
and
B
doesn't
get
anything
because
we
just
get
0
because
b
is
out,
they
don't
incur
any
costs.
B
If
B
chooses
to
enter,
we
will
sure
we
will
operate
and
they
will
incur
a
minus
C.
But
if
on
top
of
that,
if
they
are
being
censored,
they
will
not
receive
their
reward,
which
is
which
is
1.
So
if
everything
goes
well
so
the
top
top
branch
is
the
good
scenario.
B
enters
and
is
not
censored
by
a
and
the
bad
one
is
the
one
on
the
bottom
we
want.
We
want
to
have
our
equilibrium
on
the
top.
In
this
case
yeah.
B
B
We
can
look
at
the
situation
here.
It'd
be
knows
that
if
he
enters
it
doesn't
make
any
difference
for
a
to
censor
him.
If
he,
if
B,
regardless
of
what
B
does
a
does
be
able
received,
so
it
we
can
conclude
that
a
can
as
well
censor
him
B.
Then
B
would
end
up
in
a
scenario
where
B's
at
a
loss.
We
wouldn't
want
to
incur
minus
C
costs,
so
you
would
eventually
decide
decide
on
not
to
enter
because
in
that
case,
B
B
is
at
zero,
which
is
much
better
than
minus
C.
B
B
B
A
B
B
We
go
one
step
forward,
it's
specified
in
the
formulation
why
we
do
it
this
way,
so
we
we
also
make
sure
with
Tim's,
we
see,
rewards
and
we
do
it
in
a
way
without
impairing
the
performance
of
the
protocol.
If
you
were
to
compare,
compare
the
version
this
this
route,
your
distribution
scheme
with
the
previous
one
in
this
game
in
the
interior
in
the
in
the
answer-
choice
B-
has
a
possibility
where
these
payoffs
is
greater
than
zero.
So,
in
this
scenario,
B
can
still
be
making
profit.
B
B
B
B
B
If
the
oligopoly
earns
other
earns
more
rewards,
but
others
would
eventually
do
it.
It's
only
a
matter
of
who's.
Gonna
do
it
and
the
players
who
understand
the
game
would
actually
compete
to
be
that
one
so
would
actually
form
coalition's
and
it's
very
trivial
actually
in
a
client
to
enter
a
blacklist
of
addresses
and
then
ignore
all
the
messages
coming
from
those
that
addresses.
Although
it
will
ship
our
clients
with
you
know,
honest
behavior
baked
in
it's
actually
very
very
trivial
to
sensory
others,
since
it's
a
very
low
hanging
fruit
for
values
of,
do
it.
B
If
they
can
prove
that
to
themselves
that
it
earns
more
rewards,
they
will
do
it
and
it
will.
It
will
actually
lock
minorities
out
of
the
protocol.
It's
actually
a
very
big
threat
to
you
know
validators
in
proof
of
stake
protocols
in
this
in
this
presentation,
I
want
to
demonstrate
why
scale
and
symmetrical
Wars,
ensure
honest
behavior
I,
didn't
explain
why
it
doesn't
sacrifice
any
performance,
because
that's
not
in
the
scope
of
this
one.
If
you're
interested
about
it,
you
can
check
out
these
two
posts.
B
If
it's
too
much
the
type,
you
can
just
go
there
and
choose,
and
corresponding
articles
I'm
also
preparing
a
new
blog
post
for
for
our
blog,
which
will
be
you
know
along
these
lines
which
you
can
read,
which
will
have
more
details,
I
think
I
think
it
will
be.
You
know
much
more
exploratory
in
true
for
our
design
choices
that
we
took
here
at
Casa
collapse
for
highway.
A
That's
terrific,
thank
you
so
much
owner
for
that
very
good
and
thorough
explanation.
Yeah
I
would
love
for
you
to
drop
those
links
in
our
discord
and
telegram.
So
folks,
can
you
know
check
them
out
and
we'll
be
probably
sharing
this
explanation
too,
as
well
on
our
community.
Just
a
snippet
of
the
community
calls.
So
thanks.
So
much
for
that.
I
really
appreciate
it.
A
I
think
next
week,
I
want
to
see
if
I
can
get
art
Greenslade
to
come
online
here
and
show
us
the
estás
form
that
is
building
which
is
going
to
be
our
big
testing
framework
to
get
us
ready
for
a
test
net
and
maenette,
so
it'll
be
basically
a
testbed
that
will
run
through
a
variety
of
different
simulations
and
tests
of
the
network.
So
awesome
presentation,
thank
you
so
much
everyone
for
dialing
in
and
we
will
talk
to
you
next
week.
Cheers
bye.