►
From YouTube: CasperLabs Community Call
Description
Rewards Distribution presentation & status update.
A
On
youtube
good
morning,
everyone
good
morning,
good
afternoon,
good
evening,
I'm
sitting
here
in
wonderful
san,
diego
having
my
cup
of
tea
my
morning
tea.
So
wherever
you
are
have
a
sip
of
your
favorite
beverage
with
me
and
we'll
talk
a
little
bit
about
blockchain
and
about
what
casper
labs
is
doing
and
what
our
status
is
in
engineering.
So
with
that,
let's
dive
right
in
I'm
going
to
share
my
screen.
A
So
we've
entered
our
third
sprint
for
those
of
you
that
haven't
dialed.
In
some
time
we
recently
moved
to
weekly
sprints
because
we
wanted
to
actually
accelerate
delivery.
We
wanted
to
do
smaller
chunks
of
work
more
frequently
and
I,
the
team
has
been
just
doing
a
fantastic
job.
Their
sprints
have
been
smaller
and
they've
been
hitting
their
commitments.
So
I'm
really
pleased
with
how
this
is
going.
Execution
has
gone
very
smoothly.
A
We
cut
node
19
release
on
may
21st
on
schedule
and
this
release
delivered
deploy
gossiping,
and
that
was
a
big
feature
in
that
release,
along
with
bug
fixes
and
when
we
deployed
it
to
the
test
net,
we
added
we
turned
on
omega
blocks.
The
omega
block
feature
was
there
before,
but
we
hadn't
enabled
it
so
we're
enabling
omega
blocks.
Alongside
with
deploy
gossiping,
the
testnet
and
I'll.
Be
writing
a
blog
post
on
medium.
A
Look
out
for
that
that
talks
about
what
omega
blocks
are,
how
they
work
with
the
highway
proofs
and
why
they're
important,
let's
see,
let's
see
we're
gonna
we're
gonna
we're
doing
more
log
scraping
for
the
validators,
because
we
discovered
it
helped
us
in
determining
what
kind
of
fixes
we
need
to
put
in.
A
So
when
we're
debugging
issues,
it's
really
helpful
for
us
to
have
the
logs
versus
trying
to
get
them
for
the
validator,
so
we're
just
doing
log
shipping
we've
got
24
consensus,
forming
nodes
in
the
network
and
three
read-only
nodes
in
the
network.
So
it's
cool
to
see
some
read-only
nodes
joined.
These
are
not
our
read-only
nodes.
They
are
community
read-only
nodes
which
is
kind
of
cool.
These
are
folks
that
want
to
join
the
test
network,
just
waiting
in
line
to
join
up.
A
We
have
a
bunch
of
bug,
fix
stability,
fixes
that
we're
going
to
deploy
as
part
of
beta.
What's
the
number
ashok,
I
think
it's
probably
at
least
like
seven
or
eight.
B
So
we
have
total
18
bucks
out
of
which
10
are
fixed.
Three
are
three
are
in
review
and
five
we
are
prioritizing
because
a
couple
of
them,
we
don't
see
them
anymore
and
two
three
of
them
are
not
as
important.
B
Triaged-
and
this
is
overall
like.
A
B
The
test
net
has
started
so
since
31st
of
march.
A
A
A
So
yeah
on
highway
we're
working
with
fixed
round
lengths
right
proposal
blocks
if
you're
the
leader
omegas,
if
you're
not
deploy
gossiping,
that's
what's
currently
running
in
testnet
rust.
We
are
so
for
those
of
you
that
are
not
aware
we're
building
a
new
node
in
rust,
we're
finalizing
the
network
component,
we're
going
from
the
ground
up
and
a
unified
storage
model.
A
Yeah,
so
we're
figuring
out,
so
one
of
the
things
I
believe
is
that
the
deploy
buffer
functionality
is
really
important
right,
because
what
you
want
is
a
nice
smooth
stream
message
queue
of
deploys
and
then
you
only
clear
the
deploy
buffer
after
the
transactions
have
been
finalized,
and
this
prevents
replay
attacks
right,
correct.
B
Yeah,
and,
and
also
since
we
are
looking
for
a
plugable
consensus
architecture
this
time,
the
deploy
buffer
that
we
used
in
the
last
like
the
scala
node,
is
different
because
it
was
kind
of
dependent
on
the
consensus.
A
A
A
A
Yeah
we
found
that
that
issue
in
clarity.
A
We
are
hardening
node,
the
19.x
release.
All
those
bug
fixes
that
we
talked
about
those
are
all
in
lrt
our
long-running
test
bed,
where
we
run
them
for
about
seven
days.
Oh,
I
need
another
exception.
I'm
feeling
slippy.
A
So
yeah
we
like
to
harden
those
fixes
for
a
minimum
of
seven
days.
I'm
pleased
that
the
current
test
net's
been
up
for
about
10
days.
We
deployed
it.
When
did
we
launch
it?
We
launched
it
on
the
23rd
on
the
23rd,
okay,
so
we're
pushing
eight
days
now,
which
is
good
by
the
beta
test
that
time
frame,
I
have
goals,
and
I
want
the
sla
for
the
network
to
be
a
minimum
of
four
weeks,
so
we
only
balance
it
when
we
need
to.
A
Good
excellent
excellent
and
we
continue
to
make
improvements
on
s-tests
we're
working
with
s-tests
to
get
it
ready
for
the
rust
node
right
when
we.
What
are
we
going
to
start
testing
the
rust
node
with
s-test
and
comms
layer?
Have
we
started
that
already.
B
No,
so
the
first
is
standing
up
the
network
and
the
commitment
that
we
have
is
a
network
of
five
rust
nodes
with
doing
basic
basic
communication,
and
that's
the
goal
that
we
have
by
end
of
by
mid-june,
like
15th
of
june.
A
Sounds
good
great
and
one
of
the
big
features
that
we're
putting
out
there,
along
with
contract
headers
in
june
release,
is
also
going
to
be
the
multi-signature
algorithm
support,
so
multi-algorithm
keys,
basically
where
we
support
both
ethereum
keys,
as
well
as
casper,
lab
keys
and
secure
enclave
keys.
This
is
to
support.
Of
course
you
know.
Hardware
wallets
and
the
like
we're,
also
helping
out
the
crypto
chick
scales,
we're
building
a
voting,
dap
smart
contract
and
web
application.
A
So
if
you're
going
to
vote
on,
crypto
chicks,
you're
attending
their
virtual
conference
you'll
be
able
to
place
your
vote
via
the
casperlas
blockchain
on
the
voting
app
to
vote
for
who
the
hackathon
winners
are,
and
it's
our
intention
later
on,
to
use
this
same
technology
for
our
own
virtual
hackathons
that
we
host
right
so
you'll
be
able
to
submit,
submit
to
who
your
entrants
are
and
then
vote
for
them.
So
community
votes
on
the
different
projects
working
on
casper
labs,
and
we
have,
of
course
the
what
information
did
you
update?
A
Right
so
our
value
proposition
has
always
been
is
that
we
want
to
stabilize
fees,
stabilize
gas
fees
and,
as
you
guys
know,
we
have
a
partnership
with
chain
links.
That's
one
avenue
we're
definitely
looking
at
for
oracle
data
to
help
express
gas
fees
in
fiat,
but
then
there's
also
the
matter
of
stabilizing
the
gas
fees
right.
So
two
things
one.
You
want
your
transaction
fees
to
be
in
fiat
because
it's
easy
for
businesses
to
budget.
C
Actually,
the
the
proposal
I'm
going
to
do
is
like
a
complete
alternative
to
using
chain
link
or
oracles
in
a
trustless
way.
So
yeah
you'll
hear
more
about
it.
Yeah.
A
Yep
we're
also,
we
continue
to
push
forward
with
the
integration
with
chain
link,
so
we
can
continue
prototyping
and
doing
some
work
there
and
then
we're
gonna.
We've
also
been
working
on
the
token
vintage
proposal
is
basically
the
benefits
you
get
from
holding
your
token
for
a
long
time
as
a
validator
right,
it's
only
for
validators.
It
basically
is.
If
you
stake
your
token
for
a
long
time,
you
get
some
tokens
token
vintages
correct.
A
So
we
want
to
provide
good
incentives
for
folks
to
to
hold
the
token
and
weekly
workshops
are
friday.
Oh
friday,
7
a.m,
pacific
and
friday
12
a.m,
pacific
4
p.m.
Japan,
time
on
saturday,
no
on
friday.
A
Got
it
got
it
okay,
so
we
moved
into
friday
mornings
great
and
the
zoom
information
is
here
and
joe
we'll
want
to
push
out
those.
This
zoom
information
for
the
weekly
workshops
to
help
with
smart,
smart
contract
development
so
make
sure
folks
know
about
it.
For
me,
please
and
there's
also
a
calendar
link
that
we
can
push
out
there,
which
is
the
release,
count
the
release
calendar
and
we
can
throw
events
on
there
too.
A
A
C
C
That
means
like
it's:
it's
independent
of
the
the
gas
price
being
in
fiat
denominated
in
fiat,
so
it
adapts
to
market
conditions.
So
I
mean,
if
the
token
price
increases
or
decreases.
You
would
also
expect
the
gas
price
to
adapt,
but
then
the
yeah
it
will
be
dynamic.
C
I
just
want
to
introduce
the
proposal
by
talking
a
little
bit
about
control
mechanisms
in
blockchains
you
so
one
one
example
of
this
is
a
bitcoin
mining
difficulty,
bitcoin
mining.
The
difficulty
is
a
protocol
parameter
that
needs
to
adapt
to
certain
conditions,
and
you
also
don't
want
it
to
be
controlled
by
a
specific
participant
because
difficulty
directly
controls
how
long
it
takes
to
mine
a
block
and,
as
there
is
more
hash
like
as
the
hash
rate
increases,
you
want
the
difficulty
to
increase
right.
So
there
is
satoshi.
C
Nakamoto
isn't
doing
that
manually,
he's
not
submitting
a
special
transaction
every
like
few
days
adjusting
the
bitcoin
difficulty.
Instead,
it's
dumb
by
using
on-chain
data,
which
the
network
has
already
reached
consensus
upon.
So
using
this
one
chain
data
we
can
implement
a.
We
can
basically
use
control
theory
to
have
this
parameter
auto-adjust
in
time
and
like
have
a
very
simple
and
robust
system.
C
C
So
talking
about
block
difficulty,
it's
it's
a
it's
a
like
a
parameter
that
changes
in
time
you.
You
also
have
a
quantity
on
blockchains,
which
is
the
transaction
fee
or
the
gas
price,
and
that
changes
throughout
the
day.
So
the
range
of
volatility
is
even
shorter,
whereas
difficulty
changes
like
maybe
week
to
week
this
gas
price
shifts
like
and
in
this
example.
C
It
changes
from
six
to
22
in
a
single
day.
So
the
range
is
very
small
and
the
volatility
is
much
higher,
but
maybe
there
were
a
way
to
use
a
feedback.
Loop
to
you
know,
govern
govern
this
automatically
and
have
the
same
benefits
of
stability
so
yeah.
These
figures
are
from
a
previous
blog
post.
I
did
so
this.
This
is
a
from
last
year.
In
march,
I
I
took
this
window.
C
Very
characteristic,
if
you
look
at
the
peaks
and
troughs
of
the
cycles,
you
would
see
that
they
all
look
similar
and
that's
for
a
reason.
C
That's
the
world
rotating
and
people
waking
up
and
sending
transactions,
so
china
has
the
crazy
largest
demand
and
china
is
followed
by
us
and
europe.
So
that
is
what
you
would
see
in
a
day.
So
it
is
a
feature
of
the
market,
I'm
giving
this
as
a
context,
the
background
information
to
understand
the
problem
better.
C
In
any
case,
if
a
blockchain
is
popular,
it
will
hit
capacity
and
when
it
hits
capacity
it
will
experience
this
kind
of
volatility
every
day
and
so
the
I
just.
If
you
want
to
see
how
much
price
changes
during
the
day
like
the
minimum
price
and
maximum
price,
it
can
get
really
high.
But
you
see
that
on
average
it
can
get
up
to
like
four
times.
C
There
is
a
there,
isn't
any
efficiency
and
there
is
a
proposal
in
ethereum
to
solve
that
eip1559.
C
So
what
it
does
is
it
implements
a
control
mechanism
similar
to
a
difficulty
adjustment,
but
in
this
case
you
set
as
your
process
variable.
This
is
a
control
theory
term.
You
look
at
your
block.
Guess
you
usage
instead
of
block
time,
it's
yet
it's
another
quantity
that
is
on
chain.
That
is
like
the
yeah.
That
is
a
the
network,
has
already
reached
consensus
upon,
so
the
protocol
can
safely
use
it.
C
C
If
the
block
guess
usage
decreases,
you
decrease
the
price
to
move,
to
move
like
parallel,
and
the
factor
is
one
over
eight.
So,
like
you
change
approximately
twelve
point,
five
percent
in
each
block,
so
this
solves
the
efficiency
problem
at
a
at
a
given
time.
Users
don't
know
the
like
correct
price.
C
C
C
C
Like
reduce
the
incentives
to
manipulate
the
price
for
the
miners,
but
this
is
not
a
focus
of
this
presentation
so,
like
I
already
said
this
mitigates
the
efficiency
problem,
but
what
we
want
in
casper
labs
is
stable
gas
price
throughout
the
day,
whereas
with
eip
1559
it
can
change
12
per
block,
so
it
has.
It
can
move
by
12
percent
every
13
seconds
14
seconds.
C
C
C
C
So
maybe
if
you
have,
if
you
increase
your
round
length,
maybe
you
can
also
increase
your
block
gas
limit
and
maybe
have
it
set
in
the
protocol.
That's
why
I
use
this
fullness
value
to
be.
You
know
independent
of
that
context,
but
it's
sufficient
for
the
simulations
I'm
going
to
present
just
now,
so
the
rate
of
change
would
then
be
also
low
daily
or,
like
I'm
thinking
like
one
to
five
percent.
This
is
just
intuition
and
differently
than
eip1559
target
fullness
will
be
selected,
so
that
demanded
throughput
equals
blockchain's
capacity
at
its
site.
C
What
that
means
is
you
use
habit
cycle
right?
This
is
the
high
point
you
you
target
a
fullness
so
that,
like
the
aggregate
fullness
of
blocks
in
the
whole
day
like
when
you
target
a
certain
value,
for
example,
I
I
target
65
percent.
C
It
means
that
with
the
set
value,
this
is
the
point
like
the
tip
where
congestion
might
happen,
but
it
doesn't
it's
the
it's
the
limit,
so
we
we
target
so
that
35
percent
of
the
blocks
are
empty
and
65
percent
are
full
yeah.
I'm
gonna,
I'm
gonna,
introduce
the
parameters
and
from
now
on,
it's
just
graphs
and
figures.
So
don't
worry.
C
Let
me
let
me
introduce
some
simulations
to
make
it
more
concrete,
so
I
made
up
a
demand
curve
and
treated
as
a
cumulative
distribution
function.
I
sampled
the
resulting
distribution,
so
you
see
the
results.
The
red
line
is
these
actual
demand
curve
and
blue
blue
bars
are
the
histogram
of
the
sample
prices.
C
I
scale
it
to
show
that
it's
the
exact
shape
and
I
so
sampling
means,
like
I
I
sample
it
for
some
number
of
users
and
how
I
simulate
the
daily
demand
cycle
is
I
draw
like,
like
at
the
height
I
drove
7000
users
from
the
from
a
distribution
and
at
the
low
point
I
sample
3000
users
and
it's
the
it's
a
sine
wave.
C
This
is
just
to
and
the
it
repeats
daily.
So
this
is
an
analog
of
the
shape.
You
have
seen
it's
not
as
it's
not
as
distorted
as
this
one,
but
it's
still
sufficient
for
the
demo.
We're
doing.
C
I
I
included
this
explanation
here
for
reference:
I'm
not
gonna
go
over
it
in
a
lot
of
detail.
This
is
basically
ether.
So
basically,
this
agent
behavior
is
supposed
to
emulate
ethereum
users.
As
of
now-
and
you
will
see
you
see
the
gas
price
as
a
result
of
this.
So
with
this
demand
curve,
and
with
this
number
of
users
per
block
and
with
some
parameters,
you
would
see
the
gas
price
evolve
and
follow,
follow
the
follow
the
you
know,
the
demand
curve.
You
see
here
the
demand
versus
time.
C
In
fact,
the
reason
for
this
increase-
and
then
you
know
changing
into
the
sine
wave-
is
that
I
I
I
let
the
market
price
start
at
zero.
So
this
is
like
the
genesis
and
from
genesis.
Users
keep
overbidding
their
price
prices,
users
who
can
overbid
their
prices,
increase
it
until
you
know
the
market,
like
the
number
of
supply,
demanded,
sorry
supply
and
the
num
the
number
of
demand
gets
equal,
so
this
is
like
when
it
reaches
equilibrium
after
it
reaches
equilibrium.
C
It
starts
following
the
trend
you
see
in
the
previous
previous
slide
and
I
I
define
agent
behavior
so
that
they
an
agent,
looks
at
the
previous
block
and
if
they
can't
afford
the
minimum
price
in
in
that
block,
they
don't
send
a
message
like
so
if
you,
it
eventually
converges
down
to
a
value.
So
in
the
simulation
I
used
like
ethereum's
current
parameters,
10
million
gas
and
21
million
transaction
spent
the
gas
spent
by
a
transaction,
and
that
amounts
to
like
476
transactions
per
block.
C
And
you
see
the
like
simulation
converge
that
value
here.
This
will.
This
will
be
more
more
clear,
hopefully
so,
instead
of
introducing
the
value
so
this
these
are.
Some
of
these
are
the
same
values
I
used.
So,
let's
make,
let's
make
it
clear
more
so
here
I
use
like
10
million
gas
limit,
so
that's
how
much
gas
that
can
exist
in
the
block
and
like
201
million
gases
spent
by
a
transaction.
C
So
I
I
introduced
what
happens
with
the
floating
price
I
I
now
want.
I
want
to
show
you
what
happens
with
price
adjustment,
so
in
this
case
I
use
block
time
10
minutes
because
it
took
a
lot
of
time.
I'm
I'm!
C
This
is
a
huge
simulation
and
I'm
doing
it
with
python.
So,
instead
of
eight
seconds
or
or
like
13
seconds,
I
did
it
with
600
like
10
minutes,
and
as
I
told
I,
I
look
at
the
previous
day
for
the
fullness
that
I
target
and
I
let
the
price
change
one
percent
per
day.
C
So
that
means,
if
a
if
the
fullness
of
the
last
day
is
greater
than
0.65,
I
increased
the
price
one
percent.
If
the
fullness
of
the
previous
day
is
lower,
I
decrease
it.
One
percent.
C
So
I
just
I
repeat
the
same
thing
I
did
with
floating
price
case
but
for
for
40
days,
because
it's
the
rate
of
change
of
the
price
is
lower
in
this
case,
and
I
want
to
see
it
converge
to
a
value
and,
let's
remember
the
market
price
in
the
floating
price
floating
price
case.
C
B
C
C
And
you
can
see
that
there
is
a
there's,
a
there's,
a
difference
here
right
in
the
first
10
15
days,
it's
underpriced,
but
it
takes
some
time
for
it
to
reach
the
actual.
You
know,
market
price
we
desire
and
once
it
reaches
it,
stays
stable
in
this
first
period.
You'll
see
here
that,
like
some
blocks,
are
very
full
there
for,
for
the
height
of
the
day,
the
blocks
are
completely
full
and
this
is
actually
when
the
congestions
happen,
but
once
once
the
price
table
stabilizes
to
the
price
we
want.
C
You
see
that
the
maximum
fullness
in
a
day
does
not
exceed
one
like
it
exceeds
in
a
couple
of
places,
but
not
like
here,
and
you
would
see
like
the
fullness
of
that
day.
It
starts
with
one.
C
Then
it
decreases
to
the
value
and
oscillates
around
it
as
it
converges,
and
I
just
want
to
make
the
point.
So
I
said
that
there
are
congestion
here.
That
means
tran.
The
size
of
the
transaction
pool
is
increasing,
like
transactions,
don't
get
included
and
the
memory
the
transaction
pool,
increases
and
increases.
C
C
And,
of
course,
this
is
a
daily
simulation
so
and
the
it's
the
same
trend
over
and
over.
But
what
happened
if
there
was
a
long-term
increase
in
demand
like
in
a
few
months.
Let's
say
our
toy
blockchain
received
more
users
in
general.
So
what
would
happen
then
you'll
see
see
this
go
up
and
that's
what
I
wanted
to
simulate
in
this
case.
C
So,
instead
of
the
previous
range,
you
still
have
like
the
the
range
of
the
simulation
has
increased
before
it
was
40
days.
Now
it's
200
something
days,
and
then
you
see
this
increasing
demand,
and
you
see
all
this.
All
these
lines
are
the
daily
cycles
that
I
introduce
in
at
the
beginning,
and
I
just
wanna-
I
I
show
you
all
the
all
the
different
quantities
in
the
previous
example,
so
I
just
want
to
cut
cut
short
to
the
price,
but
this
is
what
price
looks
so
normally
it
should
converge
to.
B
C
I
I
make
it
intentionally
started
at
a
low
value
like
before
I
make
it
start
at
35
and
at
the
beginning
you
don't
see
a
lot
of
increase,
so
it
started
at
35
and
converges
to
40
around
40.,
but
then
keeps
increasing
per
like
follows.
The
like
long-term
demand
keeps
increasing
and
it
follows
direct
like
directly
the
trend
you
see
here,
and
this
is
what
we
want
and
once
it
subsides
once,
the
most
of
the
demand
is.
B
C
C
You,
don't
you
don't
experience
any
congestions
with
this
with
this
given
demand
profile,
I
hope
I
didn't
take
too
long.
A
No
you're
fine.
Can
you
hear
me.
C
Let
me
just
give
you
the
just
a
fit.
A
A
C
C
C
C
So
this
has
been
proven
by
bitcoins.
C
Like
robustness
and
I
believe
we
can
reproduce
the
same
robustness
with
the
fee
market,
if
you
want
to
learn
more
about
this,
you
can
read
this
blog
post.
You
can
get
it
from
the
presentation.
You
can
also
check
out
the
blog
post
I
put
here
and
if
there
are
any
questions
you
know.
B
A
A
B
C
A
How
about
now,
can
you
hear
me
yeah?
Okay,
much
better,
I
think.
Maybe
your
volume
was
turned
down.
This
is
terrific
yeah.
I
think
I'd
like
to
definitely
get
this
presentation
up
on
slideshare.net
and,
of
course
we
have
the
recording
where
you
walk
through
everything.
It
would
be
nice
to
not
have
so
many
spikes
in
between.
So
if
we
can
do
something
with
the
algorithm
to
smooth
out
the
curve,
I
mean,
if
we
can
tell
businesses
look,
this
is
your
maximum
risk
for
transaction
fees?
I
think
that
would
be
great.
A
Do
you
also
talk
here
about
omega
blocks,
because
if
we
enable
omega
blocks
literally,
any
validator
can
propose
a
block,
there's
still,
of
course,
always
a
question
of
network
capacity
and
the
number
of
transactions
the
network
broadly
can
process,
but.
A
C
And
further
like
we,
we
are
thinking
of
using
a
similar
mechanism
to
control
the
validator
slot
count
in
our
validator
auctions,
bonding
auctions,
so
the
number
of
validators
looking
at
so
this
time.
The
process
variable
here
would
be
the
finalization
percentage,
how
many
blocks
got
finalized
in
on
time.
A
Right,
yep
yeah,
so
definitely
for
those
of
you
that
don't
know
anything
about
highway
or
know
little
about
highway.
One
of
the
things
that
we
can
do
with
this
protocol
is
and
the
reason
we've
got
bonding
auctions
in
places
the
we.
What
we
want
to
do
is
have
a
consistent
time
to
finalization
and,
depending
on
network
conditions,
transaction
load,
we
can
adjust
the
validator
set
size
dynamically.
A
We
fully
expect
that
we'll
be
able
to
decentralize
the
network
more
and
more,
because
cpus
will
become
more
efficient
networks
will
become
more
efficient,
and
so
the
consensus
overhead
associated
with
this
system
will
go
down
and
down
in
in
the
sense
that
you'll
still
have
a
certain
number
of
messages
per
block,
and
you
still
will
have
to
you
know
gossip
blocks,
but
the
system
will
be
more
efficient
at
doing
so,
and
so
we
will
be
able
to
increase
the
validators
that
size
and
bring
more
security
over
time.
C
A
Yeah,
the
basically,
what
you
put
a
pin
in
is
what
is
the
acceptable
time
latency
to
finalization.
That's
what
the
validator
set
can
decide
right.
They
should
be
able
to
decide
what
kind
of
latency
to
finalization
they
want
to
support.
A
So,
while
in
ethereum,
you
have
this
notion
of
the
block
gas
limit,
that's
one
of
the
parameters,
but
the
other
parameter
is
how
long
is
the
round
length,
because
the
finality
comes
with
the
round
length
right,
and
so
the
system
auto
adjusts
round
lengths
you're
in
essence,
auto
adjusting
your
time
to
finalization.
Basically,.
A
And
yeah
it's
coming
together
nicely.
I'm
I'm
pretty
pleased
with
the
way
highway.
Is
you
know
evolving,
because
it's
it's
really
kind
of
close.
It's
close-ish
to
what
vlad
wanted
to
see
right
so
yeah,
it's
great!
I
mean
I
don't
like
that.
We
have
in
protocol
finalization.
A
That
was
something
that
he
felt
pretty
strongly
about,
but
I
recognize
why
we
need
to
have
in
protocol
finalization,
so
there's
a
few
things
that
we
had
to
step
away
from
from
the
pure
cbc
casper
protocol
paper,
but
we
did
so
to
make
the
protocol
more
secure
and
actually
you
know,
implementable,
for
lack
of
a
better
term.
So
great
great
presentation
do
send
me
the
raw
presentation
send
it
to
ashok
we're
going
to
get
that
up
on
slideshare.net.