►
From YouTube: CasperLabs Community Call
Description
Rewards Distribution presentation & status update.
B
A
All
right,
good
morning,
good
afternoon
good
evening,
wherever
you're
listening
to
us
from
my
metabolic,
are
the
scatino
of
castro
labs,
and
I
am
here
to
present
our
development
and
engineering,
update,
project
status,
update
and
own
or
is
gonna
talk
today
about
bonding
auctions.
Bonding
auctions
are
going
to
be
the
mechanism
by
which
validators
can
bid
for
seats
in
the
validator
set
and
we'll
talk
in
more
detail
about
that.
After
the
update,
so
I'll
get
started
with
the
update
right
away,
I
want
to
ask
everybody
else
to
go
in
mute,
so
there's
no
background
noise.
A
A
And
we
are
working
on
getting
that
tested
in
LRT
in
the
long
run
tests,
and
once
that
clears
our
testing
framework,
we
will
formally
cut
the
release
and
announce
it
and
then
launch
the
beta
phase
of
the
test.
That
and
I'm
working
on
the
beta
program.
Right
now,
and
primarily
like
big
big
things,
is
that
the
prize,
the
prize
set
is
double
as
well
as
the
validator
set.
A
The
validator
set
is
gonna,
be
twice
as
large,
we're
gonna
double
up
to
about
50
validators,
currently
we're
running
with
about
25
validators,
and
so
the
plans
for
LRT
testing
is
to
test
with
double
the
validator
set
size,
which
is
50.
We're
also
intend
on
runnin
a
performance
test.
Ideally,
what
I'd
like
to
do
is
have
the
validators
rather
performance
test
right.
A
So
we
want
to
modify
the
s
test
framework
such
that
the
validators
can
point
s
tests
at
their
node
and
run
the
performance
tests
against
the
system,
and
then
we
can
see
if
it
falls
over.
The
validators
are
quite
eager
to
make
the
network
fall
over
because
it's
been
very
stable,
so
the
nodes
have
been
running
just
you
know,
running
along
processing,
whatever
transactions
are
brought
to
them
on
the
on
clarity
and
by
doctor
Bela
president
Lee,
it's
probably
not
that
much,
but
the
system
has
been
pretty
stable,
which
makes
us
feel
good.
A
Let's
see
so
test,
that's
been
running
for
about
40
days
without
interruption
and
we've
got
you
know,
Omega
blocks
running,
and
that
means
you
can
get
a
new
block
of
the
network
about
every
20
seconds
and
finality
takes
about
8
to
10
minutes,
and
that
just
means
after
the
round
is
done
because
the
round
late,
that's
currently
fixed
in
about
8
minutes.
A
Ok,
so
moving
along
some
of
the
big
things
that
are
in
the
0.20
release
our
contract
headers
contract
headers.
We
had
a
workshop
on
that
last
week,
where
ma
che
walked
through
how
to
how
to
use
contract
headers,
how
to
implement
types
basically
named
arguments.
So
these
are.
These
are
new
features
that
come
with
contract
headers.
Also
the
upgrade
capability
right.
So
you
can
specify
contract
package
at
a
hash
and
then
use
that
for.
A
Use
that
basically
to
keep
multiple
versions
of
your
contract
on
the
system,
and
so
that
is
going
to
be
part
of
the
dot
to
will
release
it
available
in
the
test
net
once
that
goes,
live
those
folks
that
are
interested
in
getting
access
to
that
right.
Now
you
can
for
car
repo
and
it's
one
of
the
branches
there
and
it's
called
contract
errors.
We
also
will
be
supporting
having
support
for
aetherium
keys
very
excited
about
this.
A
We
said
that
we
would
want
multiple
key
support:
I'm
getting
clarity
if
it's
just
e
theorem
keys
or
if
it's
also
the
SEC
P
256
r1
keys,
which
are
believed
supported
by
secure
Enclave
and
believe
all
those
keys
will
be
used
for
both
bonding
as
well
as
accounts
and
transactions
signing.
So
it's
a
first-class
account
right
using
those
different
encryption,
elliptical
curves,
it's
very
exciting.
It
enables
a
lot
of
really
neat
features,
a
neat
functionality
to
be
supported
on
the
caspere
chain,
and
then
we
brought
the
token
transfer
function
host
side.
A
We
feel
that
that's
probably
one
of
the
biggest
use
cases
for
blockchain
is
token
transfer,
so
it
made
a
lot
of
sense
to
bring
back
contract
host
side.
It's
a
very
standard
contract
right,
very
standard
piece
of
functionality,
transferring
token
from
one
purse
to
another
so
very
exciting.
There
I'm
also
pleased
to
report
that
we
are
investigating
using
the
sohliang
compiler
to
potentially
support
solidity,
so
supporting
solidity
contracts,
pretty
cool
thing
that
we'll
be
able
to
support
on
the
caspere
last
blockchain.
A
We
hope
to
demo
that
in
about
six
weeks,
so
some
good
stuff
here
that
we're
working
on
so
Curt
focus
is
we're
working
on
the
rust
node.
The
team
has
fully
focused
on
the
rust
work
for
the
next
several
months.
Getting
a
clean,
full
rust
implementation
is
our
is
our
focus.
We
want
to
launch
the
Delta
test
that,
with
this
rust,
node
contract
run.
Time
is
mostly
feature
complete
right
after
after
the
contract,
header
is
implementation
and
the
multi
key
support.
We're
gonna
put
it
on
the
shelf
for
a
bit.
A
A
We're
going
to
be
exposing
the
Kasper
lab,
signer,
the
Python
client
and
the
all
of
our
example
contracts
as
public
repositories
we're
going
to
license
it
under
Apache
200.
The
same
thing
goes
for
the
clarity
block
is
for
I'm.
Sorry,
if
there's
some
noise
in
the
background
here,
clarity,
block
x4,
we're
gonna,
make
that
all
open-source
and
license
in
order
Patchi
to
do.
My
intention
is
to
do
that
before
the
beta
to
test
that
timeframe.
This
will
let
people
fork
it
freely.
People
can
still
fork
our
code
base
using
the
Casper
labs
open-source
license.
A
We
only
want
to
limit
our
source
code
for
the
use
of
private
and
enterprise
block
chains
without
being
explicit
license
for
because
we
want
to
support
the
spirit
of
open
source
software
and
open
block
chains,
permissionless
block
chains.
So
that's
a
little
NIT
about
our
license.
There
yeah,
and
so
here
we
talk
about
the
smart
contract
DSL.
We
are
building
a
DSL
that
will
make
authoring
of
contracts
a
lot
easier
and
much
more
natural.
A
We
receive
this
feedback,
one
from
one
of
the
developers
working
with
our
system
and
then,
of
course,
the
POC
using
so
laying
for
solidity
support
the
Casper
labs
signer
extension.
We
receive
some
feedback
right
now.
You
have
to
copy
paste
the
private
keys
in
there.
I
don't
want
to
do
that.
I
want
to
just
point
it
to
a
file
and
load
the
file
up.
It
should
know
where
the
dot
key
files
are
on
the
system
and
it
should
just
be
able
to
load
those
up
or
you
should
be
specify
path.
A
We
noticed
extreme
difficulty
for
users
trying
to
get
that
get
to
those
private
keys
rightfully
so
it
shouldn't
have
to
manage
private
keys
and
the
POC
using
so
Lange
goes
with
the
initial
work
for
the
ERC
20
contract
in
solidity,
and
we
are
doing
a
prototype
with
IP.
We
we're
very
excited
about
that.
We're
doing
some
research
on
their
private
chain
implementation
and
we're
gonna
see
what
some
of
the
things
are,
that
that
hyper
ledger
has
that
we
don't
so
we
can
go
and
build
that
for
private
and
enterprise.
A
Blockchain
implementations
we're
working
on
bonding
auctions
so
guys
Alexander
I
saw
you
some
notes.
Okay,
so
oh
no
you're
talking
about
the
ideal
number
of
validators
you're
talking
about
how
we're
going
to
change
and
adjust
the
validator
cell
okay
and
then
next
week,
alex
plans
are
talking
about
bonding
auctions
exactly.
B
A
Sounds
good
thanks
and
then
what
we
want
to
do
is
we
want
to
get
some
future
topics
to
Joe
and
let's
collaborate
with
Joe
I'd
like
for
you
guys
to
collaborate
with
Joe.
So
we
he
has
a
list
of
topics
and
we
could
start
presenting
those
to
the
community,
so
the
community
can
vote
on
what
they'd
like
to
hear
about.
We
should
also
in
Joe
we
can
probably
open
it
up
to,
for
the
community
serve
a
little
all
over
the
place.
This
morning.
A
My
apologies,
so
economics
research,
we're
talking
about
the
design
of
the
bonding
auctions
and
then
specification
for
gas
and
metering.
How
are
we
doing
on
gas
and
metering?
I
thought
we
were
pretty
close,
we're
pretty
we'd,
pretty
much
know
what
want
to
do
right
now,
we're
just
writing
up
the
engineering
specs
yeah
yeah.
B
A
Sounds
fabulous.
We
do
our
contract
engineering
workshops
twice
a
week.
The
zoom
meeting
is
located
here
and
then,
of
course,
our
community
calls
are
Tuesday
mornings,
9:00
a.m.
Pacific
time,
and
we
live
stream
on
YouTube
and
you
can
always
listen
to
the
recording
later.
If
you
wish,
Malika
has
also
presented
a
new
revised
governance
model,
so
I'd
like
to
update
the
proposed
governance
model
that
we
have
set
up
here
on
the
wiki
Malika
on
the
github.
So
I'll
probably
have
you
update
that
as
well
guys?
A
We
should
all
review
her
plans
for
governance,
so
get
machi
to
review
it
too,
and
maybe
the
guys
on
the
dev
Dao
have
them,
take
a
look
at
it
and
give
some
feedback.
So
we've
got
more
people
that
can
provide
some
feedback
on
the
governance
model
Malika
as
well,
so
I'll
I'll
shoot
that
out.
I'll
shoot
out
the
gun,
the
word
doc
later
today
to
them
that's
it
did
I
miss
anything.
A
B
B
The
problem
we
have
is,
how
are
we
gonna
determine
it,
and
how
are
we
going
to
update
this
value
in
time
when
network
conditions
change
or
there
are
some
other
changes?
It's
the
platform.
So
in
this
present
presentation,
I
will
talk
about
that.
So
the
problem
we
have
is
basically
in
computer
science.
What
is
called
communication
complexity?
B
B
If
you
don't
worry
about
energy
consumption
or
basically
other
environmental
factors,
because
the
more
hash
power
you
got,
the
higher
the
security
of
the
chain,
the
harder
harder
it
is
to
attack
a
double
span,
but
in
proof
of
state
you
have
a
big
difference,
so,
whereas
it
was
so
it
reached
consensus
on
a
block
or
on
a
single
data
value.
You
just
need
to
send
one
message
and
that
message
is
the
block
itself.
So
what
happens
is
one
of
the
miners
minor
block
that
gets
propagated
and
that's
it?
It's
it's
finalized.
B
B
Each
validator
needs
to
send
and
messages
for
each
other
validator.
That
square
is
the
number
of
messages
that
needs
to
be
sent
for
that.
We
and
that's
a
basic
mathematical
pattern
that
shows
up
in
this
specific
problem.
So
we
are
limited
by
this.
By
this
impracticality,
I
mean
the
result.
Is
proof
of
stake?
Is
drastically
different
and
proof
of
work
if
you
are
considering
a
number
of
evaluators,
because
more
evaluators
leads
to
more
messages.
So
if
you
need
to
send
more
messages,
then
you
need
longer
rounds
to
propagate
them.
B
Then
the
finalization
speed
is
decreased
and
also
possibly
throughput
is
also
decreased.
If
you
are
just
sending
one
block
three
round,
this
is
this
is
different
than
proof
work.
You
have
a
trade-off
between
decentralization
and
protocol
performance.
So
then
you
need
to
ask
the
questions.
There
is
an
ideal
number
of
flower
leaders
right.
It's
not
infinity.
It
should
be
somewhere
some
as
high
as
possible
without
sacrificing
the
user
experience.
User
should
not
keep
to
have
keep
to
be
waiting.
B
Okay,
I
kept
waiting
longer
and
longer
times
for
the
transactions
as
more
evaluators
enter
the
market
right.
So
you
need
to
ask
the
questions:
who
should
determine
determinate?
Who
should
update
it
and
how
should
it
be
updated?
Should
it
be?
You
know
a
protocol
parameter.
Should
it
be
hard
fought,
should
it
be
in
a
smart
contract
or
in
the
chain
spec?
So
these
are
the
questions.
B
One
of
the
proposals
that
I
that
I
made
last
month
is
to
use
these
control
mechanisms
which
are
which
is
used
to
govern,
govern
other
parameters
in
the
protocol.
For
example,
in
Bitcoin
you
have
a
similar
network
parameter
difficulty,
it
is,
it
is
actually
used
to
you
know
as
more
as
the
hash
power
increases,
your
blocks,
the
frequency
of
mining
block
link
increases,
and
you
need
to
pull
that
back
to
the
baseline
value.
If
a
block
is
being
mined
every
five
minutes,
you
need
to
correct
that
right.
B
You
need
to
increase
the
difficulty
so
that
it's
it's
back
to
ten
minutes
which
is
bitcoins
baseline.
So
you
use
these
control
mechanisms
in
block
chains
because
with
with
them,
you
don't
need
to
decide
if
you
party,
basically
deciding
for
the
for
its
value,
it
is
derived
from
an
on
chain
variable
or
data,
so
a
quantity
that
the
network
has
already
reached
consensus
upon.
So
why
not
just
use
it
to
govern
the
network?
That's
the
main
idea,
and
if
you
listen
in
the
previous
week's
I,
already
described
the
mechanism
to
adjust
the
price
price.
B
A
gas
price
in
the
protocol,
and
here
I
will
make
use
of
a
similar
mechanism
to
explain
how
it
could
be
used
for
changing
the
number
of
maximum
elevators
so
going
right
into
it.
So
I
I
called
validator
seats,
so
the
number
of
seats
is
the
maximum
number
of
violators
that
can
exist
in
the
protocol
and
similar
to
your
difficulty
adjustment.
We
planned
a
feedback
loop.
Also.
B
B
This
scheme,
which
we
ensure
limas
realms,
can
increase
their
their
length
if
the
network
is
having
having
difficulty
finalizing
blocks,
the
round
length
can
be
a
seconds
16
seconds,
32
seconds,
etc.
It's
actually
multiples
of
milliseconds.
So
it's
not
exactly
eight
seconds
but
8000
something
milliseconds.
B
So
with
this,
with
this
future
of
highway
at
hand,
we
can
make
use
of
it
right.
We
can
target
an
optimum
length
which,
in
networks,
normally
operated
operate
at
so
the
same
network
should
produce
finalized
around
every
32
seconds,
and,
let's
just
say,
for
now,
this
shorter
rounds
are
not
allowed.
It's
not
it's
not
defined
yet,
but
for
the
sake
of
explanation,
let's
say
shorter
rounds
are
not
allowed
in
this
specific
example,
then
we
calculate
the
number
of
blocks.
So
this
is
a
again
a
global
global
value.
B
You
calculate
the
number
of
blocks
that
will
be
proposed
if
everyone
had
the
optimal
round
length
in
an
error.
So
I
didn't
write
it
here,
but
our
protocol
time
is
basically
divided
into
eros,
which
is
which
contains
some
number
of
rounds.
We
haven't
exactly
pinpointed
how
long
an
error
will
be
so
by
using
the
length
of
the
error,
we
calculate
a
number
of
blocks
that
we've
proposed
and
rounds
that
will
be
finalized.
B
We
can
increase
it
by
a
fixed
value,
depending
on
how
many
of
the
rounds
were
finalized
or
blocks
are
proposed
in
the
previous
previous
error.
So
if
all
rounds
were
finalized,
that
means
the
protocol
is
doing
well
right.
The
current
parameter
is
comfortable.
Well,
they
values
are
not
running
into
an
a
limiting
case
where,
where
the
communication
complexity
is
posing
a
problem
causing
realms
not
to
be
finalized,
so
if
it
is
greater
than
95%,
we
increased
it
by
a
fixed
amount.
B
B
B
Like
apocalyptic
events,
imagine
imagine
Mount
God's
hack,
sorry,
it
was
taken
down
right.
The
Bitcoin
exchange
like
I,
think
in
2013-14,
and
that
was
that
caused
a
big
price
drop
that
caused
the
decrease
in
hash
rate,
because
no
more
revenue
for
miners,
but
it
Bitcoin
lived
on
right
and
that's
the
kind
of
robustness
I'm
talking
about.
There
should
probably
be
a
minimum
number
of
seats
as
a
protocol
around
here,
since
we
don't
want
to
draw
drop
below
a
certain
value.
B
B
B
B
Is
too
low?
Imagine
like,
if
you
can,
if
the
protocol
allows
the
number
of
leaders
to
drop
below
ten
or
something,
maybe
maybe
the
token,
maybe
the
an
attacker
attacking
cartel
will
be
able
to
buy
up
bulk
like
a
lot
of
tokens
and
submit
diamond
bonding
our
options
and
like
escalate.
This
sort
of
attack
like
in
in
successive
eras,
the
network
they
would
progressively
work
out
other
evaluators
from
the
network.
So
this
is
a
viable
scenario.
B
If,
if
someone
decides
to
implement
that,
it
all
depends
on
certain
parameters
right
how
do
validators,
how
now
they
just
entered
the
scene
in
the
first
place?
How
bonding
auctions
can
happen?
This
gatekeeping
attack,
but
attack
is
also
valid
not
only
for
this
mechanism,
but
a
cartel
could
also.
You
know
the
more
the
more
basically
say
you
have
the
more
blocks
you
can
produce
right
generally,
basic.
B
A
We
give
up
our
rewards
and
that
round,
yeah
Network,
the
bonding
auction
happens
and
a
bunch
of
validator
slots
are
eliminated.
Now
in
the
next
round,
I
go
ahead
and
I
finalize
a
bunch
of
blocks.
Do
I
get
substantially
more
rewards
enough,
such
that
forgoing
the
rewards
in
the
previous
era
is
worthwhile.
That's.
B
Substantial
here
I
mean,
if
you're
already
a
cartel.
That
means
you
have
at
least
33%
stake,
I,
think
and
then,
by
by
briefing
that
way,
you
would
increase
your
revenue
from
from
rewards
by
a
factor
of
three,
and
maybe
maybe
you
the
cartel
needs
to
be
even
bigger,
so
it
depends
on
the
size
of
the
cartel
and
whether
they
can
pull
the
attack.
B
B
Yeah
exactly
so,
the
minimum
should
be
at
least
50
or
100
or
something
we
are
still
running
benchmarks
and
tests
we'll
be
able
to
determine
Mike
safe,
a
safe
baseline
for
this
seat
count.
So
it
should
be
at
least
fifty
or
hundred
I.
Think
and
bonding
our
options
also
help
with
you
know
this
issue
because
to
attack
to
do
to
perform
gatekeeping
consistently
you
need
to
in
the
bonding
option
to,
and
that
I
mean
you
can.
B
You
cannot
split
your
split
your
vid
indefinitely,
because
at
some
point
you
will
be
splitting
it
below
other
entrance,
and
so
it
depends.
This
last
point,
the
capital
on
by
the
attacker.
Another
point
I
want
to
emphasize
this
would
use
use,
find
value
in
such
a
centralized
network,
so
the
blockchains
value
proposition
is
decentralization
and
permissionless,
at
least
the
public
ones.
If
the
network
becomes
centralized
it's
easier
to
attack
and
turn
off,
then
it
loses
value
so
attacking
in
that
way
has
externalities
that
are
inflicted
in
their
reflected
in
the
token
price,
for
example.
B
A
We
definitely
need
to
look
into
it
more
I
mean
my
intuition.
Is
that
I'd
like
to
see
200
validator
slots
at
main
that
launch
yeah
as
a
starting
and
minimum?
It's
at
a
minimum.
You've
got
200
validator
slots,
which
it's
a
pretty
good
number
it's
a
bigger
number
than
most
of
the
proof
of
State
protocols
out
there
even
completely
saturated
they
have.
You
know
less
than
200
validator
nodes,
so
my
intention
would
be
at
a
minimum.
A
B
B
The
beauty
is
like
you
have,
you
can
establish
if
you,
if
you
want,
you,
can
share
your
IP
among
other
values,
establish
much
lower
latency
connections,
just
like
how
the
Bitcoin
mining
pools
did
since
a
couple
of
years.
So
that
means
and
bitcoin
is
very
conservative.
They
don't
want
to
change
anything.
Bitcoin
could
get
away
with
one
one
megabyte
block
and
I,
don't
know
is
say
like
at
least
they
could
drop
it
at
least
five
minutes
to
increase
their
throughput
by.
A
A
B
A
Know
I
think
that's
it.
So
no
questions
so
well
we'll
talk
to
you
guys
next
week
and
do
try
to
join
our
weekly
workshops
catch
us
on
discord.
If
you
want
to
learn
about
smart
contract
development,
you
can
also
watch
the
videos
on
YouTube
for
the
weekly
workshops,
as
some
great
information
there.
If
you
want
to
get
started
on
contract
development
and
learn
about
blockchain
shares
thanks
a
bunch
folks,
bye,
bye,.