►
From YouTube: CasperLabs Community Call
Description
Engineering update, Delgator UX, How Casper is different from Ethereum, why CasperLabs is launching a separate blockchain
A
Good
morning,
good
afternoon,
good
evening,
wherever
you're
coming
to
us
from,
I
am
metta
parlikar,
the
cto
of
casper
labs,
and
today
I'm
going
to
provide
our
development
engineering
update
on
the
project
for
those
of
you
that
would
like
to
join
the
call
you
can
find
us
on
casperlabs
governance
or
casper
labs,
slash,
casper
labs
on
github,
the
zoom
information
is
available.
We
live
stream
it
so
we're
happy
to
have
folks
dial
in
on
the
zoom.
A
If
you
want
to
interact
with
us
directly,
of
course,
you
can
always
ask
questions
via
telegram
and
our
community
members
managers
will
forward
those
on
to
us
and
we're
happy
to
answer.
We
brought
a
poll
every
week
to
see
what
folks
are
interested
in
hearing
about,
and
I
believe
today
we're
going
to
talk
about
enterprise
adoption
is
that
right,
og.
A
Yeah,
so
we're
going
to
talk
about
what
I
believe
we're
going
to
do
for
enterprise
adoption
on
the
project,
so
without
further
ado,
let's
jump
right
in
and
we'll
start
talking
about
our
engineering
status.
So
we
are
in
our
september
release
this
week.
We
will
be
cutting
a
release
for
the
rust
work
that
we
have
been
doing.
We've
been
building
a
rust
client.
Our
focus
has
been
on
cleaning
up
the
scala.
A
Well,
you
know
so
moving
away
from
the
scala
logic
to
a
full,
rust
implementation,
and
I've
talked
about
this
in
previous
calls.
If
you
want
to
go
back
and
look
through
some
of
the
previous
calls
and
content,
we've
talked
about
our
reasons
for
doing
this.
It's
primarily
to
clean
up
the
architecture,
get
to
a
shared
vision
on
architectural
side
in
terms
of
concerns
around
data
storage-
and
you
know,
deploy
storage
which
enables
other
things
such
as
checkpointing
and
state
pruning,
which
are
really
really
important
for
sustainable
architecture.
In
the
long
run.
A
So
we
took
on
building
a
rust
node
in
april,
we
are
getting
close
to
being
ready
to.
We've
started
deploying
that
rust
node
into
test
beds
and
we're
going
to
be
pretty
close
here
to
having
some
of
our
validators
start
working
with
the
that
software
and
with
an
experimental
network
and
so
I'll,
be
reaching
out
to
some
of
the
validators
in
the
test
net.
To
do
that,
so
we're
looking
forward
to
that
phase
very
much
so.
A
So,
let's
carry
on
here.
Let's
talk
a
little
bit
more
about
what
we're
doing
so
yeah.
We
said
we're
getting
ready
to
set
up
a
preliminary
20
node
rust
network
on
what
we
call
our
lrts,
which
are
long-running
test
beds.
These
are
they
mimic
the
the
they're
geographically
distributed
in
aws
and
they
very
much
mimic
what
we
have
in
our
test
net
we've
been
able
to
reproduce
most
of
the
bugs
and
test
net
in
this
in
this
test
bed.
A
So
we
have
a
pretty
good
degree
of
confidence
that
it,
you
know,
resembles
a
live
living
test
net,
so
this
will
get
us
to
a
certain
level
of
internal
quality
before
we
turn
it
over
to
our
validators
right.
We
want.
We
want
to
have
some
confidence
in
what
we're
shipping
them,
even
though
it
is
experimental,
current
focus,
so
we're
focused
a
lot
on
validator
set
rotation
and
error
supervisor.
A
So
this
this
next
iteration
is
going
to
be
significantly
more
feature
feature
rich
than
the
current
test
net,
we're
pushing
towards
our
delta
testament
time
frame,
our
milestone,
which
is
going
to
involve
basically
feature
complete
right,
pretty
darn
close
to
feature
complete.
A
What
will
be
remaining
in
our
initial
from
our
initial
delta
test
net
is
probably
just
going
to
be
state
pruning,
but
and
protocol
upgrades,
so
state
pruning
and
protocol
upgrades
are
probably
the
only
two
features
we'll
attack
on
after,
but
we'll
likely
even
go
to
delta
test
net
with
checkpoints,
and
the
reason
for
that
is,
if
you
imagine
what
happens
right
now,
when
a
validator
wants
to
join
the
network,
any
network
majority
of
these
networks,
they
have
to
start
at
the
genesis
hash
with
the
genesis
block
and
they
basically
have
to
run
through
every
single
state
transition
to
get
to
the
current
block.
A
A
You
can
see
this
when
you
go
to
clarity,
even
in
our
browser,
you'll
see
the
number
of
messages
required
to
formalize
finalize
block
and
while
the
messages
themselves
are
pretty
small,
the
dag
structure
and
validating
the
dag
structure
that
all
gets
pretty
big
and,
what's
more,
is
when
a
validator
unbonds
from
the
network.
There's
no
they're,
not
providing
any
security
to
that
old
state
right
so
that
old
state
doesn't
have
security
associated
with
it
anymore,
because
those
validators
have
left
the
network.
A
A
So
if
you
can
imagine
you
kind
of
need,
checkpoints
before
you
can
do
validator
set
rotation
and
our
delta
test
net
is
going
to
be
permissionless,
so
we
absolutely
will
have
validator
set
rotation
in
place
for
delta
test
net,
which
means
we
need
checkpoints
done
before
that,
so
can't
do
checkpoints
if
you've
got
two
separate
global
state
stores,
one
for
the
scala
code
base
and
one
for
the
rust
code
base.
You
need
something
that
is
cogent
and
comprehensive
that
the
entire
node
architecture
can
use,
hence
the
rust
client.
A
So
this
is
what
we're
working
on
so
validator
set
rotation
and
error
supervisor.
This
basically
means
that,
as
part
of
an
era
rotating
the
validator
set
will
also
rotate
as
part
of
an
error.
Shifting.
A
We
are
designing
and
implementing
the
api
for
clients
and
clarity,
so
this
is
basically
the
edges
of
the
system,
how
you
will
work
with
the
rust
node
and
then
we
will
be
basically
amending
the
block
explorer
to
work
with
these
apis
and
then
design
an
implementation
of
nodejoin
and
rejoin,
and
so
I
just
talked
about
how
new
validators
will
join
come
come
into
the
network.
A
A
Even
in
these
cases
you
still
need
to
have
some
state
synchronization
right,
and
so,
even
if
we're
talking
about
state
checkpoint
state
passing
for
new
validators
joining
there's
still
going
to
be
intermittent
liveness
faults
where
a
validator
will
still
need
to
be
able
to
rejoin
the
network
right,
so
they
may,
they
may
install
an
update
where
their
node
will
be
offline
for
a
little
bit.
They
may
experience
a
short
outage,
and
so
they
still
will
need
to
provide
this.
You
know
whistling
to
provide
the
synchronization
right.
That's
what
this
join
rejoin
is.
A
If
you
have
any
questions,
feel
free
to
ask
questions,
I
don't
know
if
you
see
my
question
panel
here,
q
a
there,
we
go
and
testing
sres,
so
we're
continuing
to
work
with
our
validators
monitoring
and
reporting
for
those
of
you
that
are
validators
that
are
listening
in
tom.
Vasile
has
separated
from
the
project.
We
are
recruiting
new
devops
engineers.
A
So,
if
you're
interested
in
being
a
devops
engineer
for
the
project
do
apply
at
our
website,
there's
a
small
jobs
link
at
the
bottom
of
the
page,
and
you
can
also
locate
our
open
positions
online.
But
if
you
do
have
questions
join
us
in
the
discord,
there's
other
validators
that
can
help
you.
I
can
help
you
joe
sasher
on
the
u.s
team,
can
also
help
you
he's
providing
coverage
for
the
us
time
zone.
A
While
we
look
to
recruit
and
hire
someone
we're
continuing
to
automate
and
we're
working
on
the
20
node,
lrt
with
the
rust
node,
so
we're
starting
to
put
the
rust
node
under
load
and
performance
testing
we're
building
our
a
blog.
We
haven't
had
a
blog
on
casperlabs.io,
so
blog.casperlabs
right
now
points
to
our
medium
page,
but
we're
going
to
build
a
full
blog
site.
We
want
to
hear
from
the
developers,
so
the
devs
will
start
writing
articles
things
that
they're
interested
about
their
thoughts
on
what
it
is.
A
They're
building,
oh
noor
and
alex
have
been
providing
a
lot
of
content
for
us
and
it's
some
owners.
Content
is
on
hackmd.
Some
of
it
is
on
medium
and
so
we're
really
building
a
blog
site.
So
we
can
kind
of
break
pull
this
all
together
and
I'm
sure
all
of
you
want
to
hear
from
our
engineers
here
from
the
core
dev
team
and
so
they'll
have
an
opportunity
to
put
out
their
own
thoughts
via
this
blog
we're
continuing
to
extend
the
solidity
transpiler
to
include
enums.
A
So
there's
a
lot
of
work
to
be
done
on
the
transpiler.
Our
initial
alpha
supports
the
erc20
contract,
we're
going
to
be
deploying
the
prototype.
Basically,
the
we'll
release
a
prototype
of
the
solidity
transpiler
and
the
erc20
and
put
a
blog
post
out.
I
want
mache
to
do
that.
He's
he's
still
busy
working
on
it,
but
hopefully
that's
coming
soon
and
then
we're
enhancing
clarity,
basically
to
allow
a
parser
for
cl
types.
A
So
what
this
means
is
when
you
are
deploying
from
the
browser-
and
you
need
to
send
your
arguments
basically
it'll-
validate
that
your
arguments
are
of
the
correct
type
we're
going
to
implement
some
metrics
and
clarity.
We
want
to
understand
how
people
are
using
the
system,
how
people
are
using
the
blockchain
clarity
has
a
lot
of
this
information,
so
we're
going
to
do
that
via
prometheus
and
grafana,
and
then
we're
adding
support
for
arguments
in
the
dsl.
So
we
we
came
out
with
a
contracts
dsl
that
makes
contract
authoring
significantly
easier.
A
A
You
know
with
the
support
of
the
dsl
and
headers.
It's
a
lot
simpler,
we're
writing
tech
specs
for
our
auto
price
adjustment.
This
is
to
basically
provide
price
stability
for
transactions,
as
we
have
promised
for
a
long
time
to
our
customers
and
alex
is
also
continuing
to
work
on
the
chain
link
oracle
to
help
express
those
prices
in
fiat.
A
And
we're
designing
the
process
for
dealing
with
equivocation
catastrophes
guys
on
the
econ
team.
I
knew
yourself,
but
I'm
assuming
that
this
is
social
consensus.
Is
that
correct?
When
we
talk
about
that.
A
Okay,
great
yeah,
social
consensus
is
basically
what
happens
in
the
event
of
an
attack
or
consensus
failure.
We
assume
you
know
we're
going
to
be
turning
this
network
over
to
the
validators,
and
the
validators
will
need
to
make
sure
that
the
network
stays
up
and
is
the
correct
fork
is
what
they
are
building
on
in
the
event
of
a
fork
in
the
event
of
a
of
a
catastrophe,
we
need
to
provide
apis
and
tools
for
them
to
select
a
correct
fork
and
ensure
that
the
network
stays
up
and
is
is
doing
the
right
thing.
A
So
we
can't
assume
that
we've
found
every
possible
attack,
so
we
work
we're
actually
building
the
tools
for
the
validators
to
recover
the
network
in
the
event
of
a
catastrophic
failure.
If
there's
any
questions
coming
in
I'm
happy
to
answer
them,
I
see
some
folks
on
the
call
that
have
joined
I'm
allowing
them
to
talk.
You
can
talk
or
you
can
ask
questions
by
the
q
a.
A
A
You
know
in
terms
of
being
a
software
provider
that
provides
software
for
enterprise
companies
and,
if
you
think,
a
little
bit
about
how
enterprise
organizations
manage
their
internal
software,
you
know
rfi
rfc
process
how
they
select
software,
how
they
implement
and
invest
in
software
they
really,
they
really
have
a
very
slow
adoption
process
right.
So
there's
a
couple
of
things
that
enterprise
companies
are
going
to
be
looking
for,
you
know,
and
if
you,
if
you've
read
the
book
crossing
the
chasm
which
I
have
I
read
many
many
years
ago,
a
big
fan
of
it.
A
You
know
you're
gonna,
you're,
gonna,
look
at
what's
happening
with
blockchain
blockchain
technology
and
you're.
Gonna
know
that
we're
very
early,
very,
very
early
in
the
adoption
cycle
still
we're
still
very
much
in
the
space
of
the
enthusiasts
and
the
early
adopters
right,
and
you
know,
in
order
for
enterprises
to
cross
the
chasm
and
adopt
technology
they're
going
to
need
to
see
a
few
things.
A
They're
going
to
need
to
see
a
lot
of
use
cases
they're
going
to
need
to
see
things
like
who
the
leader
is
based
on
the
gartner
magic
quadrant
they're
going
to
want
to
see.
You
know
basically,
social
proofs,
that
this
technology
is
here
to
stay
before
they
go
ahead
and
invest
in
it.
A
They're
going
to
need
things
like
being
able
to
control
that
technology
right,
so
they
have
their
own
requirements
around
how
they're
going
to
control
that
software
and
that
technology
we've
built
a
lot
of
tools
in
our
core
platform
to
support
these
kinds
of
requirements.
Right
now,
that's
not
necessarily
going
to
help
us
with
the
adoption
curve
right.
So
what
are
some
of
the
things
that
casper
labs
needs
to
do
one?
A
We
need
to
be
fairly
zen
about
what
the
adoption
curve
is
right,
and
so
I'm
not
going
to
say
that
we're
going
to
push
for
enterprise
adoption
in
the
next
18
months
right,
enterprise
adoption
is
a
long
tail
strategy.
It
takes
a
long
time
to
get
enterprises
to
adopt.
So
what
does
that
mean?
Well,
what
that
means
is
enterprise
will
also
adopt,
I
believe,
through
middleware
right.
A
So
what
we
can
do
is
we
can
definitely
partner
with
aggressive
young
hungry
startups
that
are
building
relationship
with
middleware
providers
and
we
can
help
them
build
a
blockchain
enabled
solution
that
enterprise
uses
by
middleware
technology-
and
this
is
an
I've-
talked
about
this
a
lot
before
this
is
a
model
that's
been
adopted
by
avalera
with
resounding
success
right
it
gives
it
gives
a
very
strong
moat
in
terms
of
you
know,
because
when
you
work
with
middleware
providers,
they
invest
a
lot
of
time
and
energy
in
implementing
your
technology,
and
this
gives
them
a
competitive
advantage
against
other
middleware
providers.
A
That
may
not
offer
these
unique
value-added
services,
but
then
also
it's
very-
it's
not
very
disruptive
for
enterprise
companies
to
adopt
your
technology,
because
they're
already
using
the
middleware
right.
A
great
example
is
adapters
that
integrate
into
erp
systems
adapters
that
integrate
into
electronic
medical
record
systems.
So
you
go
to
a
large
health
care
system
and
they're
using
elec
electronic
medical
records
system
or
revenue
cycle
management
tools,
and
you
say
well,
if
you
want
to
manage
your
revenue
cycle
and
and
make
it
more
efficient,
you
can
use
this
tool
right.
A
You
can
use
this
adapter,
and
this
adapter
gives
you
proof
of
when
services
were
provided
in
an
anonymous
way
and
that
proof
of
when
service
is
provided
is
an
immutable
transaction
of
the
blockchain.
A
But
the
hospital
system
doesn't
really
care
that
it's
using
blockchain
technology
they're
only
going
to
care
that
it's
hipaa
compliant.
That
means
that
the
patient
data
can't
be
compromised
and
they're
only
going
to
care
about
the
proofs
that
the
system
provides
the
underlying
proofs
right.
They
don't
need
to
know
that
a
blockchain
is
necessarily
being
implemented
or
you
know
they
just
need
to
know
that.
A
So
this
is
a
basically
a
channel
based
partnership
model.
It's
very
much
grounded
in
business
development
right.
So
these
kinds
of
relationships
start
with
business
development.
Business
development
is
a
long
tail
process
whereby
you
come
to
an
agreement
in
terms
of
how
they
sell
the
services
and
how
everybody
is
profitable
as
a
result,
and
we
all
win.
A
So
this
is
my
you
know
a
short-term,
long-term,
medium-term
and
long-term
strategy.
For
you
know
enterprise
adoption.
In
the
long
term,
I
think
enterprise
will
ultimately
adopt
blockchain
either
through
private
blockchain,
hybrid
blockchains
or
in
eventually
public
blockchains.
I
see
public
blockchain
adoption
very
far
in
the
future.
Almost
10
years
I
see
hybrid
and
public
blocked
private
blockchains.
A
A
I
see
hyperledgers
being
very
hard
to
use
with
limited
support,
and
so
we
believe
that
we
can
carve
out
a
niche
for
ourselves
here,
based
on
some
of
the
features
that
we
built,
which
is
you
know,
the
cargo
casper
labs,
which
I've
talked
a
lot
about
the
rust
sdk
that
makes
development
extremely
easy.
The
solidity
transpiler
that
helps
solidity
developers
be
successful
even
in
rust
if
they
want
to
cross
the
chasm
over
to
rust
and
and
basically
be
able
to
test
their
software.
A
A
There
really
isn't
a
way
for
you
to
scale
out
your
software
development
and
engineering
practices
without
it,
and
so
that
and
contract
management
tools,
governance
tools
for
software,
because
I
believe
all
software
is
governed.
This
is
all
possible
on
chain
securely,
transparently
with
casper
labs,
and
that's
what's
going
to
make
us
massively
successful
in
enterprise
in
the
long
run.
I
believe
any
questions
or
comments
about
that.
A
B
B
You
know
these
decisions
that
we
have
to
make
about
actually
assigning
fiat
prices
to
different
transactions,
including
transfers
so
well,
and
my
answer
was
that
it
is
complicated
because
we
have
gone
away
from
the
ideas
that
you
know
there's
going
to
be
anything
like,
you
know,
fixed
prices,
but
you
know
we
are
not
really
going
in
the
direction
of
ethereum
as
it
is
now
right
because
ethereum
right
now
is
you
know
it's
a
first
price
auction
for
inclusion,
and
you
know
that
is
basically
what
what
guarantees
is.
B
There
is
massive
volatility
right.
We
of
course
want
to
avoid
that
we
want
to
have
something
that
is
predictable
so,
but
nevertheless
you
know
we
still
have
to
you
know
scale
prices
up
and
down
to
reflect.
You
know
supply
and
demand
now
and
well,
of
course,
in
this
system,
supply
is
effectively
fixed,
but
demand
still
varies.
You
know
day
to
day,
or
you
know,
depending
on
you
know,
other
exercise
factors.
B
So
there
are
basically
three
approaches
to
you
know
dealing
with
this
right,
these
kind
of
compromises
between
completely
fixed
prices
and
just
give
these
free
rules.
B
The
first
one
which
many
listeners
may
be
familiar
with
is
eip
1559
and,
roughly
speaking,
this
is
a
you
know:
price
adjustment
mechanism,
where
there
is
a
price
determined
by
a
base
fee,
determined
by
an
algorithm
which
is
targeting
a
certain
gas
utilization
level
and
based
on
the
gas
utilization
level
right,
whether
it's
over
under
the
target,
it
adjusts
the
prices
either
way.
Of
course
it
is
known
in
advance-
and
this
is
based
on
the
prior
history
of
the
chain
right.
B
So
it's
it's
predictable
to
a
certain
degree
and
there's
a
fairly
narrow
range
to
you
know
how
far
it
can
vary
within
you
know
the
you
know,
space
of
your
blocks,
so
on
top
of
this
base
fee,
which
is
burnt
burned
to
you,
know
avoid
being
concerned
with
any
possibly
manipulation
by
miners.
There's
also
a
premium
right,
and
this
premium
is
basically
this
is
this
still
functions
more
or
less
like
the
bits
right
that
you
normally
submit
in
ethereum
right?
B
This
goes
to
the
validator
and
the
you
know
web
is
miner
and
they
would
be
free
to
order
transactions.
You
know,
according
to
how
high
these
premiere
are.
Our
version
of
this
is
honors
auto
price.
As
far
as
I
understand
currently,
the
main
difference,
in
fact,
between
auto
price
and
eip
1559,
is
that
auto
price
in
the
last
iteration
looked
at.
They
believe
targets.
B
C
D
It's
a
different
time
scale
so,
like
eip,
1559
just
looks
at
block
fullnesses,
whereas
I
look
at
the
whole
like
aggregate
fullness
of
one
day
and
the
rate
of
time
scale
is
different,
so
it
adjusts
daily,
whereas
eip1559
I
just
adjusts,
I
think
every
block.
So
it's
it's
more
like.
B
B
Right,
but
so
my
so,
but
I
hope
you
can
see
that
these
are
you
know
somewhat
related
conceptually,
and
the
owners
version
also
allows
for
this
premiere.
B
Fundamentally,
allowing
premiere,
you
know
it
doesn't
quite
wash
the
problems
that
arises
in
the
theory,
in
the
sense
that
basically,
once
the
blocks
are
full-
and
you
know,
since
these
premier
still
allowed-
you
know
you,
I
suspect
it
will
get
the
same
volatility.
Spikes
and
or
I
mean
there
may
be
attenuated
somewhat
because
you
know
but
yeah.
I
think
they're
still
going
to
be
there
to
a
certain
degree.
B
Apartment
had
a
problem
with
my
phone
for
a
second
right,
so
I
I
think
that
allow
this
premier
to
remain
maybe
potentially
a
problem
in
the
sense
that,
like
you,
you
end
up
with
a
more
complex
systems
that
more
or
less
works
out,
like
you
know
the
regular
ethereum
options
now
this
is
not
a
total
definitive
and
I
sort
of
try
to
study
this
question
in
depth
last
week
and
they
realized
that
in
order
to
actually
model
this
formally
would
take
like
a
release
cycle
or
two.
B
The
reason
for
this
is
even
though
both
order
and
ethereum
researchers
do
have.
You
know
fairly
detailed
blog
posts.
You
know
with
accompanying
codes
that
do
the
simulations
for
how
you
know
these.
You
know
adjustment
processes
shake
out.
The
problem
is
that
for
simplicity,
both
of
them
assume
that
you
know
the
users
are
following
some
very,
very
simple:
behavioral
rules
right,
so
the
users
are
not
forward-looking
right,
so
it's
not
like
they
can
anticipate.
B
Oh
I'm
observing
you
know
a
high
activity
right
now,
I'm
not
sure
who
is
entering
right
now,
but
they
know
that
you
know.
I
have
good
reasons
to
believe
that
in
the
next
period
you
know
the
auto
price
adjustment
will
send
this
much
much
higher.
So
now
I
might
actually
want
to
pile
in
right
now
with
a
much
higher
premium,
and
you
know,
then,
then
you
recreate
this.
You
know
this
situation
with
the
spikes.
Basically
because
you
have
all
these
uncoordinated
users
trying
to
get
ahead
of
the
you
know
expected
price
increase
right.
B
D
I
think
the
best
way
is
just
to
try
and
try
and
see
you
know
like
deploy
it
out
in
the
wild,
and
then
I
mean
well.
B
I
would
like
I
I
mean
yeah,
I
mean
I
want
to
avoid
the
situation,
but
to
be
to
be
clear,
I
understand
why
owner
wants
is
a
premium
because
there's
a
very
subtle,
like
it's
very
hard
needle
to
thread
right,
because
you
know
if
we
come
up
with
some
kind
of
a
you
know,
price
system
that
is,
you
know,
not
sufficiently
flexible.
B
You
will
drive
people
towards
accelerators
right
and
at
that
point,
you're
back
in
a
situation.
That's
arguably
even
worse
than
what
happens
in
ethereum
right.
So
we
don't
want
that.
So
we
want
a
flexible
system
that
reflects
demand.
But
you
know
we
don't
want
it
to
reflect
the
demand
that
you
know
so
much
that
you
know
it
like.
It
reflects
every
single
little
spike
up
and
down
as
it.
A
B
So
well
I
mean,
I
think,
that
you
know
the
typical
business
may
not
be
necessary.
B
If
you
know
you
kind
of
strike
the
right
balance
with
how
quickly
a
price
auto
adjusts
right,
you
don't
want
to
adjust
too
quickly,
but
you
want
to
adjust
it
just
quickly
enough
that
you
are,
you
know
you
are
satisfied,
visited
you're,
not
gonna,
go
to
you
know,
you're,
not
gonna
go
to
an
accelerator,
because
that's
also
a
cost
to
you
right.
It's
a
cognitive
cost.
It's
a
time
cost
to
use
an
accelerator
instead
of
you
know,
just
like
you
know,
you
use.
B
Right,
but
we
also
don't
want
to
you
know,
open
the
gates,
the
situation
which
becomes
equivalent
to
just
a
straight
first
price
option.
I
mean,
I
think
it's
a
long
run.
I
think
in
the
long
run,
the
real
solution
is
that
there
is
going
to
be
a
reservation
system
right
and
basically
it's
what
we
previously
called
gas
futures
right.
So
at
some
point
in
time
this
problem
is
going
to
be
solved.
B
Precisely
like
you
know,
you
will
be
just
you
know
your
business,
you
show
up,
you
know,
okay,
like
I
anticipate
such
and
such
demand
over
time
right.
This
is
my
anticipated
growth
curve.
I'm
gonna
go
to
the
system,
I'm
going
to
say
I
am
going
to.
You
know
basically
reserve
this
much
gas
right
and
probably
through
some
sort
of
adoption
system,
but
you
will
know
in
advance
how
much
you're
paying
isn't
that.
B
Eos
is
really
weird,
like
I
mean
it's
it
it
just
it
does
it's
sort
of
like
it
prices
computation,
but
it's
like
you
know
they
unbundle
the
computation
right.
You
could
pay
for
the
cpu
cycles.
You
can
pay
for
the
ram.
So
that's
different
from
what
I'm
considering
here
here,
I'm
considering
basically
ability
to
reserve
some
gas
on
the
platform.
You
know
at
some
point
in
the
future
right.
D
A
I
I
I
yeah,
that's
fine,
like
I
have
no
issues
with
with
saying
that,
I'm
I'm
trying
to
understand
what
it
is
close
to,
because
I
do
know
you
can
buy
gas,
I
mean
by
buy
ram.
I
think
it's
based
on
a
reservation.
It's
kind
of
like
you're,
reserving
how
much
compute
you
want
to
use
on
the
system
is
what
it
feels
like
to
me.
D
And
then
yeah
there
is
no
really
good
documentation
about
like
I.
Maybe
I
should
go
to
source
code
and
study.
That's
also
what
once
nice
to
do
to
understand.
Bitcoins
like
difficulty
adjustment,
because
also
it's
not
super
well
documented,
and
I
had
to
do
that
for
my
research
in
auto
price
and
yeah,
but
I
don't
want
to
go
into
eos
or
source
code
likes
no.
B
B
D
Eventually,
if
you
have
limited
supply-
and
you
have
you
like
usage,
then
like
you
implement
this
gas
reservation
system,
then
people
build
derivatives
on
top
of
reserve
gas
start
trading
it,
and
now
we
have
the
same
volatility
you
know
like.
If
you,
the
the
reason
is
limited
supply,
you
don't
you
can't
solve
it
unless
you
start
something
else.
A
Yeah,
the
idea
is
to
mitigate
it.
I
mean.
Ultimately,
the
real
problem
we
have
to
solve
is
scalability
right.
That's
what
we're
here
for
right!
That's
that's
why
one
of
our
key
value
propositions
is
scalability.
Right
is
because
you
don't
want
blocks
to
get
full
because
you
can
provide
enough
scale.
So
then
you
can
add
more
validators.
A
You
can
scale
the
architecture
out
right,
so
the
idea
is
that
highway
protocol,
eventually,
particularly
with
omega
blocks,
will
ultimately
be
able
to
scale
as
bandwidth
increases,
because
you
can
have
more
and
more
validators
proposing
blocks
you're,
not
tightly
constrained
to
a
single
single
block.
A
You
know
the
leader
will
just
ultimately
wind
up
synchronizing
the
network
at
each
at
some
point
right
to
help
with
finalization,
but
there's
a
lot
of
transactions.
In
theory,
we
should
be
able
to
scale
out
the
network
over
time
right,
as
we
add
more
and
more
validators
and
then,
as
we
add,
more
and
more
shards.
B
Right,
I
mean
the
reason
I
do
not
address
this
when
they
talk
about
this
stuff
is
because
it's
sort
of
you
know,
I
mean
it's
a
problem
that
is
closely
related
to
just
the
straight
economics
of
the
system,
but
it's
really
sort
of
you
know
on
the
side
of
you
know,
for
example,
andreas
and
the
other
people
directly
working
on
consensus
right.
So
I
cannot
say
anything
particularly
meaningful
about
how
we
might
do
chardonnay.
C
A
Right
is
you
can
use
clx,
prime
in
a
given
shard
or
you
know
the
token
prime.
C
A
A
given,
shard
and
and
those
transactions
can
happen
without
being
blocked
at
the
base,
but
anytime
you're
talking
about
you
know
doing
cross-shard
transactions.
Those
will
still
get
locked
up.
I
mean
you
know,
slowed
down
by
the
the
root
shard,
so
yeah
it's
starting
is
a
non-trivial
topic.
There's
a
lot
to
unpack
in
there.
B
Yeah,
so
this
is
why
I'm
you
know
I
move
the
mind
that
will
probably
end
up
into
making
some
kind
of
a
gas
reservation
system
before
we
get
to
like
full
of
chardonnay,
but
for
the
closer
term
goal
of
actually
getting
like
sensible.
You
know,
sensible,
you
know,
price
adjustment.
B
You
know
I
mean
I
agree
with
order
to
a
certain
degree,
in
the
sense
that
yeah
I
mean
we'll
see
how
it
behaves
in
the
wild,
and
you
know
the
beginning.
The
stakes
are
relatively
small,
so
you
know
you'll
be
able
to
make
adjustment,
but
I
think
that
after
we
implement,
you
know
the
core
maintenance
features
we
should
probably
switch
to
sort
of.
You
know
like
pure
research
mode
for
a
couple
of
cycles,
because
some
of
the
stuff
that
we
implemented
actually
needs,
like
formal
analysis,.
D
B
I
mean
no,
no,
I
mean
I
looked
okay,
I
mean
it's,
not
it's,
it
might
be
a
hybrid
where
we
may
have
to
compute
the
equilibrium,
but
it's.
B
Solutions
forget
forget
about
simulations.
If
we
could
actually
improve,
I
mean
there
are
simulations
of
a
certain
kind,
but
they're
very
particular
right
I
mean
you
can't
actually
compute
equilibria.
Even
if
you
don't
have
you
know,
actual
analytical
solutions
right
and
but
I
agree,
it
is
very,
very
hard
to
set
up,
which
is
why
I'm
saying,
but
it's
worth
it
to
try
right
because.
B
D
We
can
try,
but
I
had
like
some
sort
of
understanding
of
how
this
ecosystem
works
on
like
d5
and
blockchain
in
general.
It's
just
that
there
are
so
many
projects
that
launch
and
they
all
try
at
something
without
actually
you
know
knowing
what
they're
doing
and
good
ones
eventually
survive
become
like
the
base
protocols
and
yeah
like
we
should
just
hope.
Our
sorry
trust
in
our
engineering
instincts
intuition-
and
you
know
we-
we
have
like
some
common
sense
well,.
B
I
mean
you
know
I
I
I
spent
you
know
I
spent
six
years
in
the
doctoral
program
in
economics.
I
have
no
engineering
intuition.
Okay,
that's
not
the
thing
that
part
of
the
brain
does
not
exist.
Well,
that's.
A
But
you
know
what
I
mean.
One
of
the
things
we
do
plan
on
doing
in
the
delta
test
net
is
trying
to
have
it
be
as
real
as
possible,
so
release
a
token
limit.
The
supply
turn
off
the
faucet
force.
People
make
people
have
a
way
to
acquire
the
test
net
token,
so
they
can
run
transactions
in
the
in
the
network,
and
this
hopefully,
will
give
us
some
sense
of
how
you
know
it's
actually
going
to
work
in
terms
of
you
know,
and
we
because
the
network
will
be
fairly
small.
A
Let's,
let's
see
what
happens,
maybe
we
can
try
to
run
some
real
life.
You
know
as
get
get
a
sense
of
what
the
real
economics
of
the
chain
are.
Gonna
look
like
this
was
one
of
the
ideas
I've
been
talking
with
the
dev
dao
about
as
well
having
them
help
us
launch
the
delta
test
net.
It
will
be
permissionless
and
you
know,
validators
can
come
and
go.
A
They'll
have
to
acquire
their
stake,
we'll
have
a
mechanism
by
which
they
can
acquire
their
stake,
in
addition
to
validators
that
did
participate
in
the
private
sales,
etc,
etc.
So
we're
trying
to
make
it
as
real
to
mainnet
as
possible,
and
hopefully
we
can
learn
something
about
the
tokenomics.
You
know-
and
this
will
be
a
one
big
hot
topic
right.
I
I
really
wanted
to
just
fix
prices
in
fiat,
but
I
understand
I
mean
fixed
prices,
firmly
right
that
you
would
basically
run
a
computation.
D
Protocol,
it's
just
just
a
price
like
difference
between
a
price
floor
and
a
fixed
price.
I
mean
you
know
you
have
price
floor.
People
can
bid
above,
like
the
tip
the
premium
shouldn't,
really
really
confuse
anyone.
It's
just
that
you're
able
to
over
bid
the
fixed
price.
The
protocol
is
adjusting
basically.
B
Yeah
I
mean
they
just
I
mean
like,
for
example,
my
primary
concern
was.
The
tip
is
simply
that
you
know
it
may
form
the
perspective
of
users.
It
may
end
up
looking
no
different
than
the
first
price
option,
because
I
don't
care
where
the
basically
goes,
so
it
may
not.
Actually,
you
know,
may
not
actually
do
like,
basically,
all
the
behavior,
so
the
users
may
stay
the
same
as
they
are.
You.
D
Know
I
just
want
to
like
give
an
example
from
ethereum
like
if
you
have
an
unscaled
blockchain,
your
your
my
order
of
magnitude
for
changes
is
like
100
times
like.
If
you
have
been
observing
ethereum
lately,
you
have
seen
gas
prices
like
200
300
before
it
was
like
two.
It
was
like
ten
eight.
Now
it's
three
hundred
three
hundred
two
hundred,
and
so
it
we
want
to
be
as
successful
as
that.
So
that
means
somehow
contradicting
having
a
fixed
fiat.
D
B
D
B
Basically
trying
to
model
this
formally
and
as
I
said,
it
is
difficult,
but
it
is
possible
right
it
just
it
cannot
be
done
in
the
week,
but
it
can
be
done
and
the
same
also
the
same
also
applies
to
our
auctions,
which
also
should
that
one
actually
I'm
pretty
sure
it
might
be
like
possible
to
solve
exclusively
for
equilibrium.
B
A
Be
we're
going
to
find
out
one
way
or
the
other
and
we're
going
to
find
out
one
way
or
the
other
great
discussion
guys.
I
appreciate
the
input,
and
always
the
spirited
discussion
is
great.
You
know
these
aren't.
These
are
hard
problems,
we're
trying
to
solve
and
a
lot
of
a
lot
of
what
we're
building
is
based
off
of
game
theory.
A
I
think
we
try
to
do
modeling
formal
modeling
as
much
as
we
can
that's
why
we
we're
we
talk
about
cbc
casper,
correct
by
construction,
because
we
try
to
be
grounded
in
formal
models,
even
formal
game,
theoretic
models,
because
you
can
formally
model
some
of
those
things
out,
but
then
also
there
is
a
portion
of
it.
That
is
intuition
right.
A
lot
of
a
lot
of
things
are
driven
by
intuition.
A
lot
of
the
great
people
that
have
done
great
things
in
the
world
are,
besides
being
futuristic
thinkers.
A
They
have
followed
their
intuition
because
they
see
they
see
things
a
little
differently.
So
I
think
both
of
you
are
right
and
I
think
both
of
your
perspectives,
you
know,
bring
a
lot
to
to
how
we
think
about
the
economics.
So
I'm
glad
that
both
of
you
are
on
the
project
and-
and
you
know
contributing
the
way
you
are
so
great
discussion
today.
I
love
that
we
had
six
attendees
today.
A
Please
do
join
us
next
week
again,
if
your
time
permits
and
and
feel
free
to
ask
questions
and
and
engage
more
than
just
listening
and
love
to
love
to
be
put
on
the
spot.
I
do
enjoy
that
and
challenged.
So
thanks.
So
much
for
dialing
in-
and
I
hope
to
see
you
next
week
cheers
bye-bye.