►
From YouTube: CasperLabs Community Call
Description
Rewards Distribution presentation & status update.
B
A
A
Time
is
just
flying
by
so,
let's
get
started
with
this,
so
we
released
a
patch.
You
know
18
0.1
8.1,
on
April
16.
This
was
for
additional
optimizations
for
thread,
pool
management.
We
fixed
a
bug
that
caused
some
stackable
4
errors
for
syncing
a
node.
We
saw
this
with
a
multiple
validators
in
Testament,
and
so
a
coach
got
to
fix
out
there,
and
this
has
the.
A
Net
I'm
gonna
ask
everyone
to
mute
themselves.
Well,
I'm
talking
thanks,
I
appreciate
it
and
we
went
through
some
big
organizational
changes.
We
beautified
all
the
developers
into
a
single
team,
that's
led
by
Ed
Hastings,
and
he
is
things
that
I
worked
together
when
I
was
about
a
bolero
when
I
was
senior
director
of
engineering
for
the
Abbott
axe
protocol
and
was
a
senior
developer
on
the
team
and
I
have
a
tremendous
amount
of
respect
for
him
and
he
joined
the
project
in
October
of
last
year.
A
It's
actually
September
of
last
no
sorry
July
of
last
year,
my
mistake.
His
engagement
with
us
started
at
the
Prague
off-site,
and
you
know
coming
up
on
nine
months
of
the
project
and
has
been
a
force
on
the
execution
engine
team
and
really
you
know,
has
helped
the
team
kind
of
solidify
their
processes.
So
he's
a
seasoned
engineering
leader
I'm
very
excited
to
have
him
engaging
in
this
role.
He
was
very
mindful
in
in
taking
a
back
seat
right
because
he
understands
how
engineering
organizations
works.
A
He's
like
I've
I've,
run
multiple
organizations
in
the
past,
but
it's
I
have
to
do
my
time.
In
the
back
seat
and
observe
and
learn
more
about,
the
protocol
learn
more
about
what
rebuilding
learn
more
about
rust
before
I
start
really
voicing
my
opinion,
so
he
he
took
an
active
backseat
until
such
time
that
you
know
we
felt
that
we're
ready
for
him
to
take
a
leadership
role
and
he
started
doing
that
effective
Wednesday
of
last
week.
We've
also
decided
to
implement.
A
We
have
to
make
changes
to
highway,
to
solve
for
two
security
flaws
that
we
discovered
in
that
actually
exists
in
all
dag
four
choice
based
protocols,
and
we
refer
to
those
as
the
equivocation
bomb
or
equivocation
catastrophe
and
a
long
range
like
basically
a
deep
equivocation
bomb
or
and
a
shallow
equivocation
bomb.
There's
are
two
specific
types
of
attacks
that
we
discovered
and
in
order
to
address
those
attacks,
we
need
to
go
through
and
make
some
significant
changes
to
consensus.
A
So
we've
decided
to
be
opportunistic
and
take
what
we've
learned
from
the
scallop
code
base
and
we're
going
to
start
working
on
a
new
rust
implementation
of
the
node.
This
will
enable
us
to
you
know:
sync
simplify
the
architecture
into
a
single
process.
We've
already
seen
some
issues
in
Testament
with
validators:
either
not
shutting
down
one
process
or
you
know
starting
them
in
the
incorrect
order,
because
they
have
to
start
the
execution
engine
and
the
node.
A
And
then
you
know
you
have
to
clean
up
the
state
this
certain
order,
and
so
this
architectural
decision
is
bleeding
out
to
the
edges,
the
system
and
the
user
experience
for
validators,
and
it's
not
a
big
issue
now,
because
nobody
actually
has
a
real
stake.
But
when
it's
real
stake,
it's
gonna
matter
right.
You
need
to
build
a
basically
a
really
robust
system
that
has
a
single
architecture,
single
process
and
there's
a
lot
of
things
that
matter
around
shared
data
between
the
two
processes
that
has
become
very
challenging
for
us.
A
So
we
know
we
need
to.
We
need
to
move
to
single
system,
so
we've
decided
to
do
it
now.
Test
net
validators
are
gradually
updating
to
patch.
You
know
to
the
dot
18.1
release
and
we
have
decommissioned
of
net.
So
clarity
now
points
the
test
is
so
if
you
go
into
clarity,
you're
gonna
be
directed
to
the
test
net,
so
no
longer
test
out.
A
A
Our
motive
operation
with
Tess
net
is
going
to
be
deployed.
Gossiping
is
the
next
big
feature
we're
going
to
drop
into
the
test
net
and
then
testing
that's
going
to
go
into
active
maintenance
mode
for
the
next
eight
weeks.
We
will
continue
to
work
with
the
test
net
for
performance
testing.
We
have
a
plan
to
do
some
good
performance
testing
on
the
test
net
will
give
us
an
idea
of
how
highway
performs
and,
of
course,
stability
bug
fixes
such
as
you
know,
threading
tuning
and
optimizations.
A
You
want
to
take
the
round
exponent
down
to
a
shorter
round
exponent
to
get
more
frequent
blocks
in
there
when
we
perform
that
when
we
perform
the
the
performance
test.
So
there
is
a
lot
of
work,
we're
gonna
do
in
test
net.
The
next
big
milestone,
we're
planning
for
test
net
is
going
to
be
economics.
We
suspect
that
that
node
will
have
a
lot
of
enhancements
around
economics.
That
is
all
rust
work
right.
A
big
chunk
of
it
is
likely
going
to
be
rust
work.
So
it's
not
throwaway
code
right.
A
Our
next
big
milestone
for
test
that
we're
going
to
augment
as
test
the
S
test
system
and
we're
going
to
run
L
RTS
with
a
various
combinations
to
help
us
learn
more
about
the
round
exponent
and
to
also
facilitate
the
performance
testing
we
want
to
do.
We
are
now
fully
focused
on
planning
the
rust
work
and
planning
our
final
file,
design
towards
highway,
we're
going
to
take
a
fresh
look
at
storage
and
how
what
a
storage
story
is
with
a
unified
process.
A
Our
current
storage
story
is
a
little
bit
about
of
a
hodgepodge
right.
The
execution
engine
has
its
own
storage
requirements,
and
the
node
has
its
own
storage
requirements
for
consensus
and
block
management,
and
we
can
unify
those
and
get
really
simple
and
crisp
on
it.
We're
also
looking
at
integrating
with
chain
link
for
a
cross
peg.
This
is
going
to
help
us
get
market
data
for
the
token
evaluations,
and
we
want
to
do
this
to
stabilize
gas
prices
right.
A
So
basically,
two
problems
with
transaction
fees
fluctuating
wildly
one
of
them
is
going
to
be
surges,
and
this
is
the
real
scalability
problem.
The
other
problem
is
actually
the
token
price
right,
so
you
know
that's
why
you
have
this
notion
of
a
of
a
gas
of
a
gas
exchange
rate
right
is
to
try
to
solve
two
problems,
but
this
presents
too
much
volatility
for
our
end
users.
A
A
Okay,
it's
already
escaped
my
mind,
but
in
essence
we
are
doing
the
design
work
around
the
highway
protocol.
The
idea
being
is
that
we
want
consensus
to
be
more
or
less
modular
and
able
to
plug
into
the
rest
of
the
work.
This
will
enable
us
to
push
forward
rapidly
with
the
node
work
and
the
storage
work
right.
So
we
want
to
have
a
good
storage
story.
One
of
the
goals
we
have
is
to
design
checkpointing
and
state
pruning.
We
don't
want
to
have
validators
have
to
just
provision
disk
into
infinity.
A
That
seems
silly
and
not
scalable
and
not
sustainable.
So
we
will
have
a
good,
solid
story
around
checkpointing
and
basically
state
synchronization
right.
So,
for
example,
the
way
checkpointing
would
work
is
we
would
prune.
We
would
prune
the
global
state
based
on
rent
right,
so
users
would
need
to
pay
some
kind
of
rent
to
persist
their
contracts.
A
This
is
something
that
is
still
not
present
in
etherium,
and
they
still
have
been
talking
about
it,
and
so
our
intention
is
to
have
it
by
the
time
we
launch
Mena
and
additionally
pruning
away
the
blockchain
in
a
manner
such
that
you
will
get
a
new
Genesis
block
right.
So
basically
the
blockchain
will
start
from
from
State,
and
that
will
be
the
new
Genesis
block
fruit
validators
that
would
be
joining
the
network,
so
they
don't
have
to
synchronize
from
actually
the
actual
Genesis
right.
So
a
lot
of
design
work
around
this
as
well.
A
There's
also
the
notion
of
you
know.
I
and
I
have
discussed
also
this
notion
of
transaction
receipts
right.
So,
if
you're
going
to
prune
away
the
blockchain
that
I
need
proof
that
my
transaction
happened,
that
I
could
store
from
my
own
purposes.
So
I
have
this
immutable
record
on
my
side
right
like
what
I
want
to
store,
so
validators
aren't
necessarily
beholden
to
holding
onto
that.
Nor
do
we
need
to
necessarily
worry
about
necessarily
having
full
nodes
right.
We
could
have
full
knows
if
they
don't
ever
want
to
prune
their
state.
A
So
that's
a
good
chunk
talking
about
you
know
what
we're
going
to
do
with
node.
What
we're
gonna
do
is
storage,
we're
also
working
on
the
implementation
of
contract
headers
and
we're
working
on
the
host
function
cost
model.
This
is
basically
transaction
fees.
How
the
opcodes
are
priced.
What
that
looks
like
and
again
like
I
said,
the
goal
is
to
get
to
stable
transaction
fees
for
end-users,
with
a
really
highly
efficient
system,
which
makes
it
profitable
for
validators
continue
to
work
with
tests.
A
The
test
net
validators
they've
been
amazing,
we're
so
grateful
to
have
them
participating
the
project.
We
are
working
on
uploading,
the
Python
client
to
pi
pi
we're
going
to
select
a
single
client.
We
currently
support
both
the
Python
and
a
Scala
client.
Has
the
decision
been
made
on
that?
Yet
Ashok
do
you
know.
A
A
Testing
and
the
integration
tests
have
been
hogging
up
a
lot
of
time
on
CI,
and
so
we
need
to
just.
We
need
PRS
to
go
through
more
quickly
and
so
we're
going
to
basically
pivot
away
from
an
integration
test
strategy
over
to
a
unit
trust
test
strategy,
and
so
this
should
make
the
builds
a
lot
faster
ecosystem,
we're
going
to
when
you
say,
colors
logos
and
fonts
I'm,
assuming
this
is
a
rebrand
to
match
our
website.
Is
that
correct,
Ashok.
E
A
C
E
A
Great
and
then
we're
also
working
on
deploy
signatures
and
sending
deploys
from
within
the
browser,
so
we'll
be
creating
a
little
Chrome.
Basically,
a
plugin
that
will
enable
users
to
you
know
provide
they
path
to
the
private
key
to
sign
the
deploys
before
they
send
them.
I'm,
assuming
that's
what
that's?
A
What
that's
going
to
require
we're
also
doing
a
prototype
integration
with
some
exchange
software
there's
some
open-source
exchange
software
that
most
of
the
exchanges
use
and
we're
doing
a
prototype
integration
with
the
CL
X
token,
and
this
is
going
to
help
us
flush
out
what
requirements
we
need
to
fulfill
for
integrating
with
exchanges.
So
we
got
this
feedback
from
one
of
our
advisors
over
an
Arcadian
group.
Who's
done
a
lot
of
these.
These
it's
good
advice
right.
A
We're
working
on
designing
the
integration
with
chain
link
for
token
pricing
data
I
see
a
typo
there.
We
should
probably
fix
that
team
and
company
update
mark
Brinkman
joins
the
team.
Mark
worked
with
andreas
before
on
other
projects,
particularly
proof
of
authority,
I'd
like
to
welcome
Joe
Davies
he's
on
the
call
Joe.
Do
you
want
to
say
hello
to
everyone
he's
joining
hi.
C
C
In
Cambridge,
in
England,
in
yeah,
so
I'm
in
Kate,
just
in
a
little
village
outside
of
Cambridge,
which.
C
A
You
and
so
yeah
we
have
new
timings
for
a
weekly
workshops.
7
a.m.
Pacific
and
12
a.m.
Pacific
ma
che
and
Pio
Cody
are
p.m.
in
charge
of
ecosystem.
Are
our
owning
this
and
they're
going
to
be
driving
those
weekly
workshops.
So
if
you
want
to
get
help
started,
you
know
get
started
on
contract
development
and
you
want
some
help.
Just
hop
in
on
those
calls
they'll
be
there.
They're
gonna
do
some
screencasts
and
that's
really
there
to
help
top
developers
right.
A
A
E
A
B
A
Sounds
good
real
quickly,
I
got
a
I
got
a
question
who
are
the
anchors
for
the
weekly
workshops?
It's
ma
che
szalinski?
Who
is
our
smart
contracts
developer?
Those
of
you
that
have
attended
a
workshop
or
probably
have
heard
much
his
name
and
then
Theo
to
do
Becky?
Who
is
our
program
manager
for
ecosystem,
our
project
manager,
story
for
ecosystem?
He
works
with
ma
che,
and
so
the
two
of
them
are
responsible
for
hosting
the
workshops.
So,
if
you
ever
need
help
with
rust
hop
onto
those
calls
and
ma
che,
he
also
knows
solidity.
A
B
Right
so
yeah
you,
you
said
that
I'm
working
on
putting
price
gap
calculating
gas
costs
for
these
whole
functions.
So
these
are,
as
you
know,
we
have
our
own
VM,
like
the
rust
code
compiles
down
to
was
and
hope
codes
and
they
are
picked
up
by
the
VM
and
executed
and
then
at
the
end
of
execution
you
have
a
table
of
like
which
operation,
which
op
codes
were
called.
How
many
times
from
that
and
using
the
gas
costs.
You
calculate
your
like
gas
usage,
but,
as
you
know,
we
don't
have.
B
B
So
that's
the
work
I'm
doing
each
of
these
functions
need
to
have
gas
cost
bike.
This
one
has
to
cost
like
10
gas,
this
one
s2,
but
you
have
to
assign
specific
numbers,
but
that's
what
I
did
we
have
some
benchmark
air?
We
just
drawn
benchmarks
like
we
generate
different
inputs.
If
these
functions
I
accept
arguments,
we
we
just
run
them
on
these
inputs
and
get
a
bunch
of
data
points.
Let
me
go
on
some
to
something
visual.
This
is
a
for
example.
We
have
this
function
called
add
what
it
does.
B
B
We
will
like
the
estimate.
The
prediction
will
be
correct.
Most
of
the
time
it
won't,
the
the
actual
like
real-life
runtime
we
see
on
production
environment
does
will
not
exceed
our
estimate.
So
in
this
case
we
use
a
degree
of
confidence
approach.
First
of
all,
we
generate
a
large
input
sets
like
we
generally.
We
create
a
large
set
of
data
points
and
we
assume
it
represents
the
whole
population
without
like
making
any
other
assumptions.
B
We
just
assume
that,
and
we
say
which,
which
cost
should
we
take
in
order
to
have
like
99%
of
the
points
on
the
left
of
that
cost?
So
in
this,
in
this
case,
we
should
we
should
make
it
say
that
it
consumes
zero
point
something
second
one
time
for
operation
and
that
from
that
value
we
will
calculate
the
gas
cost.
Oh,
this
is
this:
is
the
approach
and
I'm
in
the
process
of
doing
this
for
all
the
old
host
functions
and
you
saw
there
so
right.
B
We
will
very
soon
have
a
efficient
costing
of
all
these
functions
and
also
the
opcodes.
That
means
we
will
be
able
to
squeeze
in
squeeze
in
much
more
performance
than
let's
say
a
non
optimized
black
shade
was
that
was
all
technical.
Let's
have
a
non-technical,
easy
presentation,
and
this
is
a
the
second
topic
I
want
to
discuss
today.
Is
the
I
just
want
to
have
a
general
overview
of
what
gases
I
just
want
to
explain
it
with
an
analogy
in
order
to
clear
out
some
misconceptions:
like
is
what
is
gas?
Is
it
the?
B
B
So
each
K
in
each
purchase,
buyers
in
each
Co
buyers,
buy
multiple
type
of
products
and
they're
happening.
Everybody
is
haggling
or
all
of
them
like
it's.
A
typical
Middle,
Eastern
market
Bazaar,
it's
if
you,
if
you've
been
like
I've,
been
to
some
of
these
retiring
right
and
it's
also
in
a
nice
way.
It
helps
with
the
analogy,
because
everyone
is
tired
of
haggling
and
they
just
want
a
different
system.
They
just
they
want
a
supermarket,
but
in
this
in
the
system,
buyers
aren't
allowed
to
just
accept
prices
from
sellers
that
is
like.
B
This
is
a
rule
that
we
assume.
Maybe
there
isn't
an
authority
to
set
those
prices
or
yeah,
like
buyers,
have
to
give
an
input
at
least
one.
So
in
this
system
a
haggle,
fries
and
I
go
through
like
getting
rid
of
haggle,
isn't
possible,
but
what
we
can
do
is
we
can
minimize
the
number
of
files
one
for
each
buyer
instead
of
haggling
for
each
purchase,
arguing
in
each
purchase,
the
buyer
could
just
haggle
once
and
then
be
done
with
it.
B
Just
these
values
we
want,
we
won't
focus
on
how
these
they're
assigned,
but
in
a
basket.
Let's
say
there
are
three
apples
to
oranges
and
six
eggs
right.
We
know
exactly
how
many
of
these
goods
are
in
a
basket,
and
then
the
seller
organization
was
responsible
for
this
and
men's
basketball
canes,
which
correspond
to
one
basket.
So
one
token,
with
one
token,
you'll
be
able
to
buy
one
basket
of
goods,
but
these
basket
tokens
like
the
goods
are
in
sell
in
a
packaged
way.
You
can
still
buy
them
separately.
These
basket.
B
Tokens
are
used
to
buy
every
good
in
the
market
and
to
buy
one
type
of
product.
For
example,
Apple
the
total
cost
is
calculated
as
apples
bought
times.
Oh
yeah
I
forgot
some
something
in
that
form
apples
both
times
apples
and
in
basket
times
everything
in
basket
times
times
yeah.
So
the
price
the
price
is
in
basket,
tokens
right,
the
price
isn't
in
whatever,
yet
that
you
are
used
to.
So
this
gives
you
how
much
basket
tokens
it
costs
to
buy
some
amount
of
apples.
B
The
problem
is:
where
do
buyers
get
these
tokens
right
and
we
just
assumed
they'd,
get
it
up
the
market
entrance
and
how
do
they
get
these
tokens?
They
get
done
by
haggling.
They
there
there
is
a
token
dispenser
and
they
do
a
bid
in
the
dispenser
and
the
dispenser
just
after
some
iterations,
they
agree
on
a
price,
and
then
the
buyer
exchanged
his
Fiat
for
some
tokens
free
to
go
into
the
market
and
buy
whatever
he
wants.
B
It's
a
better
shopping
experience
because
like
if
they,
if
they
had
go
like
20
times
during
one
go,
they
just
have
to
do
only
once.
So
what
does
it
have
to
do
with
smart
contracts?
In
this
analogy?
Just
computers
aren't
very
you,
know
tangible
for
most
people,
but
once
you
bring
it
into
market
terms,
it's
more
tangible
products
correspond
to
like
various
computational
resources.
We
have
like
CPU
memory,
storage,
bandwidth
and
basket.
Token
is
gas
haggling.
B
The
entrance
is
billing
a
gas
price
in
a
transaction
when
you're
when
you're
sending
it.
If
you
have
sent
an
ATM
transaction,
you
say:
I'm
gonna
pay
this
much
way,
for
example.
So
to
contrast
with
so
the
title
of
the
presentation
was
coupled
versus
decoupled.
So
this
is
obviously
the
coupled
one.
You
couple
all
these
resources
to
the
basket
token,
but
the
what
is
the
decoupling
one?
The
decoupling
one
is
the
initial
initial
Commission,
it's
the
very
where
everyone
is
constantly
bidding
for
these
resources.
B
It
obviously,
it
obviously
makes
it
eat
like
the
problem,
is
solved
as
obvious
I'm,
a
user
and
I.
Don't
know
the
prices
of
all
these
resources.
If
you
have
ever
been
to
I,
don't
know,
go
shopping
in
a
market
in
the
Middle
East
or
East
Asia.
You
would
have
this
like
you,
don't
know
what
the
price
is.
B
You
don't
also
in
your
theorem,
you
don't
know
the
exact
gas
price,
but
at
least
you
can
go
to
each
gas
station
and
then
you
can,
you
can
see
it's
recovered
from
past
blocks,
the
mean
the
mean
transaction
or
the
median
median
gas
price
from
the
last
last
couple
of
blocks
is
logged
there.
But
imagine
that
you
don't
have
you
don't
have
these
this
type
of
website,
then
you
would
have
to
keep
track
of
all
these.
B
B
There
is
an
obvious
problem
when
the
supply
or
demand
shifts
for
a
given
product,
and
here
you
see
a
typical
supplying
the
mentor,
but
in
blocking
typically
the
supply
curve
is
vertical
because,
like
you
have
capacity
and
then
beyond
capacity,
you
can
you
can't
have
that
like
it's?
Not
it
doesn't
it's
not
possible
instantaneously,
for
supply
to
grow
and
some
resources,
like
bandwidth,
do
not
grow
due
to
like
global
bandwidth.
B
They
simply
do
not
grow
fast
because
for
them
to
grow
fast,
all
the
world
would
need
to
upgrade
their.
You
know:
infrastructure
and
fiber
optic
cables
under
the
oceans,
so
we
will
need
to
have
superfast
internet
in
order
to
have
more
supply
of
some
resources.
So
let's
say
the
supply
or
demand
shifts
you
have.
You
should
have
a
new
price,
but
the
basket
contents
prices
prices
a
certain
product
at
the
old
price.
So
what
what
should
you
do?
You
should
then
update
the
basket
contents.
B
So
if
you
have
twice
less
twice
the
last
apples,
you
should
also
decrease
the
number
of
apples
in
your
basket
and
then,
if
the
tokens,
with
the
same
token,
you
will
be
able
to
buy
like
twice
less
apples-
and
this
is
in
order
to
this
is
in
order
to
keep
the
economic
efficiency
at
maximum.
So
otherwise,
if
you,
if
you,
if
you
don't
update
your
basket
contents
contents
as
your
supply
ships,
now
there
was
you're
in
trouble.
B
You
you,
you,
create
these
inefficiencies
and
well
in
this
in
this
market
analogy
some
some.
What
would
happen
is
some
sellers
of
products
would
be
paying
for
the
expenses
of
some
of
the
other
sellers,
but
in
the
blockchain,
that
analogy
doesn't
work
so
in
blockchain
knows
like
an
old
operator,
all
provides
all
the
resources
that
you
see
and
they
all
partake
in
and
everybody
as
an
open
as
a
node
operator.
I
have
both
a
CPU
and
a
Raman.
Also
storage
and
I
also
pay
for
my
internet
connection.
So.
B
B
Gas
like
provided
then
that
can
actually
supply
like
in
any
theorem.
A
block
can
have
maximum
10
million
gas,
and
the
reason
is
as
if
that's
the
amount
of
maximum
resources
that
can
be
provided
like
if
you,
if
you
went
about
that,
the
syrup
service
would
fail
blocks,
wouldn't
get
fine
and
finalized
or
there
the
miners
wouldn't
be
able
to
process
them
in
time,
so
I'm
making
the
analogy
with
respect
the
theorem
all
the
because
for
our
viewers,
this
is
a
community
called
so
for
our
viewers.
We.
B
B
We
will
set
it
constant
at
a
value
that
we
will
be
sure
that
wil
will
not
exceed
so
we
will
never
have
a
surge
or
it
will
be
very
that's,
that's
absolute
it.
So
it's
it
will
be
very
highly
probable
that
we
will
not
have
a
search.
We
will
be
99%
sure
that,
like
90%
of
the
time,
we
will
not
experience
in
any
sort
of
search
in
these.
B
You
know,
that's
that's
how
we
will
approach
our
pricing
ooh,
the
sort
of
strategy
where
you
use
pricing,
to
curb
the
demand,
and
it's
very
analogous
to
what
cloud
services
how
cloud
service
is
left
launched.
So
I,
imagine
you're.
You
want
to
be
a
cloud
service
provider
and
you
construct
a
data
standard,
and
you
know
some
capacity,
you
you
have
thousand
dedicated
instances
and
let's
say
you
can
just
have
maximum
ten
thousand
virtual
instances.
So
you
have
a
given
capacity.
B
You
you
would
no
matter
what,
like
you,
wouldn't
be
able
to
provide
more
than
that
to
provide
more
than
that,
you
will
need
to
construct
a
second
one,
but
it
will
take
some
time
it
will
take
one
year.
Maybe
so
what
you
do
is
you
make
sure
with
your
price?
First
of
all,
you
you
want
to
address
your
customers
use,
so
you
want
to
find
your
customers,
and
once
you
find
them,
you
want
to
make
sure
your
price
like
it's
it's.
B
It
gives
them
the
value,
so
I
mean
when
you
when
they
buy,
they
receive
the
value
they
they
paid
for.
Second
of
all,
like
actually
directly
a
direct
conclusion
in
order
for
them
to
get
the
service
they
need,
you
shouldn't
have
surges
and
you
their
servers
should
not
just
stop
at
some
point
due
to
due
to
excessive
demand.
B
Talking
about
our
pricing,
if
you,
if
you
fix
gas
price,
like
that,
the
user
experience
you
have
is
that
users
actually
do
not
choose
a
gas
price
when
you
in
the
Cask
astrolabe
boxing,
when
you're
gonna
send
a
transaction
you
actually
like
me,
theorem,
you
can
choose
now
fast,
slow,
very
fast
and
that
corresponds
to
the
gas
price
you're.
Giving
implicitly
so
before
you
didn't
have
those
options
in
meta
masks
in
the
meta
mask
is
the
most
commonly
used
wallet
before
you
directly
gave
your
like
gas
price.
B
Now
you
now
they
improve
the
UX
a
bit.
But
what
is
doing
in
the
background
is
it's.
We
want
away
if
you
wanted
to
be
included
very
fast.
You
just
put
a
very
high
gas
price,
so
in
our
case
you
will
you
won't
even
have
that
it
will
be
very
simple
and
easy.
You
will
click
confirm,
and
that
will
be
that.
So
you
won't
have
to
think
about
any
of
the
gas.
It
will
be
like
yeah
it
will
you
will.
You
will
see
how
much
gas
it
will
cost
so
how
much.
B
B
B
Once
we
achieve
scalability,
it
will
lead.
The
upside
for
users
will
be
amazing,
like
they
will
see
you.
Oh
my
god,
I
pay
ten
times
less
I
pay
hundred
times
less
for
a
transaction
amazing,
so
that
will
be
their
user
experience.
But
up
to
that
point
they
will
still
have
the
advantage
of
not
having
to
do
it
with
gas
prices
volatility.
They
will
just
pay
for
the
services
and
they
will
get
done.
B
Yeah
they
III
didn't
have
enough
time
to
add
more,
but
I
I
want
to
complete
this
completeness
as
like
a
introduction
to
like
economic
economics
of
smart
contract
platforms.
I
think
it
would
be
useful
for
people
to
read
yeah.
A
I
think
so
too
I
think
you
know
it
shines,
a
real
light
on
the
scalability,
the
real
scalability
problem
behind
blockchain
technology
right.
So
a
lot
of
is
this
real
supply
demand
problem
because
there
isn't
a
way
to
increase
the
amount
of
computation
available
using
blockchain
technology
prices
go
up
as,
and
it's
actually
one
of
the
things
that
the
chains
there
they
become
hobbled
by
their
own
success.
How
is
what
happens?
A
B
A
D
Well,
there
are
a
few
options,
so
you
know
if,
for
example,
we
make
a
major
breakthrough
on
you
know
the
cross
spec.
We
could
conceivably
return
to
that
subject.
I
mean
otherwise
something's
a
build
like
look
more
about
the
future
is
actually
sort
of
the
software
aspects
of
design
in
the
single
platform
for
builders-
and
you
know
ever
referred,
you
know,
I
self-promoted,
my
own
blog
post
on
the
SPT
sounds.