►
From YouTube: CasperLabs Community Call
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
I'm
doing
okay,
I'm
doing
okay
yeah!
You
can
talk
more
after
yeah,
okay
good
morning,
good
afternoon
good
evening.
A
Everyone
where
you're
coming
to
us
from
my
name
is
metta
parlikar,
I'm
the
cto
and
one
of
the
co-founders
of
casper
labs,
providing
our
update
to
you
today.
I
hope
everyone
is
safe
and
well
out
there
let's
get
started
so
the
team
has
started
their
third
weekly
sprint
of
the
21.01
release
cycle,
sprint.
9.3.
A
We
are
dropping
in
a
brand
new
node
discovery
and
gospeling
protocol
and
we're
also
fully
focused
on
hardening
and
robustness
this
this
round
this
iteration.
So
let's
continue
forward.
We
currently
have
a
delta
test
set
up
with
four
validators.
A
A
Yeah
so
stay
tuned
in
the
discord
channel
for
updates
on
that,
and
hopefully
this
one
will
have
shaken
out
enough
issues
with
the
code
and
it's
a
very
new
version
of
the
code.
It's
3.1
right.
We
just
pulled
that
off
of
master,
so
it's
got
a
lot
of
good
fixes
in
it,
and
we've
heard
good
things
from
the
validators
so
far
in
their
preliminary
testing
and
another
piece
on
the
delta
test
net.
A
A
Probably
I
have
to
do
a
form
for
that.
We'll
probably
do
a
forum
for
that
current
focus.
So,
let's
see,
we've
got
our
implementation
of
finality
signatures.
This
improves
the
trust
and
security
of
the
network
disabling
bids
on
liveness
failures.
So
basically,
what
happens
is
if
this
brought
the
delta
test
that
down
a
couple
of
times
that
if
you
are
a
big
validator,
you
got
a
lot
of
bond
in
the
network
and
you
happen
to
go
offline.
We
have
to
disable
your
bid
from
consensus.
A
Otherwise
we
wind
up
having
liveness
failures
in
the
network
crash
recovery
is
part
of
hardening.
So
when
the
network
crashes,
how
does
it
recover?
We
also
finish
the
paper,
so
we're
going
to
be
submitting
it
to
archive
and
start
the
peer
review
process,
but
we're
now
we've
got
a
spanx
ci
to
get
it
to
build.
So
we
get
the
latest
paper
up
on
the
website
on
the
node
rust
side.
Lib
p2p,
gossip-
and
this
is
this-
is
a
direct
message.
Behavior.
This
is
the
the
net
small
network
component
we
talked
about.
A
We
are
going
to
be
removing
the
account
hashes
implementation
will
still
support
multiple
accounts,
but
internally
in
the
global
state,
we're
going
to
represent
that
as
just
the
public
key
with
a
pre-pended
zero
one,
two
three:
it
can
support
up
to
10
different
key
types,
10
different
encryption
types.
A
At
least
this
is
going
to
this
is
going
to
make
things
a
little
bit
easier
to
track,
so
you
can
see
the
public
key
right
in
the
global
state
network
discovery
again,
that's
related
to
lib
p2p,
so
we
are
changing
out
to
cadembly
from
the
small
network
component.
That
was
the
prototype,
the
merkle
proof,
yeah
merkle
proof
key
enumeration.
This
is
research
on
fast
sync
that
we're
doing
so.
We
know
we
need
to
do
fast,
sync
both
for
for
upgrades.
So
this
is
really
to
support
hard
forks
protocol,
hard
forks.
A
B
I
think
this
is
just
to
reduce
the
size
of
events.
I
read
the
ticket.
It
says
that
the
current
event
size
is
pretty
high
and
it's
not
required-
and
this
is
just
discovering
and
trying
to
bring
it
down.
Okay,.
A
Check
account
when
submitting
deploys,
so
what
this
basically
means
is,
if
you
submit
a
deploy
from
an
account
and
because
we
have
touring
complete
payment
code,
it
is
possible
to
do
complex
things
with
payments.
The
account
that
signs
the
deploy
is
basically
the
guarantor.
A
That's
what
this
is-
and
this
is
again
to
prevent
another
attack
vector
put
deploy.
Rpc
endpoint
will
return
an
error
if
the
deploy
is
invalid.
So
right
now
we
just
send
it
200
and
it
just
means
we
get
your
deploy,
but
we
don't
perform
any
validation
on
the
deploy.
So
now,
what
we'll
do
is
we'll
get
a
200
you'll
have
some
guarantees
that
the
deploy
is
at
least
valid.
A
S-Test
we're
extending
workload,
generators,
we're
creating
some
scripts
to
help
provision
systems
inside
of
ec
ec2,
and
we
are
trying
to
get
create
style
right
into
the
publishing
for
the
tag
pipeline
to
speed
up
a
release
process
we're
building
our
javascript
sdk,
you
can
get
to
it
on
npm
and
we're
making
more
improvements
on
the
event
store
and
clarity.
B
This
is,
I
think,
for
the
clarity.
A
Okay,
this
is
for
automated
testing
yeah.
A
A
The
contract,
runtime
folks,
are
also
looking
at
the
removal
of
account
hashes.
So
that's
another
piece
that
that
henry
is
working
on
they're
also.
A
A
Gas
spot
market
specification
so
alex
is
working
on
this
and
creating
some
auction
ceps
so
on
the
gas
spot
market
specification
alex
is
also
doing
a
bunch
of
work
around
the
ecom
paper
as
well.
So
what
are
the
pieces?
How
is
our
thinking
on
gas
spot
market?
Do
you
want
to
talk
a
little
bit
about
that.
C
Alex
right
so
my
thinking
pretty
much
it's
more
or
less
exactly
where
it
was.
You
know,
broadly
speaking,
several
weeks
ago,
which
is
that
you
know
we
are
going
to
implement.
Initially,
you
know
a
system
that
is
very,
very
similar
to
ethereum,
except
that,
because
we
have
you
know,
consensus
before
execution,
you
will
be,
you
know
paying
for
all
the
gas
that
you're
reserving
all
the
time,
and
you
know
in
order
to
you,
know
kind
of
sweeten
this
medicine.
C
C
You
know-
and
you
know
we
had
some
internal
pushback
even
against
this,
because
you
know,
for
example,
massey
you
know
feels
that
this
is
still
too
harsh,
potentially
for
certain
applications
like
d5,
but
it
turns
out
that
even
these
partial
gas
refunds
have
potentially
really
really
strange
effects,
because,
basically
you
know
it
introduces
another.
C
You
know
another
set
of
actions
for
bidders
to
potentially
takes
that
might
screw
up
the
order
right,
because,
ideally,
what
we
want
is
that
the
priority
order
and
goes
by
how
much
you
are,
you
know,
willing
to
pay
per
unit
of
gas,
and
I
don't
just
mean
stated
willingness
to
pay.
I
mean
your
actual.
C
You
know
internal
valuations
that
is
known
to
you
right,
because
you
want
to
allocate
the
gas
to
the
people
who
have
the
best
use
for
it
right,
and
so
it
turns
out
that
there
is
that
there's
some
strategy
around
these
refunds
that
I
have
to
investigate
to
figure
out
if
that
might
potentially
ruin
this
order
in
equilibrium.
A
Right
right,
yeah,
I
mean
so.
The
interesting
thing
is
because
we
have
execution
before
consensus.
There
isn't
a
way
for
the
validators
to
actually
know
what's
going
to
be
profitable
for
them
versus
not
right,
it's
just
going
to
be
based
on
how
much
you're
willing
to
pay.
They
don't
have
any
idea
of
the
actual
amount
of
effort,
a
given
actual
amount
of
compute
a
given
contract
is
it's
literally
just
the
willingness
to
pay
if
they
had
to
if
they
wanted
to
order
transactions.
C
Yeah,
well,
it's
obvious
right
because
you
also
have
the
stated
gas
reservation,
but
the
point
is:
if
you
were
not
charging
for
those
entire
gas
reservations
and
yeah,
you
could
have
very
strange
outcomes
right
because
you
know
so
you
could
just
you
know,
say:
okay,
like
I
reserved
the
entire
block,
you
know,
and
I
you
know
put
in
very
high
willingness
to
pay
on
it.
But
then
I
run
you
know
like
five
units
of
gas
worth
of
compute,
and
so
now
you
know
nobody
else
gets
to
exit
right
right.
A
C
Yes,
and
again
I
mean
I
think
we
may
have
discusses
before
in
these
community
calls
I
mean
we
do
intend
to.
C
You
know,
support
you
know
estimation
of
gas
usage
right,
I
mean
we,
you
know
in
fact
there's
like
really
no
option
for
us,
but
not
to
do
so
right,
because
we
do
need
like
a
working
test
net,
and
so
you
know
the
test
net
you're
going
to
go,
and
you
know,
deploy
your
assemblage
of
contracts
and
you're
going
to
run
your
workloads,
and
you
know,
estimate
how
much
you
know
how
much
gas
you're
going
to
be
using.
A
A
C
Yeah,
because
you
are
submitting
bids,
you
are
submitting
bids
in
tokens
currently
well,
I
mean
so
not
currently
but
like
in
this
cp
right
you're
submitting
your
bids
and
tokens,
and
you
know
basically,
the
concept
of
you
know.
A
gas
price
in
terms
of
modes
in
this
specification
corresponds
to
the
last
the
willingness
to
pay
of
the
last
guy.
In
the
last
block
you
observed
right.
That
is
the
current
going
market
price
of
gas.
C
Because
that's
you
know
if
everything
is
stable,
that's
what
you
have
to
pay
in
order
to
get
into
a
block
and
the
anything
that
you
pay
above.
That
is
just
you
know.
Basically,
a
defense
against
uncertainty
of
you
know
what
other
you
know.
People
are
arriving
to
billions
of
gas.
C
Essentially,
yes,
right
I
mean
I
again,
I
don't
want
to
call
it
the
gas.
You
know
exchange
rate
really
right,
because
what
you're
passing
in
passing
the
villain
to
pay?
That's
what
you're
passing
it
yeah
right
and
then
you
know
the
gas
exchange
rate.
Let's
say
it's
not
really
a
things
that
exists
in
the
code.
It's
just
it's
an
aggregate
statistic
about
the
behavior
of
the
system,
yeah.
A
C
Well,
I
mean
it
should
perform.
I
think
it's
a.
It
is
very
similar
to
ethereum.
I
mean
there
may
be
some
implementation
differences,
but
the
id
is
exactly
the
same.
A
C
A
C
A
Yeah
yeah
yep,
so
our
focus
between
now
and
launch
really
is
around
performance
hardening
hardening
and
stabilization
crash
recovery
and
upgrades
right
we're
getting
the
account
hashes
fixed
because
we
need
that
for
custody
solution,
but
that's
basically
it
so
over
the
next
six
to
eight
weeks.
That's
what
our
big
push
is
to
get
upgrades
working,
so
we
can
do
protocol
forks
and
the
upgrades
and
fast
synchronization
feature
associated
with
upgrades
will
enable
us
to
do
everything
else
that
we
need
to
do
post
main
net
launch.
A
We
do
fully
intend
on
putting
out
plenty
of
protocol
forks
once
this
product
launches
once
the
main
net
launches.
We
feel
that
it's
our
responsibility
to
maintain
the
system
until
we
have
delivered
on
all
the
promises
we've
made
around
the
feature
and
functionality
system,
so
big
features.
We're
planning.
Post-Mainnet
launch
is
the
gas
futures
market
which
delivers
on
the
promise
of
price
stability.
We
also
plan
on
implementing
features
where
we'll
do
the
account
and
contracts
unification,
simplification
that
will
enable
contracts
to
pay
for
themselves.
A
That
will
be
a
very
large
protocol
fork,
so
we'll
probably
want
to
get
that
done
shortly
after
release.
So
big
exciting
features
coming
even
after
mainnet
launch
and,
of
course,
we're
going
to
start
on
our
sharding
research
right
after
mainnet
as
well
so
good
stuff.
A
Anything
else
you
guys
would
like
to
talk
about
on
the
marketing
team
is,
I
had
to
push
back
one
more
week
because
matt
had
to
get
go,
get
a
covet
test
so
and
he
could
only
get
a
slot
during
this
time,
so
he
plans
on
providing
that
update
next
week.
So
apologies
for
the
delay
community,
we'll
get
you
the
update
on
the
marketing
and
launch
plan.
A
Terrific
short
update
today
merry
christmas
to
everyone
happy
new
year.
I
will
be
here
next
week
to
give
the
update.
So
I
wish
you
guys
happy
new
year
then,
but
have
a
safe
and
wonderful
holiday
season
and
be
safe
out.
There.