►
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,
everyone
from
wherever
you're
coming
to
us
today
my
name
is
metta
parlikar.
I
am
the
cto
of
casper
labs,
and
this
morning
this
lovely
day,
I'm
going
to
give
you
an
update
on
our
engineering
progress
on
the
project.
The
community
has
requested
a
road
map,
update
ashok,
so
we'll
be
providing
we'll
do
our
best
to
provide
a
status
on
where
we
are
with
the
roadmap.
A
A
I'm
doing
this
that'd
be
great.
Thank
you
all
right,
so
we
are
in
now
in
sprint.
6.3,
we
are
working
on
the
20.10
release
cycle.
We
did
meet
our
september
milestone.
We
had
a
september
milestone
of
launching
the
charlie
test
net.
We
did
launch
the
charlie
test
net.
It
was
charlie
peaked.
His
head
up
said
hello
to
the
world
and
hung
around
for
about
five
days
before
the
logs
filled
up
a
lot
of
the
disc,
and
we
ran
into
some
issues
with
stability,
but
these
were
not
issues
that
we
didn't
know
about
already.
A
A
I
believe
joe
has
configured
and
fine-tuned
some
of
the
logging,
so
it
should
be
a
little
bit
more
stable.
We
also
made
a
mistake
when
we
launched
charlie
tested
in
terms
of
how
we
configured
the
fault
tolerance
threshold,
so
we've
learned
from
that
and
we're
getting
ready
to
launch
another
round
this
morning.
Actually,
both
the
test
nets
will
be
balanced.
This
morning.
The
beta
test
net,
I
think,
kicks
off
at
11,
am
pacific
and
the
charlie
test
net
will
kick
off
at
10
o'clock.
A
If
you're
interested
in
participating
in
the
test
net
program,
please
do
sign
up
if
you've
signed
up
before,
with
keys
from
clarity,
there's
probably
an
issue
with
your
keys
ashok.
We
should
make
it
so
folks
can
update
their
keys
on
the
google
sign
up
form.
I
think
we
only
allow
we
don't
allow
people
to
change
their
submission,
but
I
think
in
hindsight
we
should
probably
allow
them
to
fix
their
keys.
There's
new
instructions
in
the
in
the
google
doc
for
you
to
update
your
keys.
A
So
if
you
had
an
issue
with
keys
before
you
can
probably
get
new
keys
worked
out
that
way,
beta
testnet
is
going
to
launch.
Today,
like
I
said,
we
had
an
error
in
updating
the
security
group
and
we
ran
a
test
on
what
it
would
look
like
to
have
a
network
partition.
So
we
now
know
what
happens
in
the
event
of
a
network.
Partition
we'll
be
re-running
this
test
with
charlie,
to
make
sure
that
the
protocol
can
heal
relatively
easy.
A
The
charlie
testnet
immediately
upon
launching
a
couple
of
the
validators
got
together
and
they
scripted
up
a
token
transfer
test.
I
was
very
pleased
with
the
results
they
were
able
to
process
a
rather
large
bulk
of
transactions
through
consensus
in
a
very
short
amount
of
time,
so
it
was
somewhere
in
the
order
of
2500
transactions
that
were
processed.
You
know,
in
about
a
minute,
it
took
very
very
little
time
to
process
all
those
transactions
which
is
a
nice
nice
performance
rate.
A
Considering
we
haven't
done
any
kind
of
optimization
performance
tuning
of
the
network
at
all.
Our
current
focus
and
highway
validator
set
rotation
slashing
and
then
implementation
of
the
core
security
features.
These
are
defenses
against
equivocation-based
spam
attacks.
These
are
the
attacks
that
we
discovered
in
our
research
that
all
dag
based
protocols
are
susceptible
to.
We
are
making
updates
in
the
highway
paper
to
outline
to
chronicle
these
attacks
and
our
our
mitigation
against
these
attacks.
A
So
there
is
an
update
to
the
highway
paper,
cardinal
cryptography,
whom
we're
partnering
with
put
in
a
pull
request
that
was
merged
recently.
So
there's
a
new
version.
A
Validator
set
rotation
also
ties
in
with
bonding
auctions,
so
there's
also
work
being
done
in
bonding
auctions
that
we
expect
to
deliver
in
the
next
milestone.
So
by
the
october
release,
we
actually
fully
have
implemented
validator
set
rotation,
bonding
auctions,
and
you
know
basically
the
entire
functionality
to
make
the
network
permissionless
on
the
node
rust
side.
This
is
where
the
real
bulk
of
the
heavy
lift
is
happening
with
the
rust
implementation.
A
We've
got
our
event
stream,
so
there's
going
to
be
a
slightly
different
architecture.
You
wouldn't
you
wouldn't
want
to
necessarily
query
the
nodes
directly,
but
the
nodes
will
offload
events
in
a
stream
that
can
then
be
put
into
a
kafka
type
interface.
A
We
are
expecting
that
some
validators
will
offer
this
data
service
data
as
a
service
from
the
blockchain.
That
will
be
a
separate
service
from
the
validation
service
itself.
We
don't
want
our
nodes
to
be
burdened
by.
We
don't
want
the
nodes
to
be
burdened
by
having
the
you
know,
providing
data
services.
A
All
right
so
we're
going
to
be
offloading
that
data
service
problem
over
to
a
kafka
data
service,
we're
enhancing
the
rust
client.
We
want
the
rust
client
to
be
feature
parity
in
the
sense
that
it
will
provide
all
the
necessary
functions
that
we
were
used
to
with
the
scala
and
python
client
right
on
the
scala
node,
we're
implementing
economics
of
the
blockchain,
so
you'll
see
here
in
highway
they're
implementing
slashing
we
can
already
detect
equivocations
and
now
we're
going
to
slash
for
equivocations
and
we're
implementing,
rewards
distribution
right.
A
A
So
rewards
and
transaction
fees
are
different,
rewards
come
from
inflation
of
the
token
supply.
Transaction
fees,
of
course,
come
from
transactions
fees
that
are
paid
for
by
users
and
right
now,
the
transaction
fees
just
pool
in
the
proof
of
stake
contract,
and
so
what
we're
going
to
be
doing
is
distributing
those
to.
A
Validators
ability
to
fetch
linear
chain
blocks
by
height.
This
is
this,
is
a
client
feature
right
or
it's
client
and
node
feature.
A
C
No,
I
think
this
is
this
is
more
related
to
no
joining
up
and
rejoining.
C
C
A
A
Sorry,
my
bad.
A
So
here
we
go
yeah,
so
yes,
so
we're
ability
to
fetch
the
linear
chain
blocks
by
height.
This
is
for
no
joining
rejoining.
We
can
do
that
right
now,
but
we
can,
but
we
can
also
fetch
dependencies.
So
you
have
to
remember
that
the
network
messaging
message
passing
is
asynchronous
right.
You
don't
you
can't
expect
that
you're
going
to
get
the
message
linearly.
That
would
be
very,
not
very
efficient,
so
you'll
get
block
information,
kind
of
randomly
and
so
being
able
to
fetch
a
its
dependence
right.
A
If
I
have
blocked
number
10-
and
I
only
have
block
number
six-
I
want
to
get
seven
eight
nine
that
that's
currently
a
pull
request.
I
believe
that's
in
review,
node
rs,
374,
yeah,
test
yup
test
and
sre
rust
network
launch
with
five
validators.
I
think
that's
looking
more
like
set
the
total
number
of
nodes,
it
was
about
eight
or
nine.
My
node
was
there.
A
We
had
three
of
our
nodes
and
we've
added
a
couple
more
validators,
so
we're
slowly
increasing
the
network
size,
we're
also
testing
relaunching
beta,
we're
doing
lrt
testing
of
rust
node.
So
I
instructed
archit
to
put
on
his
chaos
monkey
hat,
so
I
really
do
want
him
banging
on
this
thing.
We
also
welcome
our
validators
to
use
chaos,
monkey
type
thinking
with
the
rust
node
and
also
attacking
it.
We're
also
deploying
our
brand
new
blog
presently
on
our
website.
A
You
see
our
medium
link
to
our
blog,
but
we're
going
to
have
our
own
blog,
so
you
can
start
hearing
from
our
developers.
Our
devs
have
a
lot
of
really
cool
things
to
say
and
they're
really
smart
guys,
and
so
we've
implemented
a
github
workflow
that
will
enable
them
to
create
blog
posts
by
github,
so
you
guys
can
hear
from
them
directly.
A
It'll
be
good
if
you
want
to
pop
in
on
discord,
and
you
want
them
to
talk
about
a
particular
topic
by
in
all
means
feel
free
to
tag
them
in
discord.
If
you
go
to
our
discord,
you'll
see
that
the
core
devs
there
are
different
colors,
so
they're
right
here.
These
are
the
core
devs
on
the
team,
so
you
can
certainly
ask
them
if
you
want
to
hear
learn
about
something,
some
of
them
stay
in
hidden
mode,
but
you
can
still
tag
them
in
a
message
so
you'll
see
them
here,
there'll
be
yellow.
A
This
is
alex
leminov
who's
on
the
call
here.
Abner
who's,
not
on
the
call
arko
life
is
one
of
our
core
devs,
so
the
core
devs
are
there
you
just
gotta.
You
can
definitely
tag
them
if
you
wanna
hear
from
them,
so
that
you
know
can
get
them
out
of
their
shell,
calling
them
out
ecosystem.
So
we
got
to
get
clarity
up
and
running
for
the
rust
network
and
we're
going
to
make
some
other
ui
ux
improvements.
A
The
rust
network
is
slightly
different
than
the
scala
network,
so
the
the
clarity
ui
is
going
to
be
slightly
different
and
we're
going
to
look
to
improve
that
as
well.
We're
designing
payments
2.0.
These
are
some
important
features
that
make
the
casper
blockchain
more
friendly
for
financial
and
economic
transactions.
Of
course,
that
is
hugely
important.
A
B
A
We're
also
going
to
start
doing
some
research
on
gas
futures
and
we're
going
to
write
some
specifications
on
how
that's
going
to
work,
we're
going
to
implement
gas
futures
as
a
way
to
get
us
real
stability.
The
you
know,
ed255,
not
ed25519,
but
it's
the
eip.
It's
an
ethereum
improvement
proposal
for
gas
stabilization.
A
A
That
basically
gives
the
user
no
benefit
of
downside
and
only
upside
risk,
and
while
it's
probably
better
than
what's
happening
on
ethereum
now,
it's
not
what
we
promised
our
customers,
so
we're
going
to
implement
something
closer
to
gas
futures
which
will
enable
a
customer
to
purchase
space
in
future
blocks
and
we're
going
to
try
to
enforce
it
in
protocol,
and
so
this
is
something
that
we're
doing
some
research
on
now.
A
The
scala
node
is
in
code.
Freeze
we're
not
doing
any
new
work
on
the
scala
node,
because
the
rust
node
is
around
the
corner,
and
the
intention
is
that
by
charlie
testnet,
we're
going
to
upgrade
all
the
current
validators
and
more
onto
the
charlie
and
there's
the
charlie.
Basically,
the
rust
node
for
the
next
delta
test
net
we're
going
to
restart
some
work
on
contract
runtime.
Once
we
have
our
contract
account
and
contract
unification,
spec
and
payments
2.0
specified.
A
The
developer
trainer
and
advocate
did
not
join
the
team,
so
ashok
we've
got
to
update
that.
He
was
supposed
to
join
this
week,
but
he
didn't
join.
So
we
are
looking
for
someone
that
wants
to
be
a
dev
trainer
and
advocate
to
join
our
team
and
we're
searching
for
talented
sres
to
join
our
team
and
are
we
hosting
a
weekly
workshop
this
week?
Do
we
know.
A
We
are
okay,
fantastic
we've
adopted
a
casper
enhancement
proposal
process
to
vet
and
accept
important
feature
updates
right
now.
The
core
dev
team
is
working
on
implementing
that
they're.
Basically
using
the
cep
process,
there
is
the
validator
association
cep
that
probably
is
good
for
folks
to
take
a
read
at.
I
know.
I
certainly
owe
it
a
review
amongst
others.
So
definitely
check
that
out.
That's
the
first
portion
of
our
cep
process.
Our
grant
process,
hopefully
will
come
via
the
dev
dao.
A
I
don't
see
members
of
the
dev
down
on
this
call
today
they
usually
on
here
they
must
be
busy
but
yeah.
The
dev
dao
is
going
to
be
the
ones
really
supporting
supporting
us
in
terms
of
grants
and
that
grant
was
approved
by
the
board
of
directors
last
week.
So
they
are
in
business.
C
Yeah,
and
and
and
before
just
to
clarify
the
the
ticket
that
you
were
talking
about
me
that
fetching
the
block
from
the
linear
chain
by
height.
This
is
basically
work
being
done
on
the
storage
component,
so
that
we
support
ability
to
fetch
block
by
specifying
its
height.
It
requires
adding
additional
key
value
store
to
the
storage
component.
A
Got
it?
I
understand
yes,
because
in
the
scala
node
that
information
came
from
a
different
block
store.
So
when
a
coach
built
the
scala
node,
he
worked
around
it
that
way,
and
so
now
we
don't
have
it
in
the
rust
side.
So
this
is
something
that
knew
that
when
we
moved
away
from
the
scala
implementation
we
had
to
implement
it
got
it.
C
A
A
C
So
this
this
is
the
this
is
a
road
map
for
for
our
march
to
the
main
net.
What
we
have
here
we
are
in
september
now,
and
we
are,
we
are
on
track
so
far.
We
have
met
all
the
goals
that
that
we
we
have
been
planning
so
september.
We
have
launched
a
charlie
net
with
with
rust
a
hand-picked.
You
know
five
validators.
C
C
As
well
as
slashing,
we
should
be
able
to
support
more
than
50
validators
here.
A
minimum
of
50
is
what
we
are
gaming
for,
and
this
this
essentially
yeah.
It
is
a
big
milestone
for
us
in
terms
of
our
roadmap.
C
Implementation
of
payments
2.0
and
subsequent
month,
plus,
is
what
has
been
going
to
be
used
for
hardening
of
our
software
and
making
sure
that
you
know
we
test
it
out
with
all
different
possibilities
of
attack
scenarios,
we're
going
to
go
through
a
security
review
in
november,
so
the
so
the
work
after
in
in
and
after
november
is
more
towards
just
making
our
system
more
robust
to
be
able
to
face
face
the
real
world.
You
know
situations
and
attacks
and
scenarios.
C
C
A
Once
we're
pencils
down
in
november
and
we're
code
complete,
then
we're
just
going
to
be
focused
on
performance
bug,
fixes
hardening
and
seeing
what
test
net
and
security
review
yield
right.
So
we
are
opening
the
delta
test
net
up
to
attacks
we're
encouraging
people
to
attack
the
network.
A
We
want
this
to
be
secure
and
stable
and
and
permissionless,
and
so
you
know
making
sure
that,
above
all
else
that
the
protocol
is
correct,
that
it
is,
it
passes,
security
review
that
the
paper
clears
peer
review
is
the
most
you
know
is
the
most
important
piece
for
us,
and
so
that's
going
to
be
our
focus
once
we
enter
security.
C
Review
right
and
then
we
we
have.
We
have
a
process
here
where
we
plan
out
each
month
we
have
monthly
release
cycles
and
weekly
sprints,
so
we
plan
out
our
monthly
release
at
the
beginning
of
entering
entering
the
release
cycle
and
then
all
the
features
that
we
want
to
deploy
during
work
on
during
that
those
four
sprints,
because
it's
a
weekly
sprint
for
next
one
month
are
laid
out
and
we
track
it
very
closely.
On
jira.
A
Yeah,
why
don't
we
give
them
a
peek
inside
of
the
there?
We
go
the
at
the
kanban
board.
C
C
There's
just
one
ticket,
it's
already
in
review
and
charlie
code
actually
already
takes
cherry
picked
it.
So
then,
the
current
release
that
we
are
working
on
is
is
here
and
we
track
our
progress
here.
C
We
already
have
started
planning
what
we're
going
to
do
in
the
next
next
release
and
all
that
that
is
required
for
mainnet
is
also
being
tracked
here.
So
we
know
these
are
all
things
that
we
need
before
I
need
to
complete
before
we
go
to
magnet.
A
That's
fantastic,
can
you
show
them
the
reports
too?
Please.
The
reports
is
really
cool
on
the
left-hand
tab.
There
whoops.
A
On
the
left-hand
tab
under
features
board,
there's
reports
and
let's
just
show
them
node
rs-
go
up,
go
down.
Oh
you
were
right,
there
yeah
yeah.
There
you
go
and
then
let's
go
down
to
the
version
report
on
the
right
hand,
side
there
and
we'll
show
them.
Node
rs
for
2009
and
not
2010,
show
them
2010.
A
So,
as
I
indicated
like
one
of
the
really
big
portions
of
this
work
is
the
node
right
and
so
the
note
up,
yeah
there
you
go
and
the
work
that's
being
done
in
the
node
right
now
is
a
line
share
the
work.
So
here's
here's
how
we
as
project
managers,
program
managers,
really
track
the
release
really
carefully.
A
So
we
make
sure
that
all
of
the
issues
you
know
have
estimates
and
the
team
has
become
quite
good
in
terms
of
their
estimation
accuracy
over
time,
because
we've
got
like
four
months
right
and
jira
is
really
good
at
a
little
bit
of
a
jira
plug,
but
it's
really
good
at
giving
you
these
forecasts,
so
ashok
tracks
this
very
carefully.
So
if
you
go
to
the
very
top
here,
the
release
date
is
october,
8th
here
and
right
now
we
can
see
we're
like
running
a
little
bit
behind
a
couple
weeks
behind
the
goal.
A
So
we
know
we
need
to
button
this
up
and
tighten
it
down
and
potentially
descope
non-essential
features
for
the
release
or
we
need
to
limit.
You
know
we
basically
need
to
inform
the
team
that
they're
not
on
track
right.
So
if
you,
if
you
scroll
up
you'll,
see
here
that
it's
looking
like
the
optimistic
is
october
18th
and
then
the
pessimistic
is
november
1st,
and
if
you
scroll
down,
you
can
see
all
the
tickets
that
are
in
there.
The
port,
you
know
so
predicted,
is
26
and
so
find
that
this
is
pretty
accurate.
A
Scroll
down,
you
can
see.
All
the
tickets
that
are
completed
shows
the
tickets
that
are
completed
and
the
ones
that
are
in
progress
or
not
completed
right
and
the
amount
of
stories
associated
with
it,
and
then
it
also
gives
you
incomplete
unestimated
issues.
So
this
probably
will
be
the
piece
that
will
get
de-scoped
because
it's
not
been
estimated
yet
right.
We
probably
won't
do
this
in
october.
A
B
Well,
it's
out
of
cycle
for
me
technically,
I
should
be
sharing
next
week,
yeah
so
next
week.
Presumably
I'll
do
another
pass
on
what
we're
doing
this
chain
link
and
specifically,
is
there
gonna,
be
some
changes
relative
to
what
was
previously
presented,
but
these
changes
make
it
easier
for
us
to
actually
use
it.