►
From YouTube: Ethereum Core Devs Meeting #58 [2019-03-29]
Description
A
A
All
right
good
morning,
everybody
welcome
to
Cornell's
meeting
number
58
I
just
updated
the
agenda
like
a
second
ago
with
Martin's,
EIP
and
I.
Think
that's
it
so
go
ahead
and
refresh
on
the
agenda.
If
you
haven't
already,
let's
start
with
the
roadmap,
I,
don't
think
there's
any
updates
on
that.
But
I
did
add
a
link
to
the
etherium
wiki.
So
you
all
can
keep
up
with
some
of
the
updates
and
proposals
and
the
timelines.
B
C
A
It
so
so
what
we
had
before
was
what
offer
he
came
up
with
as
far
as
the
timeline
goes,
that
we
all
agreed
to
at
the
time,
but
I'm
happy
to
change
it
with
suggestions
right
now.
What
we
have
is
May
17th
a
hard
deadline
to
accept
proposals
such
as
a
IPS,
and
then
we
have
a
soft
deadline
of
the
19th
of
August.
No
wait
July
to
have
major
client
implementations,
it's
under
the
timelines
section
on
the
link
on
there.
C
And
that
means
that
if
we
keep
doing
things
this
way,
we
would
not
be
able
to
do
smaller,
hard
Forks,
because
for
the
smaller
hard
Forks
it
there's
needs
to
be
some
kind
of
pipeline
going
where
there's
multiple
kind
of
hard
works
in
the
in
the
process
of
being
prepared,
and
that
includes
the
testing.
So
you
know
I
would
like
to
get
some
comments,
maybe
from
Martin,
because
he
recently
responded
in
that
thread.
D
The
hard
folks
we
have
done
historically
are
very
intensive
on
the
test
section,
because
the
eeap's
that
are
going
in
they
can
have
into
you
know
related
effects.
So
as
we
change
something
and
finalize
stuff,
we
might
have
to
redo
earlier
test
cases
regenerate
a
lot
of
test
cases,
and
this
is
yeah
so
I'm
thinking
that
it
might
be
easier
and
more
spread-out
work
if
the
hard
forks
are
smaller
and
more
spread
out.
That
means
it's
not
that
big
of
a
catch
effect
on
their
testers
and
then.
C
D
C
Either
the
question
for
you,
Dmitri
and
Marcin-
is
that
so
the
in
terms
of
the
time
is
spent
on
preparing
a
test.
Where
is
the
biggest
time
span?
Is
it
the
human
labor
or
is
the
test
regeneration
or
what
is
the
kind
of
the
biggest
component
that
we
would
need
to
kind
of
paralyzed
and
push
it
onto
each
sort
of
each
parallel
line
of
the
our
pipeline
processing
thing
I.
E
Think
the
biggest
time
is
to
actually
write
down
the
best
case
in
a
JSON
format,
and
sometimes
it's
there's
about
hundreds
of
test
cases,
and
that
would
be
good
if
those
who
propose
EAP
also
propose
some
test
scenarios.
So
I
could
just
so.
We
could
have
process
of
converting
those
scenarios
into
JSON.
C
E
C
G
Totally
agree
so
I
think
it's
unfeasible
to
have
one
person
so
means
we
currently
have
a
truck
factor
of
one
if
Demetri
gets
run
over
by
a
truck.
We
don't
have
any
other
knowledge
within
the
system
of
how
to
do
testing
so
I
think.
A
really
good
takeaway
is,
let's
rally
some
people
to
have
more
humans
working
on
testing
actually.
D
H
It's
tricky,
there's
been,
there's
been
a
parallel
thread
in
the
fellowship
forum
about
just
generally
setting
higher
standards
for
your
IP
authors
and
I.
Do
think
that,
like
putting
some
of
burden
on
authors
or
submitters
to
like
include
tests
or
even
just
the
types
of
tests
they
want
would
be
a
good
thing.
I
think
better
tooling
could
make
that
easier.
D
Simple
shift
operations,
those
test
cases
can
be
defined
pretty
well
or
if
it's
a
gas
repricing
there's
not
much.
Those
test
cases
are
very
quick
to
write,
but
it
was
something
more
intricate
like
the
s
s
story,
pricing,
where
we
have
this
other
extra,
weird
edge
cases
where,
if
you
enter
back
in
the
same
thing
and
you
do
it
three
times
in
a
row
and
munis
and
half
mode,
those
are
really
complicated
to
write
and
they
take
a
lot
of
time
to
write.
But
it's
this
big
variance,
but.
A
Think
that
the
way
we
do,
that
is
to
change
a
IP
one,
to
make
it
a
requirement
to
have
some
test
cases
for
Cori
IPS.
The
IP
one
needs
a
complete
overhaul,
as
it
is
so
maybe
setting
a
deadline
for
VIP
one
changes
over
the
next
month
or
whatever,
and
then
get
a
lot
of
people
working
on
a
single
PR,
for
it
would
maybe
be
the
most
beneficial
but
I'm
ready
to
hear
other
opinion.
G
H
A
H
A
Okay,
well
we'll
we'll
communicate
and
the
and
the
chat
like
I'm
telegram
or
dinner,
or
something
and
keep
up
with
that.
I
have
that
as
a
note
here
and
I'm
there's
already
a
note
taker
today,
but
I'm
taking
personal
notes
for
basically
to
dues
for
next
time
and
then
we
will
come
up
with,
like
maybe
in
the
notes.
They'll
be
like
a
you
know,
what's
been
accomplished,
what
needs
to
be
accomplished
to
dues
for
next
time
kind
of
a
thing,
so
we
can
keep
up
with
that
that
one
suggestion
from
Gregg
thank
you,
Greg
I.
H
H
Question
is
just
so.
This
has
come
up
a
few
times
now
in
conversation,
this
May
17
dad
line
I'm
just
reading
here
it
says
hard
deadlines
to
accept
proposals
for
Istanbul.
Did
it
like?
Is
there
any
requirement
on
how
mature
those
GIPS
are
at
that
point?
In
time
like?
Could
someone
hypothetically,
like
just
file
a
placeholder
AIP
it'd,
be
good
to
have
some
consensus
on
this.
C
C
Because
I
don't
think
it's
a
good
idea
to
just
put
the
placeholder.
So
I
think
the
eyepiece
needs
to
be
worked
out
pretty
well
before
and
and
then
the
problem
with
this
deadline.
That
now,
with
the
plan
of
having
our
hard
works
every
nine
months-
and
these
deadlines
is
that
now
obviously,
if
I
miss
the
deadline,
I
have
to
wait
another
year
or
something
to
pour
it
something
to
happen.
C
C
Well,
here,
I,
don't
I'm
not
sure
about
what
to
go
into
the
specifics
of
ERP
process,
because
I'm
not
very
well
I,
know
I,
haven't
thought
it
very
well.
But
for
me
it's
not
really
like
yeah,
it's
something
which
you
propose
to
be
definitely
proposed
to
be
included
into
the
hard
work,
whether
it's
a
IP
or
like
fully
specified
or
I,
don't
know
how
what
is
it
so?
G
C
It's
fine
to
have
it
proposed
like
written
and
proposed
and
merged,
and
that's
that
that's
fine,
but
the
eeap's
that
are
kind
of
slated
for
the
inclusion
into
the
heart.
Worksheet
have
a
certain
and
very
high
requirements
in
terms
of
readiness.
So
we
can
create
hundreds
of
each
and
then
get
the
merged,
but
it
doesn't
mean
that
we're
gonna
find
them
in
for
the
hard
floors.
Absolutely
all.
G
Battling
to
e2
3/3,
which
basically
has
a
proposed
session
so
that
people
know
that
they're
signalling
that
they're
gonna
try
and
get
this
EEP
in,
but
it
has
to
move
through
other
sections
and
that
may
include
tests
and
everything
else
to
actually
get
in.
But
I
think
that
the
sooner
you
get
in
a
merge
deep
that
isn't
just
filled
with
to
dues.
That's
a
useful
step
to
signal
that
you're
serious
about
engaged
with
the
process.
G
D
I
would
I
would
propose
that
we
changed
the
process
insofar
as
not
have
it
hard
for
eccentric,
but
have
it
EEP
centric.
So
that
say,
the
first
step
is
that
they
need
get
accepted,
and
the
second
step
is
that
it's
implemented
in
just
parity,
and
the
third
step
is
that
there
are
tests
that
fully
cover
it
and
everyone's
helped
with
those
tests
and
the
clients
pass
it
and
they
step
up
with
that
is
okay.
It
passes
all
these
three
or
four
or
five
steps.
So
when
can
we
include
it?
I
Kind
of
drafted
a
process
and
I'm
trying
to
propose
that
I'm
from
Canada,
so
I'm
doing
it.
From
that
perspective,
what
we
can
do
is
light
curve.
We
can
suggest
EEP
order
to
send
a
man
open
an
issue,
each
saying
that
each
should
be
considered
for
next
hard
food
for
this
hard
work.
If
we
are
going
for
Istanbul,
we
can
do
that
and
I
have
created
a
tracker
at
github
of
ezh,
where
we
are
listing
all
the
eats
it's.
It
states
that
these
eats
are
at
proposed
status,
and
then
it
goes
for
the.
I
Then
we
bring
it
to
the
coda
discussion.
If
it
is
accepted,
we
will
change
the
status
and
and
eventually
we
will
take
it
to
client
implementation
test
nets
and
finally
included
in
there
and
I'm.
Also
referring
to
the
tracker
that
Boris
have
created
about
the
client
implementation,
we
can
go
ahead
and
do
that
so
I'm
just
yeah.
Please.
D
Sorry,
so
what
I
my
alternative
is
that
would
be
to
not
focus
on
like
xanthium
is
next
one
and
which
one
are
much
enough,
but
to
instead
focus
on
one
by
one
like
oh
this
week,
it
looks
like
this.
Eep
is
now
finished,
it's
implemented
everywhere.
This
has
tests
for
everything.
It
is
well
written
everything,
so
we'll
just
schedule
that
EEP
for
one
month
from
that
time,
yeah.
J
A
H
I
very
strongly
agree
with
Martin
and
I
like
to
say
here
that
smaller
hard
Forks
in
the
nation,
with
a
more
AIP
centric
process,
I
think
it's
pretty
clear,
would
produce
higher
quality,
VIPs
I'm
gonna,
move
that
we
table
this
conversation
and
take
it
to
the
fellowship
of
athira
magicians
forum
and
like
work
out
some
more
concrete
proposals
there
and
not
take
up
more
awkward.
That's
bandwidth
on
this
right
now.
A
A
If
he
comes
later,
we'll
just
have
him
go
back
to
this,
but
the
one
that
I
included
here
was
the
IP
1599.
If
the
market
change
for
the
eath
1.0
chain,
there's
a
link
to
that.
If
people
want
to
check
it
out,
the
next
one
is
one
that
I
think
Danno
did,
which
is
early,
which
is
the
etherium
network.
Upgrade
windows
and
I
think,
that's
kind
of
related
to
what
we
just
talked
about.
But
just
a
little
bit
more
defined
is
that
right,
Dan,
oh
yeah,.
K
The
gist
of
it
is
that
we
would
set
aside
four
weeks
out
of
the
year
that
we
would
target
launching
our
network
upgrades
in
and
Friz
make
a
discussion.
I
picked
the
third
week
of
January
April,
July
and
October
to
start
the
discussion
and
they're
right
there,
each
four
weeks
apart
they
dodge
August
and
December
and
fourth
of
July
for
the
US.
You
know
big
vacation
times.
People
want
to
take
so
on
that
subject.
One
of
the
big
advantages
is:
it
provides
engineers
a
calendar
that
they
can
take.
K
A
vacation
I'll
have
to
worry
about
a
hard
fork
being
plucked
in
the
middle
of
it.
You
know
when
they
plan
their
stuff
in
advance
and
also
for
network
operators
node
operators
when
they
know
they
need
to
do
upgrades.
They
also
don't
need
at
the
same
time,
worried
about
and
network
upgrade.
Unless
it's
in
the
window,
so
some
of
the
details
of
it
I
split
the
network
upgrades
into
three
broad
categories,
just
for
the
sake
of
convenience.
First
ones,
roadmap,
which
is
once
we've
always
done.
K
I
created
a
new
category
priority,
which
is
something
more
important
than
a
road
map
upgrade
but
less
important
than
the
third
category,
which
is
a
critical
network,
upgrade
that's
when
something's
broken
and
has
to
be
fixed
and
we
need
to
do
whatever
we
have
to
do
so.
The
three
hard
Forks
we've
done
before:
homestead
Byzantium,
Constantinople
movies
roadmap
upgrades
and
like
the
DAO
fork,
spurious
monkey
and
tangerine
muscle,
would
be
the
critical
upgrade
examples,
and
you
know
there's
some
standards
around
priorities.
You
know.
K
C
Yeah
I'm
completely
supporting
this.
It
would
be
very
good
in
combination
with
what
we
just
discussed
before,
because
it
sort
of
looks
at
the
problem
on
from
the
other
side,
from
the
side
of
people
who
actually
have
to
run
the
nodes,
and
they
need
to
certainty
about
when
these
and
grades
can
actually
happen.
L
Quite
different
opinion
about
that
I
mean
I,
don't
mean
I,
don't
like
the
idea.
What
I
think
we
try
to
put
a
lot
of
constraints
and
decisions,
how
the
process
should
work
at
the
same
time
and
we
haven't
proven
yet
we
can
keep
the
deadlines.
The
skill
for
Istanbul,
for
example,
so
I
would
like
white
with
applying
this
one.
Until
we
we
actually
have
a
smooth
one
for
the
first
time.
L
K
They
don't
process
work.
Studies
in
the
absence
of
deadlines
tends
to
result
in
a
longer
work
cycle
with
the
same
or
lower
quality.
So
you
know
putting
some
sort
of
certainty
and
it's
not
you
know
if
we
miss
a
window,
it's
only
three
months
and
that
gives
us
you
know,
there's
less
haggling
about
well.
Can
we
do
in
January?
Do
we
do
it
February?
What
if
we
could
sneak
an
image
in
Christmas
a
new
year
it?
K
It
removes
some
of
the
arguments
of
the
process,
so
there
is
there
isn't
a
theorem
thread
on
this,
since
that's
the
theme,
because
we're
taking
a
lot
of
time
on
these
I
would
encourage
everyone
with
an
opinion
on
it
to
come
in
express
it,
especially
if
you
want
to
change
the
dates
it's
still
not
in
draft
I.
Haven't
I
need
to
do
some
upgrades
on
it.
I
think
one
of
the
authors
put
some
comments
in
there
I'm
going
to
upgrade
it
but
yeah.
So
there's
there's
new
three
magicians
threads
mention
them.
So
if.
H
I'm
understanding
public
correctly
he's
not
saying
that
it's
a
bad
idea,
I
also
strongly
support
this
idea.
I
think
Pablo
is
kind
of
suggesting.
Let's
stick
with
the
plan,
we
have
the
deadlines
we
have
for
the
next
hard
fork
and
then
like
make
this
something
that
happens
after
that.
Assuming
that
hard
work
happens
smoothly
and
I
think
that
might
be
an
okay
idea.
Yeah.
C
H
A
G
M
A
Danny
next
up,
we
have
an
EIP
proposal
from
Martin.
It
has
really
pretty
graphs
on
it.
Oh
the
link
doesn't
work.
Let
me
try
to
find
the
right
link
and
post
it
in
here,
but
like
the
graphs
like
and
stuff
like
that,
put
PowerPoint
graphs
to
shame,
so
everyone
should
check
it
out
and
it
says
page
not
found
Martin.
Did
it
go
private.
D
D
Yeah,
but
the
thing
is,
this
focus
is
on
trying
to
make
the
operations
which
have
deteriorated
over
time,
put
them
back
in
balance
again
and
those
specific
ones
that
I've
in
for
our
s
load
and
balance,
and
there
are
two
versions
to
this.
In
both
of
them,
the
s
load
would
increase
by
a
factor
of
four
from
200
to
800
and
in
version
a
the
balance
would
be
repriced
to
700
from
400,
so
they
imperative,
with
Xcode
sighs
in
Xcode.
D
D
There
are
other
things
that
could
be
done
to
make
all
ol
op
codes
more
equivalent
equivalent
to
heavy,
but
that's
not
a
really
concern
for
this
eat,
because
the
the
main
goal
is
to
make
sure
that
the
things
which
are
deteriorating
and
might
deteriorate
in
a
bad
manner
that
we
fix
that
problem
for
now
and
maybe
how
to
fix
it
again
later,
but
that
all
depends
on
how.
How
would
how
this
state
size
what
happens
with
the
state
size
in
the
future,
with
all
the
one
of
the
X
initiatives?
D
A
A
C
C
So
basically,
you
probably
saw
some
of
you
saw
that
yesterday
publish
the
stateless
client
prototype,
which
is
also
outlining
so
essentially
collecting
data.
How
big
these
proofs
would
be
and
what
are
the
next
steps
and,
in
short,
the
next
steps
could
be
either
to
try
to
bring
those
sort
of
to
bring
the
the
bandwidth
requirement
for
the
by
using
some
more
statefulness
or
to
use
darks
so
that,
but
the
thing
is
I
really
wanted
to
get
some
questions
about.
Is
the
there's
something
that
come
out
of
very
sort
of
some
time
ago.
C
It's
called
probabilistic
estimation
of
contract
sizes,
essentially
one
of
the
issues
that,
in
the
real
estate,
rent
and
now
is
in
the
in
the
thinking
protocols.
Is
that
how
to
to
get
to
figure
out
the
size
of
the
contract,
meaning
how
many
items
just
the
contract
storage
have
and
so
for
the
state
rent
proposals.
C
That
would
mean
if
we
could
bring
forward.
So
this
is
in
number
four
on
the
in
my
post,
my
agenda.
So
there's
this
change.
J
in
the
latest
state
fee
proposal,
which
is
the
fixed
repayment
for
contract
storage-
and
this
is
the
basically.
This
is
the
requirement,
from
my
point
of
view,
to
increase
the
block
size
limit
because
it
will
stop
kind
of
acceleration
of
state
growth,
so
I'm,
hoping
that
this
probabilistic
estimation
of
contract
size
could
help
bring
this
change
forward
to
the
first
hard
work.
C
So
we
can
actually
do
the
sort
of
harmless
block
us
limit.
Increase
earlier,
and
so
basically
that
that
is
the
main
thing
everything
else
has
already
been
out
of
the
out
there
and
yes
and
after
the
number
six
as
well.
It's
I'm
also
planning
to
do
some
more
research
on
linear
storage
because
it
could
be
integrated
better
with
you
as
an
engine,
but
I'm
kind
of
I'm
still
waiting
for
some
further
updates
from
he
wasn't
team
before
I
kind
of
launched
into
this.
C
A
Okay,
awesome
so
there's
still
the
Istanbul
and
ëthe
1x
roadmap
planning
meeting,
which
is
gonna,
be
quite
different.
If
we
do
like
upgrades
that
are
more
frequent,
but
we'll
still
go
over
the
EIP
s
that
we're
wanting
and
whatever
the
next
hard
fork
is
that's
April,
17th
and
18th
Boris.
Is
there
any
updates
on
that
that
are
worth
mentioning.
G
We
are
confirmed
for
full
node.
Thank
you.
Gnosis
team
for
hosting
more
and
more
people
are
I
believe
using
that
event
and
the
deadlines
we
have
in
place
to
propose
oops
I
just
saw
one
from
Virgil.
A
ton
of
them
are
related
to
pre
compiles,
and
we
can
also
use
that
live
format
to
I
think
discuss
what
a
more
continuously
process
looks
like.
G
Is
Trent
is
helping
me
out
with
the
live
stream,
so
we'll
capture
all
of
the
formal
stuff,
so
we'll
have
things
to
refer
to.
I
know
that
Virgil
is
gonna,
be
in
Korea
and
Vitalik
wants
to
dial
in
so
Trent
and
I
are
gonna
work
on
timing
to
make
sure
that
we
don't
have
to
make
people
get
up
in
the
middle
of
the
night
if
they
want
to
present.
A
C
H
Okay,
yeah
just
one
more
quick
thing
if
we
do
end
up
adopting
this
kind
of
like
quarterly
release
cycle
or
doing
more
often
during
hard
Forks,
more
often,
and
having
a
more
concrete,
fixed
schedule
for
it,
I
think
it
would
be
much
easier
to
begin
planning
events
like
this
way
ahead
of
time
to
get
people
time
to
plan
to
be
there,
which
I
think
would
be
really
awesome.
Just
another
nice
kind
of
knock-on
benefit
of
doing
that.
A
G
So
that
the
etherium
wiki,
that's
in
the
schedule
right
at
the
top
roadmap,
I'm,
I
and
others-
are
keeping
that
up
to
date
and
with
two
three
three
that'll
ideally
be
in
github.
Really
the
deadline
is
bringing
people
out
of
the
woodwork,
so
I'm
basically
constantly
curating
this,
and
it
should
be
relatively
up
to
date,
but
ideally
it
will
move
to
github.
So
you
can
actually
track
a
feed.
A
H
A
H
A
Yes
good
point
and
as
far
as
the
problem,
how
audit
update
goes
for
some
reason:
people
don't
want
to
give
me
money
for
stuff
or
the
cat
herders
money.
I
should
say
because
we're
we've
been
going
around
talking
to
people
and
it's
not
really
working
very
well,
because
people
are
kind
of
skeptical
of
the
cost.
I
will
say:
Atlantic
crypto
gave
10,000
died,
which
was
really
nice
of
them.
We
do
have
a
get
coin,
set
up
that
I'll
link
to
and
we're
just
continuing
to
fundraise.
A
That's
the
only
update,
I'm,
confident
that
I'll
have
a
lot
more
info,
the
next
time
that
I
am
on
the
court,
have
call
and
then
I'm
gonna,
post
the
get
coin
in
the
YouTube
chat
and
and
Oh
someone
put
5,000
more
and
and
the
zoom
if
anyone's
interested
in
sharing
that
link
around
so
there's
the
link
and
I
think
that
it
am
I
missing
anything
cat,
herders.
Anybody.
A
O
A
O
A
O
A
Is
my
understanding
right
now
that
could
change
if
people
speak
up
within
the
core,
Deb's
or
loud
enough
within
the
community
that
the
core
devs
start
to
listen,
because
as
it
is
right
now,
the
core
devs
are
the
ones
who
are
coding
this
into
their
clients.
That
is
my
understanding.
I,
however,
am
NOT.
This
isn't
my
decision
I'm
just
saying
this
is
my
understanding
of.
What's
going
on.
O
H
Yeah
I'm
sorry
to
be
to
be
rude
here
and
butt
in
but
like
I,
think,
we've
sort
of
reached
consensus
and
I
think
you
may
agree
with
this.
That,
like
these
are
not
technical
questions
and
that
this
awkward
is,
is
probably
not
the
right
channel
for
this
stuff.
So
are
you?
Okay,
with
our
taking
this
back
to
git,
er
or
femm,
or
one
of
the
other
other
characters
or
something
not.
O
A
Guess
the
the
answer
that
I
have
in
my
opinion
is
that
if
the
carbon
vote
says
something
wildly
different
than
it
does
now
or
if,
for
some
reason,
there's
enough
people
within
this
call
or
outside
of
this
call
who
raise
enough
us
decisions
can
change.
It's
not
like.
We
make
the
decision
and
we're
going
to
be
stubborn
and
never
change
it.
I
would
say
that
right
now
we're
still.
M
When
we
get
more
information,
so
I
think
it's
not
unreasonable
that
if
we
get
more
information
over
time
that
it
is
no
change,
obviously,
because
a
very
technical
reason
like
if
there's
some
massive
bug-
and
we
came
in
forward
everything
at
CBC
ideas,
Dalton
rapport
with
them.
But
there's
also
like
a
lot
of
folks
in
the
community
that,
like
a
change
of
decision,
that's
not
unreasonable
that
we
keep
our
ears
to
the
ground.
M
A
Q
E
So
hi
everybody
I'm
glad
to
announce
to
you
that
I
finished
stabilizing
retested
application
on
blockchain
tests.
So
now
it
could
run
over
the
blockchain
tests,
including
the
general
state
test,
in
a
blushing
form
on
any
client
that
support
RPC
interface,
which
is
described
on
retested,
wiki,
page
and
I
also
have
instructions
for
you
to
use
docker
build,
so
you
could
come
when
retested
and
run
it
on
your
client.
If
you
support
this
interface,
see
how
it
works.
Your
client
also
I
was
thinking
about
this
Jenna
Jenna.
E
E
So
once
we
come
up
how
it
remains
a
port,
we
could
talk
about
Genesis
format,
for
every
button
for
every
client
and
the
bare
minimum
of
Genesis
is
described
to
be
50
to
80
page,
and
this
pair
minimum
is
required
to
run
blockchain
state
test,
so
we're
ready
to
switch
to
mp3
test
execution.
This
is
a
CBP
client,
not
test
generation.
Yet,
but
at
least
test
execution
is
already
possible.
Then
you
could
think
about
debugging
through
the
RPC
and
I
feel
no
more
blackbox
testing
they
use
Excel.
E
A
Ok,
usually
by
now
we
do
client
updates,
and
then
he
was
them
in
research
updates,
but
I
wanted
to
kind
of
get
a
quick
feel
of
the
room
on
if
people
think
these
are
valuable.
These
used
to
be
the
main
focus
of
the
core
dev
meetings,
because
there
was
a
lot
of
movement
on
implementing
the
IPS
and
some
of
the
clients
were
earlier
on
and
needed
to
talk
about
their
progress.
A
R
R
Oh
we're
doing
like
a
million
things
on
the
client,
otherwise
and
I'd
like
there's
not
enough
space
or
time
to
really
dig
into
those
details
either
so
I
like
that
that
doesn't
make
it
in
so
I
think
it
might
be
good
and
having
an
on-demand
might
be
good.
But,
like
me
personally,
I
don't
know
what
your
report
BOCES
are
I.
D
A
Cool
I'll
start
having
a
comment
in
there.
That
just
says,
please
add
a
comment
to
the
agenda.
If
you
have
any
updates
for
now,
we'll
keep
the
testing
updates,
because
I
feel
like
that's
pretty
important
personally,
but
if
someone
wants
to
get
rid
of
that
talk
to
me
privately
I
only
think
it's
important,
because
testing
is
something
that
happens
in
every
client
and
we
need
to
keep
up
with
what's
happening.
There
go
ahead.
A
M
A
H
Something
we've
seen
this
worked
really
well
in
the
e
azam
team.
Is
that
people
post
their
updates
to
the
agenda
in
text
format
and
then
only
we
only
like
on
the
actual
call.
We
only
discuss
questions
that
people
have.
Ideally,
everyone
would
have
a
chance
to
get
their
updates
in,
like
let's
say,
24-48
hours
before
the
call,
and
everyone
would
have
a
chance
to
review
them
before
the
call
that
definitely
saves
time.
Okay,.
A
C
The
finality
gadget
and
I
know
that
there
are
some
proposals
from
italic.
You
can
see
the
link
in
the
in
updater
that
posted
an
agenda,
but
I
would
like
to
explore
kind
of
various
ways
of
doing
it
like
because
vitalik
was
suggesting
to
do
it
by
let's
say
BLS
pre-compile,
but
maybe
it's
a
practical.
Maybe
it
could
be
done
in
native
code
and
things
like
this,
but
I
wanted
to
get
this
work
started
somehow
I
mean
I,
don't
necessarily
mean
that
I
will
be
leading
it,
but
I
want
to
kind
of
launch
it.
C
M
Have
a
conversation
yeah,
and
this
item,
at
least
from
my
understanding,
seems
to
be
a
high-value
community
and
people
want
it.
So
something
we
should
probably
I
mean
is,
except
that
we're
driving
forward
kind
of
in
parallel
and
making
sure
that
the
processes,
when
the
API
side
are
such
that
they
have
with
usable
path
the
embedded
entity
one
step,
but
having
somebody
kind
of
drive
it
a
little
bit
more
quickly
works
well.
M
That
the
phase
one
specifications
are
very
being
done
very
actively
where's
the
phase
two
conversations
continue
kind
of
emerge
and
happen.
The
design
space
there
they
have
said
before
is
pretty
large,
so
we
need
to
narrow
things
down.
A
lot
of
of
these
clients
that
are
working
on
the
tube
you
can
chain
are
getting
near.
Pressed
net
data,
so
I'm
excuse
me
interesting
stuff
when
I,
probably.
M
A
N
Here
we
had
a
beta
release
this
week
or
actually
last
week
that
introduced
a
new
algorithm
for
gas
estimation
that
is
I'm
gonna,
say
we
call
it
gas,
exact
nation.
It's
accurate
down
to
two
one
gas,
even
for
the
IP
114
opcodes,
and
it
doesn't
use
a
binary
search.
So
it
only
requires
one
pass
through
the
transaction
out
codes
to
get
the
estimation.
A
All
right
thanks,
everyone
for
coming
I,
won't
be
on
the
call
in
two
weeks,
because
I'm
going
to
a
giant
star
wars
convention
that
should
be
pretty
awesome,
and
so
someone
else
will
have
to
take
over
I'll
find
someone
in
the
cat,
herders
too,
or
maybe
a
few
people
to
group
up
and
do
it
or
if
anyone
wants
to
volunteer
from
this
group.
We
have
right
now
feel
free.