►
From YouTube: NEAR EVM Working Group Update [2021-02-19]
Description
https://gov.near.org/t/2021-02-19-evm-update/691
Follow the latest from NEAR Protocol on,
Website: https://near.org
Discord: https://near.ai/discord
Medium: https://near.ai/medium
Twitter: https://near.ai/twitter
GitHub: https://near.ai/github
#WhiteboardSeries #Blockchain #FutureIsNEAR
A
A
A
B
B
So
the
post,
it
changed
a
bit
so
the
way
how
it
so
we
still
want
to
do
it,
and
it's
more
like
an
informational
post
and
to
show
how
things
are
going
to
work
and
where
pieces
of
the
code
are
going
to
be
located.
So
there
must
probably
going
to
be
some
pieces
that
are
going
to
go
into
the
protocol
level,
but
they
are
going
to
be
quite
small.
B
They
are
not
going
to
affect
any
specifics,
like
the
economics
of
the
tokens
that
are
working
on
the
on
the
blockchain
or
introduction
of
some
specific
things
like
a
json
serializing
code
to
the
protocol
level
yeah,
and
so
so
that
post
is
has
a
little
bit
of
a
different
direction.
But
but
I
have
created
the
draft
now
it's
under
review
of
arto
and
once
once
we
are
done,
we
are
going
to
publish
into
the
governance
form.
B
A
Yeah
yeah,
I
think
we
might
might
post
it
today,
although
you
know
friday
not
the
best
time
to
be
posting
anything,
let's
review
after
the
call
right.
So
on
my
end,
I
continued
work
on
the
evm
contracts,
which
is
getting
upgraded
to
the
upstream
version
of
the
evm
machine
and
continuing
to
debug.
The
balance
of
tests
also
pretty
much
finished,
with
the
pull
request
for
adding
everything
we
need
to
near
core
to
support
performance.
A
Free
compiles
need
to
do
some
more
debugging
of
runtimes
params
estimator
to
get
the
actual
cost
for
any
operations
and
yep.
Let's
call
it
the
evm
release
proposal,
so
I
did
a
bunch
of
edits
on
on
on
your
post
alex
and
we
need
to
sync
on
that
again-
and
I
listened
in
on
the
last
several
all
core
devs
meetings
from
the
ethereum
foundation
regarding
when
they
plan
to
to
release
berlin,
harford.
Finally,
and
now
increasingly
plans
for
the
london
hard
fork,
which
will
be
summoned.
A
C
For
that
week
we
have
discussion
with
alexander
and
kirill
about
s
connector
architecture
and
also
some
important
things
with
killer
related
to
serum,
smart
contract
and
and
connections.
I
have
good
progress
for
that
week.
I
almost
complete
all
contract,
but
unfortunately
I
have
some
compilation
problems,
so
our
current
algorithm
is
completed,
but
my
compilations
are
not
not
successful,
so
I
need
some
time
for
fixing.
A
C
But
for
now
it's
I
think
it's
really
good
progress,
because
it
was
implemented
as
connector
look
at
balances,
integrated,
fungible
tokens
and
integrated
some
proofs
for
for
checking,
ethereum
contracts.
So
on
so
probably
all
algorithms
is
completed,
but
I
should
for
for
the
next
week
I
need
fix
compilation
box
and
after
that
we
will
start
with
scale
testing
our
contacts
from
sram
side
and
from
near
side.
B
B
B
That's
that's
it,
and
at
least
this
literally
this,
this
repo
is
going
to
implement
the
connector
side
and
the
near
side
contract
eventually
is
going
to
be
immersed
with
the
evm
contract
yeah.
So
kirill
already
added
his
test,
his
his
code
there.
It
was
an
interesting
discussion
about
the
about
the
how
how
users
are
sending
the
fees,
so
I
I
suggested
to
to
enhance
the
user
experience
there,
but
we
potentially
might.
If
we
have
some
time
we
potentially
might
discuss
it.
I.
B
D
While
we're
here,
I
wanted
to
mention
that,
if,
if
you're,
building
a
near
side
of
the
seize
connector
using
sdk
then
like,
if
you
will
need
to
rewrite
the
bench
so
like
because
evm
contract
doesn't
use
sdk
at
all,
it
has
kind
of
directly
uses.
So
it
doesn't
use
any
like
lookup
maps
or
storage
state
like
through
the
binding
generation.
So.
A
Just
take
now,
let's,
let's
come
back
to
that
in
a
second
in
a
discussion
so
for
frank's,
frank's
updates
he's
sick,
but
he
continues
working
on
the
riding
the
transactions
via
the
layer
that
we
call
evm
and
also,
of
course,
on
migrating.
The
initial
genesis
states
which
will
be
needed
to
to
replay
transactions
from
another
chain.
A
Interestingly,
on
the
all
core,
devs
call,
not
the
one
today,
but
one
two
weeks
ago,
I
heard
that
the
ethereum
foundation
is
also
now
working
on
something
like
a
replayer,
very,
very
similar
to
what
frank
is
doing
specifically
for
testing
the
berlin
heart
fork.
So
we
might
check
if
we
can
collaborate
with
on
that,
maybe
over
some
duplication
of
effort.
C
C
The
proxy
rpc
server
and
then
a
local
node
set
up
with
near
cli
link
drop
your
wallet
as
well
as
the
near
explorer,
and
then
I
just
read
through
the
materials
that
you
posted
that
I
thought
I
should
get
started
on
and
I'm
going
gonna
be
continuing
through
the
weekend
on
those
okay,
so
quite
short.
But
I
I'm
pretty
sure
I'm
ready
ready
to
go.
A
A
A
A
On
the
other
hand,
we
had
certain
items.
For
example,
we
had
the
boolean
hard
fork
upgrade
originally
scheduled
for
next
monday.
That's
not
gonna
happen
one
way
or
another.
Now
the
good
news
on
that
front
is.
We
can
probably
discover
this
all
together
from
q1,
given
that
the
boolean
hard
fork
keeps
sleeping
on
the
upstream
and
as
in
it's
already
more
than
a
year
late
and
looks
like
it's
going
to
be
a
little
later
still
so
april.
14
is
the
is
probably
a
pretty
certain
date.
A
On
the
other
hand,
we
don't
necessarily
have
to
make
that
date
for
our
berlin
hard
fork
support,
mostly
the
problem
from
our
end
concerns.
When
will
people
start
using
a
solidity
compiler
which
will
target
by
default?
The
boolean
heart
fork
specifically
solid
table
will
certainly
change
to
using
the
new
op
codes
for
subroutines.
A
It
will
make
for
faster,
faster
contracts,
less
gas-
that's
unlikely
to
happen
in
april,
so
I
need
to
check
on
the
solidity
team's
plans
on
that,
but
we
probably
have
some
time
in
q2
for
the
berlin
heart
fork.
So
realistically
we
have
no
option
but
to
describe
this,
given
that
we
have
more
fundamental
questions
to
address
still
and
at
the
same
time
it's
also
not
a
pressing
priority.
A
A
Well,
I
mean,
depending
on
what
we
push
down
to
nearport,
but
currently
the
whole
interpreter
is
in
the
contract,
and
that
means
we
will
deploy
a
new
version
of
the
contracts
yeah.
It
doesn't
need
changes
in
the
protocol,
but
in
but
but-
and
this
is
a
big-
but
if,
if
they
do
things
as
in
berlin,
hard
fork
of
introducing
new
pre-compiles,
you
know
for
let's
say
new
cryptographic
primitives,
then
it's
pretty
unlikely
that
we
would
want
those
to
stay
as
user
level.
You
know
contract
level
operations.
A
A
No,
I
think,
that's
that's
particular
so
hereby
d
scoped
and
we
will
revisit
the
question
okay
and
so
the
other
thing
alex
mentioned.
Let's
see
here
right,
so
we
we
are
almost
on
track
composting.
This
probably
we
will
post
it
today
to
the
forum
to
begin
within
that
discussion
regarding
well
in
general,
the
evm
released
on
test
nets.
A
A
So
that's
going
out
soon
and
let's
look
at
the
other
things
we
had
for
for
february
here.
We
don't
really
have
too
much
for
february.
If
connector
is,
is
probably
on
on
track
to
check
with
you
guys
on
on
progress
closer
now.
One
one
big
question
we
have
is
the
bug
bounty
program
alex
you
want
to
take
this
one.
B
B
C
B
B
So
after
after
the
launch,
we
need
to
you
know
to
to
deal
with
it.
We
need
to,
you
know,
make
sure
that
we
are
working
with
it
correctly,
though,
we
are
like
already
learning
how
to
do
it
on
the
test
net,
but
like
still
the
mainnet
deployment
is,
is
the
different
thing.
So
I
think
that,
like
at
least
a
month,
it
would
take
us
to
to
make
sure
that
we
are
completely
okay
with
it
and
then
in
the
middle
of
april.
This
is
going
to
be
a
like
an
appropriate
time.
B
A
B
Actually,
no,
I
believe
we
need
to
do
it
for
for
the
mainnet
okay,
because
it's
go
well.
We
have
the
main
net
deployment
yeah.
Another
thing
that
we
are
fixing
with
this
is
is
this:
is
the
incentive
for
people
for
hackers
to
to
submit
bugs?
Not
now
when
the,
when
there
is
a
test
net
bug
bounty,
but
either
later
when
the
bridge
is
live
on
the
mainnet?
So.
A
All
right
well,
on
the
on
the
evm
front,
we're
certainly
not
going
to
make
the
match
first
on
this,
given
that
we
we
don't
really
have
a
final
final
implementation
nearby.
Yet,
on
the
other
hand,
the
backbone
program
was
also
much
more
important
when
we
were
doing
the
protocol
level
evm
as
not
the
current
contract
approach
architecture.
A
So,
for
example,
we
had
here
a
decision
point
on
the
dismal
release,
meaning
whether
we
would
go
ahead
with
the
protocol
upgrade
or
not,
and
so
some
extent.
This
whole
thing
has
been
overcome
by
events,
but
we
should
still
have
it
in
terms
of
perception,
but
yeah.
Okay,
it
needs
to
get
pushed
pushed
out.
I
would
say
at
minimum
it
needs
to
get
pushed
out
two
weeks
and
if
it's
getting
pushed
up
two
weeks,
then
that's
pretty
close
to
what
you're
running
with
the
bridge.
In
any
case,.
B
B
A
A
A
B
Yeah
one
one
additional
thing
that
we
need
to
not
forget
about
about
is
the
protocol
upgrade,
so
it
depends
on
what
we
would
like
to
put
to
to
the
math
api,
as
you
mentioned,
so
if,
if
we
so
it
would
be
a
really
good
thing
for
us
before
the
end
of
this
well,
all,
potentially
we
can
do
this
next
quarter
at
some
point
in
time.
Obviously
we
would
need
to
do
some
some
performance
testing
and
understand
what
we
actually
would
like
to
put
to
the
to
the
protocol
level.
B
A
This
also
is
easy
recover
now,
okay,
which
is
the
most
important
one.
We
need,
oh
and
by
the
way,
funny
thing
about
that
in
in
doing
that
implementation,
I
discovered
that
our
easy
recovery
was
quite
buggy.
I
need
to
check,
with
with
bo
nelly
on
that
offline,
how
how
come
it
was
working
at
all
if
it
was
now,
the
other
milestones
remain
on
track.
I
think
we
are
okay
for
the
three
player
and
mike's
got
the
the
how
to
under
under
control.
A
Yeah
I
I
saw
that
I
thought
that
discussion,
but
it's
good
to
think
with
ilya
here
so
yeah,
the
bottom
line
being
that
the
contract
runtime
team
doesn't
feel
that
the
llvm
approach
is
worthwhile.
Necessarily
given,
given
that
okay,
it
would
execute
faster,
but
there
would
be
no
difference
in
the
gas
problem
or
couldn't.
E
D
Okay,
I
think
I
think
stuff
running
faster
underneath
is
still
a
very
useful,
useful
addition,
because,
like
at
the
end,
you
know
block
needs
to
execute
in
you
know
some
real
time
and
like
if
we
know
some
contract
will
be
gets
executed
a
lot
and
it
is
trusted
like
having
it.
You
know,
running
faster
is
a
pretty
you
know,
can
be
a
big
deal
for
just
performance
of
the
whole
system.
A
E
A
A
A
The
time
timing
is
is
currently.
Timing
has
a
hard
constraint
due
to
a
difficulty
bomb,
that's
said
to
happen
in
the
summer.
So
that's
that's.
Why
they're
rushing
the
london
hard
fork,
but
what
could
end
up
happening
is
that
they
solve
the
difficulty
bomb
and
then
london
gets
pushed
out.
That
seems
more
more
likely
in
london.
They
are
currently
debating
which
eeps
will
be
included,
and
on
that
front
I
wanted
to
ask
alex
so,
given
that
we
have
now
the
the
pessimistic
bridge
approach.
B
So
this
is.
This
is
one
thing,
however,
also
an
additional
thing
that
will
increase
the
performance
in
terms
of
the
the
time
for
the
transfer.
It
is
going
to
be
an
additional
it's
like
an
additional
feature.
It's
like
pretty
much
orthogonal
to
what
evgeny
is
doing.
This
is
the
realistic
bridge
yeah
and
before
that
it
seems
like
we.
By
that
moment,
we
will
have
almost
everything
done
on
the
protocol
level
and
but
if
bowen
is,
is
working
on
this
or
he's
going
to
work
on
this
in
the
beginning
of
the
night.
B
F
A
A
B
B
So
at
the
moment,
as
as,
if
guinea
said,
almost
everything
done,
he
is
fixing
some
issues.
The
stability
side
of
the
connector
is
is
done,
tests
are
underway,
so
well,
it's
done
in
terms
of
like
it
is
there
still
need
to
test.
It
need
to
need
to
do
the
integration
test
and
well,
because
you
understand
that
this
is
like
complicated
thing,
so
so
lots
of
things
in
front
of
us
still,
but
the
code
is
already
there.
B
There
are
some
questions
that
are
yet
to
be
disc,
yet
to
be
discussed
at
the
moment.
The
way
how
the
transfer
is
implemented
is
a
little
bit
different
from
the
way
how
it
was
done,
how
it
was
done
with
the
ear
seat.
One
is,
and
the
main
reason
is
to
allow
for
the
people
with
meta
mask
only
to
be
able
to
finalize
the
transfer
for
those
of
you
who
who
doesn't
well
just
just
a
quick
reminder
for
everybody.
B
And,
moreover,
if
we
are
going
to
ask
a
person
to
you
know
to
interact
with
the
interface
in
seven
minutes
after
the
moment
when
he
started
the
transfer
it
all.
It
is
also
not
a
very
good
user
experience,
so
taking
into
account
the
high
liquidity
of
of
ether
and
that
everybody
would
be
able
to
consume
it
as
a
payment
for
the
relaying
of
the
finalization
transaction.
B
This
approach
is
good
in
terms
of.
If
a
user
specifies
the
zero
fee,
then
he
is
able
to
relay
by
himself
if,
for
example,
he
possess
the
the
near
account-
and
he
is
on
the
safe
side
here,
because
if
somebody
else
will
will
will
relate,
then
obviously
you
will
just
have
this
finalization
transaction
for
free,
yeah
yeah.
B
The
specification
of
the
fee
is
is,
is
done
seven
minutes
before
the
moment
when,
when
the
actual
transaction
on
near
side
is
submitted,
obviously,
within
the
seven
minutes,
the
the
the
fluctuations
of
the
near
price
and
ethereum
price
is
not
going
to
be
drastic,
but
still
some
some
bad
things
may
occur
there
and
the
fee
can
be
too
low
and
the
user
need
to
have
an
ability
to.
You
know
to
still
finalize
the
transfer.
B
B
B
D
A
D
A
A
Hopefully,
foreign
some
of
the
code
for
sure,
but
the
the
way
to
look
at
it
is
that
kiril
and
evgeny
have
basically
they're
working
on
a
proof
of
concept.
It's
a
proof
of
concept
which
can
be
thrown
away
and
substantially
it's
still
onboarding,
as
in
they
are
learning
how
an
ndp41
is
written.
They
haven't
done
it
before.
A
D
Just
just
for
clarity
like
there's
a
bunch
of
stuff
that,
like
on
nearly
the
m
side,
because
it's
the
same
balances,
so
you
actually
need
to
use
exactly
the
same
data
structure.
That's
right!
That's
right!
That's
already
there!
So
yeah!
That's
fine!
But
I
mean
the
the
the
piece:
that's
like
lock,
event,
dot,
rs
and
stuff
like
this.
They
share
the
same
so
yeah.
That's
fine!
Yeah,.
D
Yeah,
I
guess,
like
one
question
I
had
is
like
the
code
in
the
like
in
this
pretty
much
rust
side
connector.
It
is
pretty
much
the
same,
almost
the
same
as
the
code
in
in
the
token
connector.
So
I'm
just
like
a
little
bit
cautious
that
we're
duplicating
that
code
in
like
and
when
the
bridge
will
be
changing,
we'll
need
to
like
go
hunt
for
all
the
places
where
yeah.
B
D
B
B
E
B
D
I
think
there's
just
like
there's
some
difference
in
data
structures.
So
that's
why
it
does
it
doesn't
need
to
be
copied
right
now,
it's
just
like.
Maybe
anyway-
and
I
mean
this
is
just
like
engineering
it
a
little
bit.
So
we
can
yeah
make
sure
that,
because,
like
we'll
have
like
an
ft,
connector
will
have
like
a
bunch
of
those
and
so
yeah.
B
B
D
B
Yeah
yeah
yeah.
Well,
absolutely
so
so
that's
why
we
we
have
posted
that
upgradability
and
governance
thing
I
was.
I
was
saying
there
what
at
what
stages
of
the
upgrades
of
the
bridge
the
connectors
need
to
be
also
upgradable,
because,
obviously
the
bridge
is
going
to
change
the
code
going
to
change
the
interfaces
from
time
to
time
and
going
to
change
potentially
well
at
this
moment
is
changing
actually
addresses
so
until
we
have
not
implemented
the
proxy
pattern,
it
changes
the
addresses,
so
the
connectors
also
need
to
be
upgradable.
D
A
A
D
D
That's
I
mean
that
can
be
discussed,
I
think
separately,
but
the
the
main
I
think,
point
that
max
made
is
that
we
planning
at
some
point
to
change
the
storage
layout
and
when
we
change
the
storage
layout,
the
iterators
might
not
be
possible
like
if
we,
for
example,
switch
to
like
a
hash
map
underneath
and
which
means
like
you
know.
This
operation
will
go
from
being
like
super
cheap
to
being
super
expensive.
If
we
can
wouldn't
continue
supporting
it,
and
so
it's
like
a
very
suboptimal
to
add
it.
D
D
F
So
we
can
just
not
actually
delete
it
and
just
have
some
way
of
marking
them
as
being
marked
for
deletion.
So,
for
example,
in
fact,
so
the
problem
is
that
the
ethereum,
it
is
actually
possible
to
recruit
a
contract
at
the
same
address,
use
it
using
crate
to
instruction,
and
when
this
happens
it's
stop.
The
new
contract
storage
should
be
empty.
F
If
this
were
not
possible,
then
we
could
just
leave
this
storage
in
place
because
we
as
nothing
would
be
able
to
use
it,
but
not
given
that
in
ethereum
it
is
possible
to
recreate
a
contract
with
the
same
address.
We
can,
for
example,
have
some
sort
of
generation
number
on
that
account
and
in
each
storage
entry
and
then
compare
the
numbers
if
the
numbers
are
different.
This
means
that
the
storage
entry
is
from
the
old
contract,
and
so
it
should
be
ignored.
D
F
F
A
So,
for
now
like
for
now
and
right
yeah,
so
for
now
we
we
are
not
blocked
on
it.
We
we
just
need
to
work
around
it
and
we'll
come
back
to
this
question
later
then
sounds
good,
and
thanks.
Thank
you
for
your
thoughts
on
that.
Okay,
do
we
do
we
have
anything
anything
else
today
that
people
wanna
discuss
yeah.
E
A
E
Well,
no,
but
oh
sorry,
my
question
is:
how
exactly
are
you
going
to
so
what
is
the
interface
for
this
replay
like?
Do
you
just
how?
How
do
you
plan
to,
for
example,
map
ethereum
blocks
to
near
blocks
or
like
what
exactly
is
being
replaced
are?
Are
you
just
replaying
all
the
transactions,
or
are
you
do
you
want
to
like
re,
do
some
mapping
from
block
to
block
and
then
replace
the
blocks
or
what
exactly
is.
A
E
Okay,
okay,
yeah
I
mean
there
is
a
I
mean
it
just
depends
on
how
what
is
the
setup
right?
If
your
setup
is
like,
you
actually
run
on
your
node
and
then
you
want
to
submit
your
rpc,
then
that's
obviously
more
difficult,
but
if
you
just
want
to
have
something
that
replaces
all
the
transaction
without
actually
running
a
near
node,
then
that's
much
easier
because
you
can
do
like
low
level.
E
A
A
We
also
need
to
sync,
with
whatever
people
are
doing
with
regards
to
testing
the
berlin
heart
fork,
because
they'll
have
to
be
thinking
about
exactly
the
same
things.
So
a
number
of
people
on
the
orca
all
core
dev
school
had
expressed
that
they
would
like
a
tool
like
this
in
order
to
to
have
some
faith
in
the
building
hot
fork,
and
so
that
tool
is
being
developed
somewhere.
I
haven't
found
the
repo.
A
Yet
okay
come
back
to
this
yeah,
okay,
okay,
so
for
for
next
week,
yeah
time
I
can,
I
can
compose
the
summary
for
next
week.
After
the
call,
I
think
it's
pretty
clear
what
everybody's
working
on
with
frank
about
his
state
and
he's
available
joshua
will
will
come
back
to
you
on
monday
latest,
with
some
some
tickets
assigned
your
guinea
again,
you
have
everything
you
need.
You
have
any
anything!
That's
blocking
you.