►
From YouTube: NEAR EVM Working Group Update [2021-02-26]
Description
https://gov.near.org/t/2021-02-26-evm-update/780
Follow the latest from NEAR Protocol on,
Website: https://near.org/
Discord: https://near.chat
Blog: https://near.org/blog/
Twitter: https://near.ai/twitter
GitHub: https://near.ai/github
#Blockchain #FutureIsNEAR #NEAR #NEARProtocol
A
A
Nothing
comes
to
mind,
okay,
let's
proceed
and
see
if
we
can
come
up
with
something
along
the
way.
So,
let's
begin
with
the
current
goal,
we
remain
on
the
current
goal
of
looking
towards
the
destinator
release
by
the
end
of
q1.
So
that's
about
one
month
away
alex.
Let's
do
the
weekly
updates,
beginning
with
you.
B
Sure
not
major
from
my
site,
not
a
lot
of
updates
from
my
site.
I
was
working
with
evgeny
a
little
bit
helping
helping
with
the
with
the
work
on
the
near
side,
f,
connector
and
like
lots
of
just
small
tasks
regarding
the
aurora
thing,
but
in
general
nothing
crucial
for
for
the
current
stage
of
us
of
the
project.
A
A
So
there
is
some
work
on
going
for
the
berlin
hard
fork,
actually
pull
request,
indeed,
called
code
being
written,
and
that's
that's
great,
so
that's
coming
along
and
they
tell
us
that
it's
a
good
idea
for
us
to
upgrade
as
quickly
as
possible
to
the
latest
version,
because
the
latest
version
includes
significant
performance
improvements
up
to
5x
multiplier.
In
some
cases
they
also
thank
us
for
using
sputnik
vm,
as
in
they
are
very
happy
that
near
is
using
it
and
or
at
least
considering
using
it.
A
A
C
C
D
Yeah
I
am
kept
working
on
the
evm
bully.
I
added
some
option
that
we
can
change
the
test
net,
so
you
can
switch
to
test
net.
Now
I
read
out
the
block
context
the
issue
alex
raised
and
that
we
discussed
about
that.
We
need
an
additional
context.
I
read
out
all
of
that
information
made
it
ready
for
submission
to
the
to
the
evm
and
also
write
out
the
genesis
state.
D
A
C
A
C
C
C
A
A
B
Yeah
we
we
were
connecting
with
eugenie
over
the
course
of
this
week,
trying
so
so
yeah
we
are
trying
to
at
the
moment.
I
believe
evgeny
is
trying
to
to
deploy
the
connector
itself
to
the
near
testnet
and
and
and
robson
for
example,
and
try
the
the
transfer
itself
so
so
that
will
be
the
perfect
outcome
for
each
connector.
I
believe
it's
not
yet
there.
A
A
B
Yeah
from
that
already
more
than
three
and
a
half
thousand
slides
of
gold
yeah,
one
of
the
one
of
the
things
here
that
is
still
still
important
for
us
is
the
fungible
token
connector
yeah,
and
so
I
was
suggesting
to
to
kenny
to
not
to
focus
a
lot
on
this.
We
still
we're
do
not.
We
don't
need
to
publish
at
connector
together
with
the
bridge
right
now,
so
we
can.
We
can
use
what
whatever
we
want
to
use
it
right
now,
so
we
can
take
the.
B
I
know
the
current
point
of
view
on
how
the
fungible
token
should
look
like
before
it
is
finalized
and
just
use
it
like
this
and
then
at
the
moment,
when
we're
going
to
merge
at
connector
and
the
near
evm,
then
like
update,
update
it,
remove
the
usage
of
the
std
and
and
also
insert
the
the
correct
version
of
the
fungible
token
implementation
that
is
created
together
with
the
standard.
So
I
believe
the
fungible
token
standardization
process
is
not
blocking
us
here.
C
B
B
A
A
All
right,
so,
if,
if
no
questions
on
on
that,
let's
move
on
to
the
bully.
So
now
I
remember
that
what
we
talked
about
the
block
context-
frank,
we
might
want
to
summarize
here
to
let's
see
here
here
we
are
yeah.
You
know
what
we're
gonna
summarize
for
for
everybody,
so
that
they
know
what
the
the
difficulty
here
is.
A
As
in
the
there
exists
a
number
of
op
codes
in
the
evm,
specifically
five
five
op
codes
that
obtain
a
block
context
and,
of
course
in
in
solidity,
you
have
the
primitives
for
accessing
the
results
of
this.
So
in
other
words,
authority,
the
contract
can
get
not
only
the
box
current
number,
but
also
indeed,
something
like
the
blocks.
Current
the
current
box
difficulty,
the
gas
limit
used
even
the
timestamp
and
address
of
the
of
the
current
blocks
minor
and-
and
these
are.
A
A
A
Benchmarking,
benchmarking
mode
or
or
indeed
development
mode.
It's
it's
not
something
that
we
would
make
available
as
a
facility
to
end
users
when,
when
the
evm
contract
is
deployed
on
chain,
but
rather
it's
it's
just
in
in
order
to
support
a
tool
like
the
bully,
we
will
be
able
to
run
benchmark
scenarios
where
this
facility
is
available.
So
it
will
be
something
like
a
compile
compile
flag
in
the
evm
contract,
and
so,
let's
frankly,
do
you
have
anything
to
add
to
that
description,
or
is
that
the
correct
description.
D
A
Where
do
we
get
the?
I
don't
find
that
here,
but
I
mean
there's
an
erp
for
an
opcode
that
gives
the
chain
id.
So
I
guess
this
is
not
updated
for
for
that
right.
So
so
indeed
this
in
overall,
the
seven
seven
op
codes
that
we
have
thought
about
so
far
where
we
have
to
provide
this
overridden
runtime
context.
D
B
B
A
A
B
A
A
D
Just
one
thing:
why
I'm,
when
I'm
pushing
on
having
a
separate
contract,
because
I'm
worried
that
the
the
state
might
interfere?
Otherwise,
because
if
we
have
a
separate
contact
contract
just
for
testing,
we
basically
start
from
a
clean
slate
with
the
genesis
state,
and
then
we
can
replay
everything
if
we
have
like
put
it
all
in
in
a
different
in
the
same
contract,
we
might
have
state
interference.
A
Yeah
sure,
but
but
ultimately
I
mean
the
the
benchmarking
you're
doing
right
now
you
have
a
as
I
understand
it,
a
dedicated
server
where
you
will
be
running
near
core
and
you're
running
it
using
this.
Let's
say
benchmarking:
flag,
go
mode
deployment
and
then
you'll
collect
numbers
there.
So
in
that
case
it's
pretty
ephemeral
in
terms
of
yeah,
yeah,
yeah
sure
on
beta
on
it
yeah
sure
we
wouldn't
ever
put
this
on
the
main
evm
account.
C
A
A
C
A
B
B
C
A
So
I
believe
that
in
in
other
evm
compatible
chains,
these
these
are
pretty
dummy,
dummified
values
as
well.
So
it's
not
nothing
different
on
our
part.
In
this
case
yeah,
I
think
the
block
block
hash
is
the
biggest
problem
both
for
the
boolean
and
force
generally,
we'll
have
to
visit
that
one
again
go
ahead.
D
A
Yeah
I
mean
you
have
to
remember
that
there
are.
There
are
f1
test
nets
that
are
not
actually
proof
of
work,
so,
in
other
words,
we
should
just
stimulate
whatever
you
get
returned
in
a
proof
of
authority
or
a
proof-of-stake
version
of
that,
so
that
that
needs
a
ticketing.
Let
me
note
here:
this
is
not.
A
D
A
A
A
A
D
A
A
Yeah,
well,
the
I
think,
preferably
we
don't
keep
stage,
but
the
of
course.
On
the
other
hand,
in
terms
of
providing
the
context
for
the
call,
it
would
not
be
great
if,
if
you
had
to
on
every
every
transaction,
provide
the
block
hash
context,
which
means
a
massive
json
object.
You
know
with
256
block
hashes
at
that
point,
our
benchmarking
starts
to
be
fairly
fairly
limited
by
json,
passing
and
serialization.
A
D
A
A
We
we
aren't,
we
aren't
so
this
is
all
about
the
benchmarking
mode.
So
the
other
approach
I
had
suggested
is
that
potentially
this
isn't
used
that
much
with
block
block
heights
larger
than
you
know,
a
handful
for
just
the
most
recent
blocks,
and-
and
in
that
case
we
could
provide
those-
those,
let's
say,
the
most
most
recent
8
or
16
block
hashes.
A
We
could
provide
that
as
part
of
the
context,
in
which
case
the
the
evm
contract
doesn't
need
to
keep
any
additional
state
for
for
the
benchmarking
mode,
and
then
we
could
just
bail
out
in
case
it
turns
out
that
something
is
calling
a
block
hash
with.
Let's
say:
actually
you
know
255
or
whatever
this.
This
might
be
a
worthwhile
thing
to
to
start
with,
because
unless
there's
some
kind
of
test
contract
somewhere
that
uses
indeed
a
very
deep
block
block
hash,
then
it's
not
much
sense
spending
time
on
this.
A
D
A
A
A
C
D
A
E
A
B
B
E
A
A
D
A
I
mean
there
you
should
check
again
their
plans
in
that
regard.
Came
after
that
call.
You
had
with
him,
which
is
why
I
thought
they
might
have
been
inspired
by
your
call.
But
if,
if
that
work
is
proceeding
somewhere,
then
please
find
that
work
and
and
let's,
let's
sync
with
those
people
doing
that
work.
Okay.
D
D
A
The
first
one
is,
is
basically
about
the
bully
now,
what's
called
the
bully
now,
so
this
is
the
the
replayer
actually
taking
one
chain,
some
of
the
test
nets
and
replaying
that
to
get
some
some
figures
and
frank,
what
would
you?
What
would
you
estimate
is
the
completion
percentage
of
of
this
assigned
to
you,
okay
or.
A
D
B
A
C
B
B
Yeah,
okay,
so
I
think
that
potentially,
since
we
have
and
actually
evgeny
kapoon
is
on
the
call,
so
so
we
are
not
going
to
do
the
bug
bounty
for
it
right
now,
but
maybe
evgeny
capuan
would
be
able,
in
the
end
of
the
quarter,
to
take
a
look
at
at
the
evm
and
try
to
you
know
to
hack
the
the
smart
contract
and
review
the
code
of
it
so
evgeny.
What
do
you
think?
Is
it
something
that
that
is
interesting
for
you.
E
E
E
A
E
It
is
in
one
of
the
internal
part
of
blake
2,
and
this
version
has
some
extra
arguments
like
number
of
rounds,
which
is
usually
like,
usually
fixed,
but
this
this
pre-compile
allows
the
caller
to
specify
it,
so
we
will
need
our
own
implementation.
I
I
had
a
look
at
how
gif
is
implementing
this
this
precompiles,
so
they
made
the
copy
of
black
2
code
from
gauss,
not
exactly
standard
library.
E
A
A
A
Yes,
this,
this
is
a
definitely
the
case.
We
have
not
reviewed
that
as
yet.
So
that
is
a
good
thing
to
review.
This
is
being
landed.
For
other
reasons,
I
think
this
is
being
landed
to
support
the
xero
pool
project
in
particular,
so
we
should
make
sure
that
we
can
support
both
the
evm
and
zero
pool.
Ideally,
okay,
let
me
add
also
those
three
nodes,
so
this
is
an
additional
item.
A
C
A
Yeah
right
so
at
the
moment,
indeed,
everything
everything
is
implemented
in
web
assembly.
Only
the
performance
improvements
will
take
some
time
to
land
on
on
particularly
testnet
in
in
any
case,
so
alex.
What's
your
current
concept
for
for
when
we
would
actually
begin
the
bug
bounty
it's
about
the
bridge.
Also
on
this
thing,.
B
A
A
A
B
A
B
D
A
B
A
A
Although
we
want
to
support
that,
so
I
would
I
would
skip
assigning
a
score
there.
We
are
at
time
in
terms
of
the
length
of
the
meeting,
so
anybody
who
needs
to
hope
go
ahead
and
hope
we
will
just
continue
going
through
this
and
then
wrap
up
so
protocol
upgrade
this
this
you
have
alex.
You
have
posted
the
protocol,
upgrade
proposal
essentially
to
the
forum.
A
B
A
Okay,
now
for
partner
support
partners
being
able
to
succeed
basically
enough
everything,
so
they
certainly
begin
to
have
soon
enough
to
documentation.
Tooling,
tooling,
is
very
easy.
I
think
they
should
just
be
able
to
change
the
end
point
and
that's
what
we
enable
with
the
current
proxy
approach
and
support
yeah.
I
think
I
think
we're
doing
well
on
this.