►
From YouTube: ECDC - DAY 2: Round Table Eth1.0 - 2.0 by Danny Ryan
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
A
We're
still
narrowing
down
on
some
of
the
details,
but
the
difference
between
eighth
100
and
e2o
are
huge
and
the
links,
at
least
in
the
immediate
foreseeable
future,
between
the
two
or
there's,
not
that
many
links
there's
ways
to
get
into
this
new
system,
but
not
necessarily
ways
to
talk
to
the
solises
olds.
So
in
that
respect
there
are
a
lot
of
things
to
think
about
beyond
just
the
core
design
of
the
protocol,
and
this
room
is
full
of
people
that
it's
going
to
be
very
important
that
you
begin
to
think
about
these
problems.
A
A
First
of
all,
the
first
kind
of
thing
that
comes
to
mind
is
clients
right,
so
we
are
going
to
have
this
idea
of
an
EVM
legacy,
client
and
a
shard
claim,
and
they
can
either
be
one
in
the
same
or
a
client
could
support.
Both
the
client
could
support
one
or
the
other,
so
we
have
things
we
need
to
start
thinking
about.
In
terms
of
is
my
clients
this
or
that
if
I'm,
an
up-and-coming
client
do
I
focus
my
efforts
on
sharding
and
then
maybe
add
legacy
support
in
the
future.
A
If
I'm,
an
existing
client
did
I
focus
my
efforts
on
keeping
the
immediate
EVM
strong,
letting
other
people
experiment
and
then
moving
into
the
sharding
landscape
and
then
from
there
we
have
like
three
way:
people
interface
with
the
chain.
Are
we
making
the
the
way
that
people
have
to
conceive
of
these
two
systems?
There's
a
lot
of
design
decisions
there
into
the
interfaces
that
we
expose.
Do
they
have
to
think
about
these
two
different
landscapes?
So
we
have
two
separate
modules.
We
have
kind
of
an
unified
language
to
talk
to
both
of
these
systems.
A
That
user
doesn't
have
to
think
about
it.
We
have
the
contract
languages.
These
guys
are
working
on
you'll,
which
is
going
to
allow
solidity
to
begin
to
compile
into
wasum.
We
were
talking
already
vipir
and
some
of
these
other
languages
maybe
should
consider
moving
towards
the
standard.
Maybe
should
be
some
sort
of
VIP
to
standardize
you
all
and
try
to
promote
you'll.
Then
we
have
kind
of
the.
Then
we
started
moving
more
and
more
into
the
user
area.
We
have
existing
contracts,
we
have
existing
gaps.
A
A
It's
probably
depends
on
what
their
community,
with
what
they're
doing
with
their
community
and
what
their
token
actually
does
and
stuff,
but
things
that
we
need
to
start
thinking
about
and
then
just
kind
of
be
the
narrative
and
the
user
experience
in
general,
like
we
have
a
lot
of
people
that
are
involved
in
some
levels:
community.
What
is
the
narrative
moving
to
beef
to?
Oh
and
and
how
do
we
kind
of
capture
the
community
and
bring
everyone
over
the
next
time
anyway?
A
Those
are
just
the
large
problems
as
I
see
them
and,
as
I
said,
we've
begun
to
have
that
conversation
organically,
but
we
have
some
time
now
if
we
want
to
talk
about
anything
in
particular,
I'm
sure
a
lot
of
the
research
teams
thought
about
some
of
these
things.
I
know
the
awesome
guys
and
language
guys.
People
are
starting
to
figure
out
these
problems.
So
it's
you
know
a
good
time
for
people
to
chat.
B
So
one
thing
that
I've
been
thinking
about
I
got
really
well
I
got
mildly
worried
when
we
were
like
okay,
full
Casper
instead
of
Hybrid
Casper
just
from
a
messaging
standpoint,
and
so
then,
but
then
I
started
like
looking
into
the
full
Casper
and
I
kind
of
feeling
it
out,
and
it
feels
like
a
basically.
We
are
taking
those
old
Casper
messages
and
we're
kind
of
adding
more
information
and
that
more
information
is
bootstrapping.
B
This
this
beacon
chain
and
so
I
feel
like
an
important
part
of
at
least
that
part
of
the
messaging,
which
is
I,
think
only
the
first
bit
of
this,
like
larger
picture
of
messaging.
Where
we
talk
about
you
know,
awasum
and
the
EVM
I
think
that's
gonna
be
another
whole
challenge,
but
basically
saying
you
know
we're
we're
doing
this
hybrid
Casper
thing.
B
But
then
we
realized
that
at
the
same
time
as
we
build
on
this,
this
hybrid
group
of
work
for
the
state
chain,
we
can
also
start
bootstrapping
a
full
proof
of
stake
chain
so
that
we
can
like
kind
of
very
easily
transition
from
hybrid
to
full
and
not
just
be
stuck
on
hybrid
forever.
So
anyway,
this
is
something
that
I've
been
thinking
about
and
I
think
somewhere
reasonable.
A
C
C
C
We
feel
that
we
have
an
opportunity
to
work
on
the
shuttle
from
the
perspective
of
somebody
that
is
going
to
bring
it
to
the
masses
of
mobile
phones
to
really
pushing
this
angle
of
the
project,
just
what
a
valuable
experience
to
be
had
for
hospitals,
also,
thankfully,
a
theorem
community
to
contribute
to
this
process
right.
So
it's
not
much.
E
A
C
G
A
If
I
were
to
develop,
if
I
decided
to
develop
a
client
a
year
from
now,
I
think
it
would
be
sharding
first,
like
I
would
start
working
on
a
sharding
client
and
then,
if
I
chose
to
support
the
evm,
we
should
come
up
with
the
right
term.
For
it,
I
know
there
was
some
pushback
on
legacy
chain,
but
to
support
IBM
100
chain
that
would
be
kind
of
a
secondary
thing,
but
right
now
you
are
we're
in
limbo
or
I.
A
Think
if
you,
if
you're
ready
to
put
the
pedal
to
metal
on
a
client,
it's
like
build,
build,
EVM
one
if
you
but
there's
also
an
opportunity
to
be
a
like
prominent
sharding
client,
because
if
you're,
if
you're
involved
right
now,
you
can
kind
of
be
on
the
cutting
edge,
would
be
kind
of
have
the
information
get
in
there.
Early
didn't
get
to
use
it
early.
So.
D
It's
almost
what
I
was
gonna
say
about
the
limbo
it
says
so
worked
for.
My
feeling,
like
from
what
I
could
see
is
that
the
car,
the
existence
of
theorem
clients,
are
kind
of
living
in
a
permanent
state
of
crisis
at
the
moment,
because
there
is
a
certain
thresholds
which
are
now
being
reached
in
terms
of
the
capacity
or
I'm
not
talking
about
all
the
client
older,
like
it's
like
terribly.
You
know
you've
heard
about
this.
D
These
hysterical
thing
about
terrified,
reaching
the
terabyte
limit
and
stuff
like
this
and
still
every
week
there
are
some
messages
coming
on
like
what's
coupling
with
a
network,
you
said
like
gas
prices
going
up
again,
so
it
creates
the
feeling
that
we're
living
in
this
permanent
crisis
mode
and
that
again
translates
into
most
of
the
developers
and
the
current
teams.
Unfortunately,
I
broke
down
with
the
maintenance
of
the
existing
clients,
so
they
can't
really
spend
a
lot
of
time
and
looking
into
shouting
and
they
trying
to
avoid
that
situation.
D
I
think
a
lot
of
them
would
postpone
this
work
until
it's
like
the
very
last
stages
that
all
you
know,
they
still
figure
these
things
out.
If
you
start
implementing
that
now,
just
like,
we
might
need
to
completely
rewrite
the
stuff
again,
and
another
thing
I
was
going
to
say
is
that
in
this
usually
in
the
software
engineering,
if
you
decide
to
rewrite
something
like
I
see,
the
charr
team
is
completely
rewrite
for
this
area.
D
The
important
thing
to
do
is
that
to
also
change
the
way
you
or
you
do
software
engineering,
because
otherwise
you're
gonna
just
write
the
same
kind
of
same
thing
right.
You
have
to
change
your
practices
around
how
your
your
develop,
how
your
test,
how
you
shipped
otherwise
you
get
exactly
the
same
problem
as
before,
like
you
might
start
living
in
a
firm
crisis
again,
so.
A
In
that
respect,
you
might,
if
I
had
a
new
client
and
I
had
the
resources.
I
could
have
guys
working
on
EVM
one.
Oh
I
have
another
team
working
on
shard
and
keep
them
actually
like
kind
of
isolated
because
as
the
EVM,
when
a
client
actually
reaches
main
that,
like
you're,
probably
gonna,
start
putting
out,
fires
like
it's
just
gonna
like
consume,
consume
more
and
more
resources.
H
Kind
of
struggling
with
us
a
little
bit
in
the
II
wasn't
world
as
well.
So
obviously,
up
to
this
point
he
wasn't
has
been
designed
to
be
a
hundred
percent
EDF
compatible
way
back
to
the
original
roadmap
was
for
it
to
sort
of
replace
EVM
on
the
maintain.
That's
not
the
case
anymore,
and
so
this
is
a
question.
H
We've
been
asking
ourselves
a
lot
like
should
we
feel
bit
like
we
have
our
hands
tied
by
by
this
BPM
compatibility
question
and
there's
a
lot
of
things
that
we
would
change
or
redesign
we
sort
of
begun
this
conversation.
What
is
he?
Was
it
the
2.0?
It
look
like
now,
as
we
begin
to
approach
sort
of
a
group
of
concept
than
having
he
wasn't
sort
of
one
point
of
working.
Should
we
break
out
a
little
bit
from
the
EVM
mindset?
Are
there
you
know
like
just
as
one
concrete
example
like
we're?
H
Up
to
now,
we've
been
trying
to
match
the
gas
pricing.
We
have
been
matching
the
gas
pricing
exactly
one
for
one,
and
so
we
can
pass
all
the
existing
state
tests
and
stuff,
but
that
may
actually
be
a
waste
of
our
time
and
I.
Guess
if,
if
it
wasn't
code
is
never
gonna
have
talked
it
UVM
code,
then
you
know.
Maybe
we
don't
need
to
be
perfectly
compatible.
It.
A
A
A
I
We
have
a
solution
for
you,
you
don't
need
to
split
people.
There's
this
thing
called
EVM
to
wasn't,
so
you
only
need
to
have
the
Charlaine
client
who
it
wasn't
and
you
can
use
this
compiler
which
compiled
CPM
2000,
and
you
can
run
it
on
Watson,
and
this
goes
back
to
what
Lane
is
mentioned
on
matching
gas
prices,
but
to
be
clear,
it's
not
about
being
compatible
with
the
EVM
on
the
sense
of
was
encode
talking
to
a
vm
code,
whatever
you
don't
need,
like
a
hundred
percent
match
of
gas
costs.
I
I
If
you
only
have
it
wasn't
base
client,
but
you
still
going
to
support
the
main
chain,
but
secondly,
what
we
actually
use
it
for,
because
nobody
wants
to
write
tests
at
all
and-
and
there
was
a
decision-
they'll
accept
Dmitri,
but
we
couldn't
couldn't
hijack
Dmitri,
for
he
wasn't
because
he's
like
the
only
person.
Writing
tests
for
each
area,
and
so
there
was
a
decision
that
we
could
just
step
into
the
work.
I
Dmitri
is
doing
and
use
the
state
test,
and
to
do
that,
we
have
used
EVM
to
wasn't,
but
we
just
don't
have
enough
resources
to
like
keep
it
up
today,
as
you
mentioned,
they
would
need
to
split
a
couple
of
people
and
that's
one
of
the
questions.
We
actually
have.
What
should
be
there,
the
future
of
EVM
to
us
on
whether
this
concept
makes
sense
I'm,
not
sure
where
to
ask
that.
J
A
I
I
Yeah
I
mean
the
assumption
was
that
he
wasn't
gonna
be
next
to
the
EVM
on
the
main
chain,
but
it
doesn't
mean
that
you
need
to
you
need
this
tool,
because
the
clients
already
have
EVM,
so
you
could
just
put
he
wasn't
next
to
it.
It
was
really
just
about
having
like
a
clean
implementation,
if
you
don't
want
to,
if
you
just
want
to
get
rid
of
EVM
entirely,
but
still
support
all
contracts.
A
D
So
I
was
gonna,
offer,
maybe
offer
a
solution,
but
if
that
requires
to
go
back
on
some
of
the
roadmap
decisions
that
have
made
before
even
go
back
to
2017
I
think
what
Alec
was
talking
at
DEFCON
3
about
this
new
roadmap.
Where
are
we
going
to
keep
the
existing
chain
pretty
static
in
terms
of
functionality
and
start
building
things
on
the
on
the
side?
D
But
I
start
having
doubts
about
it,
because
I
think
we
end
up
actually
splitting
the
developers
into
two
groups,
the
ones
that
are
maintaining
the
current
clients
are
kind
of
super
busy
and
they
can't
work
of
work
of
the
people
were
the
the
the
the
sharding
and
faster
research
and
I
will
have
a
glue
people
who
were
doing
for
Charlene
clasp
research,
but
they
only
writing
the
prototypes.
And
so
how
do
we
not
do
that
and
just
take
everybody
together
to
the
fight
to
the
end
bow?
D
B
Think,
probably
one
of
the
hardest
parts
is
going
to
be
that,
like
EVM
awasum
transition,
I
think
like
from
the
developers
side.
That
is
definitely
super
hard
from
a
like
protocol
side.
I
would
say
that
the
like
casper,
FF
g,
/,
beacon,
change
stuff
shouldn't
be
too
radical
from
a
user
perspective,
and
so
that's
definitely
like
a
unambiguously
very
good
first
step,
at
least
so
far.
I
have
not
yet
felt
too
ambiguity
there,
but
then,
when
it
gets
into
like
okay,
how
do
we
do
crush
our
communication?
B
How
do
we
like
bring
contracts
down
to
the
the
shards?
How
do
we
actually,
like
you
know,
bring
ease
to
the
shards?
That's
when
it's
like
a
lot
more
ambiguous
and
hopefully
I
mean
that's.
Why
I
definitely
think,
as
you
said,
breaking
it
into
small
chunks
and
making
sure
that
each
time,
each
time
we
have
a
kind
of
clear
story
for
why
that
is
the
next
step
and
like
where
why
it's
it's
a
manageable
chunk
and
why
it's
still
kind
of
core
aetherium
cannon.
E
D
Of
you
worked
in
the
kind
of
big
organizations
with
lots
of
software
developers.
You
probably
witnessed
that
a
situation
where
there
was
like
there
was
an
existing
system
which
worked
at
everything
that
actually
made
money,
and
then
there
was
this
team
working
on
and
then
they
bring
the
new
people
to
work
with
a
new
version
and
I
thought.
Two
things
can
happen,
so
they
divert
really
quite
a
lot
and
so
either
the
new
team
gets
fired
and
then
the
whole
thing
gets
card
because
that
they
never
managed
to
to
actually
implement
this
or
the
magically.
D
A
K
Think
the
final
temptation
odd-
maybe
yes,
the
only
one
that
we
actually
seem.
You
thusly
that
increment
the
type
in
well
1.0
and
2.0
altogether
and
I-
don't
know
how
Piper
has
much
time
to
do
that.
He's
amazing,
you
hope
not
more
the
crime
and
tender.
Please
take
a
look
at
that
pay.
Some
time
for
the
shouting,
research
and
I
know
I,
don't
know!
K
Maybe
it's
because
those
Easter
Foundation
we
have
some
resource
to
do
that,
but
we
do
need
more
researchers
and
deaths
to
pay
attention
and
like
the
p2p,
next-generation,
p2p
topology
of
shutting
because,
as
we
mentioned
in
Taipei
three
months
ago,
also
collapse
from
that
permit
might
be
the
first
bottleneck
of
the
whole
row
map
of
Plantation.
So
we
can
talk
about
tomorrow.
G
Hope,
I,
don't
I,
have
an
answer
from
like
two
stalkers
ago.
Heavy
has
have
any.
How
are
the
clients
written
and
I'm
curious
if
they
have
swappable
components,
for
example
the
VM?
If
we
have
some
sort
of
interface
pavo's
doing
work
on
this
everett
Hildebrandt
is
doing
work
on
this.
To
have
VMs
be
swappable
or
you
can
coexist
so
then
we
we
have
we're
living
in
a
pedal
level
of
abstraction
that
we
can
swap
things
in
and
out
so
we're
more
flexible.
So
I'm
just
curious.
If,
if
we're
thinking
about
doing
things
like
this.
A
Yes,
depending
on
how
the
clients
design,
like
some
of
these
things,
are
walking
a
module
with
modular
can
be
pretty
modular,
but
like
we're
talking
about
total
like
this,
is
there
are
blocks
and
there
are
signatures,
but,
like
it's
very,
very
different
in
terms
of
like
the
closest
thing
you
have
is
that
you
have
the
you're
gonna
have
like
a
conversation
engine,
but
you
have
a
radically
new
landscape
and
like
using
the
same
components.
Is
it
will
take
some
effort.
L
L
You
know
if
you
you
worked
on,
would
it
be
something
that
you
have
to
abandon
later,
but
also
I
can't
promise
that
unless
you're,
a
researcher
and
and
kind
of
like
Mike,
the
way
that
I
think
I'm
thinking
like
oh
I,
guess,
like
my
plan,
is
just
continue
working
on
the
protocols
and
producing
evidence
that
it's
cool
and
you
know
producing
like
like
prototypes
and
eventually
test
nests
and
stuff
and
I'll.
Try
to
like
convince
people
to
use
them.
I,
don't
know
how
it
fits
into
the
rest
of
the
except
for
that
I
figure.