►
From YouTube: IETF114 SATP 20220726 1900
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).
B
B
We
might
have
a
long
one,
I'm
wes
hartaker
for
those
that
don't
know
me
and
there's
a
mailing
list
for.
B
Little
bit
of
discrepancy
in
terms
of
what
this
group
is
called,
it
was
originally
called
sat
and
the
mailing
list
is
called
sap,
but
the
both
is
called
sat
p
with
the
p
for
protocol.
So
don't
let
that
confuse
you
some
of
the
slides
might
say
one
thing
and
other
slides
might
say
another
thing:
it's
still
the
same
thing.
B
So
it's
all
good.
This
is
the
agenda
for
today,
we'll
start
off
I'll
rant
for
a
little
bit
about
stuff.
Then
we're
gonna
have
the
brunt
of
the
material
which
will
be
both
problem
statements
and
assumptions.
Some
use
cases
on
how
this
is
expected
to
be
used,
as
well
as
a
protocol
overview
for
where
the
work
might
start
from
with
an
initial
draft
that
need
to
be
later
adopted
by
the
working
group.
Should
it
go
forward.
F
B
That
and
where
people
can
bring
other
stuff
note
that
the
the
three
sections
of
questions
following
those
other
presentations
are
really
just
clarifying
questions.
We're
not
going
to
get
into
the
brunt
of.
Is
this
a
good
idea,
or
does
this
belong
in
the
ietf
until
later?
We
need
to
get
through
those
three
presentations
and
understand
the
background.
First
and
the
chartered
discussion
should
shake
say:
20
minutes,
I
thought
I'd
change
that
apparently
I
didn't.
B
We
do
have
a
draft
charter,
it's
on
the
mailing
list
and
there's
some
on
the
slides
and
then,
of
course,
we'll
have
the
standard
ending
of
above.
Is
this
a
good
idea
and
should
the
work
proceed
in
the
ietf,
I'm
sure
most
of
you
have
at
least
seen
the
note.
Well,
I
don't
know
how
many
of
you
have
read.
H
B
We
have
a
bunch
of
documents
that
talk
about
this
that
and
that
talk
about
everything
from
how
the
standards
process
works,
to
how
working
groups
work
to
anti-harassment
procedures
and
code
of
conduct
and
copyright
and
patents
and
stuff.
Everybody
needs
to
be
aware
of
that,
especially
if
you're
representing
your
your
company,
our
code
of
conduct
guidelines,
especially
this
you
know,
really
holds
true
for
boss.
Things
can
get
heated
and
above
if
differing
opinions,
you
know
kind
of
come
around.
C
B
Limit
the
use
of
slang,
you
know,
use
use
english,
that
most
people
will
understand
dispute
ideas
using
region,
reasoned
argument,
in
other
words,
don't
say
that
that's
a
horrible
idea
explain
why
you
think
it's
a
horrible
idea,
use
your
best
engineering
judgment
and
then
find
the
best
solution
that
works
for
the
whole
internet.
The
the
ietf
is
designed
to
work
for
the
whole
internet,
not
just
a
segment
of
the
internet
and
contribute
to
the
ongoing
work.
Note
that
the
ietf
has
a
masking
policy
for
for
this
ietf.
B
Everybody
should
be
wearing
a
mask
unless
you
are
taking
a
sip,
no
taking
one
hour,
long
drinks.
That
does
not
count.
I've.
B
So
you
know
they're,
the
the
cases
really
have
been
rising.
Some
of
the
iuts
staff
are
sick.
It's
actually
not
good.
So
please,
everybody
keep
your
masks
up.
J
Russ
you're
actually
gonna
stay
in
the
policy,
there's
no
exception
for
eating
and
drinking
in
these
rooms.
Thank
you
and
dan.
That
means
you.
I'm
serious,
put
your
mask
on
you're
doing
this
all
week.
B
C
B
Correct
and
eckerd's,
probably
quoting
it
correctly,
because
he
always
does
the
ietf
tool.
So
if
you
want
to
speak
in
the
room,
please
do
use
the
qr
code
over
there
to
log
in
you
need
to
log
in
with
the
data
tracker,
and
we
will
be
using
only
people
at
the
queue
so
lining
for
the
mic.
Won't
do
do
well.
You
need
to
make
sure
you
use
the
queue
by
clicking
on
the
little
button
in
the
agenda
or
you
can
scan
the
qr
code
either.
One
of
those
will
work.
B
In
general,
boss
have
a
particular
goal
right
we
need
to
decide.
Does
this
problem
need
solving?
Is
the
ietf
the
right
place
for
it?
Are
there
enough
people
interested
in
doing
the
work?
Is
the
scope
well
designed?
Is
it
too
big?
Is
it
too
small?
Are
the
deliverables
right?
Do
we
think
that
this
is?
You
know
this
is
what
should
come
out
of
the
working
group?
You
know
for
it
to
succeed,
and
if
we
do
put
all
those
other
things
together,
will
it
actually
succeed?
We
think
that
there's
a
good
chance
of
success
and.
B
5434,
which
is
a
great
document
for
bots
in
general,
but
really
about
getting
work,
started
so
back
to
the
agenda.
This
is
the
agenda
and
we
are
going
to
go
on
to
thomas
with
the
problem,
statements
and
assumptions.
B
B
L
Sure
so
I'm
going
to
talk
about
that's
the
problem
definition
and
the
assumptions
that
we're
going
into
this
okay.
So
so
this
slide
and
the
next
slide
next
slide
is
diagram
is
terminology.
So
this
is
particularly
important
because
we
want
to
keep
the
discussion
sort
of
generic,
so
I'm
using
the
word
network
as
a
group
of
nodes
that
share
state
about
a
digital
asset
and
you're,
going
to
turn
around
and
say:
what's
the
digital.
L
Okay,
I
won't
sing
so
but
I'll
just
speak.
I
promise
so
validity
means
that
the
data
object
is
legally
recognized
and
the
word
legal
is,
you
can
talk
about
that
later,
which
means
it
has
some
economic
value?
That's
been,
you
know,
legally
recognized
with
inside
jurisdiction.
To
have
them
gateway
is
a
computer.
L
L
Okay,
so
moving
forward
so
problem
statement.
What
we
want
is
an
interrupt
protocol
that
permits
the
movements
of
a
digital
asset
from
one
network
to
the
other,
and
we
have
some
conditions,
and
that
is
that
we
have
to
do
this
while
retaining
the
validity
of
the
digital
asset,
and
it
has
to
be
done
in
such
a
way
that
when
the
there's
a
third
party
independent,
that
is
asked
to
verify
post
transfer,
of
course,
that
they
can
do
that.
There's
sufficient
evidence
for
somebody
to
come
in
to
say
yes,
this
is
true.
L
This
is
not
true,
because,
there's
always
you
know,
system
failures,
and
so
there's
always
cases
where
disputes
occur.
So
that's
that's
kind
of
the
condition
at
least
one
condition
so
assumptions.
So
we
assume
that
both
networks,
please
share
common
semantics
about
what
the
data
object
is
and
the
notion
of
validity.
L
So
and
this
notion
of
ability
is
going
to
come
come
up
again,
it
really
depends
on
the
legal
legal
jurisdiction
of
where
that
you
know
network
operates.
We
assume
one
or
both
networks
are
opaque,
meaning
that
we
assume
that
you
can't
see
into
the
resources
into
the
constructs
in
the
network.
So
I'm
I'm
using
the
analogy
of
routing
where
you
have
an
isp
network
and
you
know
nicely,
is
not
going
to
let
you
look
into
all
the
subnets
within
the
asp
network.
It's
the
same
same
concept.
This
is
the
autonomous
systems
principle
concept.
L
The
gateways
are
trusted,
so
we
can.
We
can
talk
about
that
later,
but
it's
trusted
to
do
this.
One
thing,
and,
and
only
that
thing
information
subsets,
are
what
we
call
views
maybe
share.
So
views
are
not
asset
transfer
views,
it's
just
an
extract
of
information,
that's
queried
to
a
gateway,
so
you
could
say,
hey
gateway.
I
know
you've
got
you
know
behind
you,
there's
this
type
of
network.
I
want
to
ask
you
about
this
particular
state
of
an
object
in
that
network
and
the
gateway
is
going
to
respond.
L
L
So
what
we
want
to
do
this
is
use
the
gateway
model,
because
you
know
we've
got
a
lot.
A
lot
of
lessons
learned
from
the
history
of
the
internet.
You
know
many
many
people
don't
realize
the
the
internet
is
not
one
long
contiguous
network,
there's
lots
of
networks
in
between
there's
backbone
networks
and
in
in
many
cases
you
know
the
the
interconnection
point
between
two
networks
are
in
fact
implemented
by
some
gateway.
You
know
gateway
function,
so
we
assume
there's
one
or
more
gateways
in
front
of
each
network.
L
And
so
our
proposed
area
of
work
in
the
itf
is
the
yellow.
It
looks
green
on
the
screen
here,
but
it's
it's
the
yellow
bit
the
in
the
middle
right.
So
it's
between
gateway
and
g1
and
g2.
L
We
don't
care
or
we
it's
out
of
scope
for
us.
You
know
to
how
the
network
is
is
defined.
What
is
the
network
and
so
on
and
so
on?
And
if
we
do
the
gateway
protocol
correctly,
it's
possible
that
one
of
the
networks
or
both
of
them
could
in
fact
be
monolithic
systems.
Right
so
think
about
you
know
existing.
You
know,
I
don't
say
banking
systems
rtgs.
You
know
clearing
systems
that
some
of
you
might
be
aware
of
it
could
be.
One
of
them
could
be
both
of
them
right.
L
C
L
Adapter,
a
connector
whatever
with
the
network
behind
you,
and
why
why?
Why
is
this?
Because
this
is
too
many
networks?
There's
too
many
network
types?
So
it's
it's
just
you
know
it's
just
too
much
work,
so
we
want
to
look
at
three
particular
modes.
We
just
followed
modes.
First,
one
is,
you
know,
asset
transfer.
L
The
second
one
is
this
data
sharing,
so
you
know,
there's
a
gateway,
there's
an
api.
If
you're
authorized
to
ask
a
question,
you
go
to
that
api
and
it's
not
an
asset
transfer
you're,
just
extracting
information,
the
third
one
which
is
down
the
road
somewhere.
Sometimes
it's
going
to
be
what
we
call
asset
exchange
or
asset
swap,
which
is
a
coordinated
swap.
That's
you
know,
occurring
through
the
two
gateways.
L
The
scope
of
work,
api,
endpoint
definitions,
resource
identifiers,
the
message
payloads,
the
message:
format,
the
message
flows
so
for
the
floats
we
are
looking
at
a
the
classic
three
pc
three
phase
commit.
This
is
classic
database
textbook
if
you're
in
college.
This
is
database
one
or
database
two.
Of
course
you
know
for
those
for
those
people
who
don't
know
what
the
2pcsvpc
is
and
there's
there's
acid
properties
there's
also
classic
properties.
L
Okay,
so
current
main
technical
graphs,
we
have
an
architecture
drive,
we
have
a,
we
have
a
protocol
draft
and
we
actually
have
a
use
cases
draft,
that's
not
yet
uploaded,
but
you
know
it's
it's.
We
have
material
for
use
cases
and
in
fact
we
are
looking
for
people
to
provide
us
with
use
cases
that
you
know
that
would
be.
You
know
be
useful
for
the
for
the
discussion
is
that
my
last
slide.
I
think
that's
my
last
slide.
B
Okay,
so
with
that,
I
will
take
a
clear
clarifying
questions,
as
I
mentioned
coming
up.
Next
is
some
discussion
about
use
cases
on
where
this
is
envisioned
and
then
followed
by
that
as
an
early
protocol
description?
So
if
your
questions
about
those
do
table
them
until
later
dave,
please.
O
Dave
taylor
and
I
could
have
saved
my
question
for
the
next
presentation,
but
I
looked
at
the
slides
and
the
question
is
actually
more
appropriate
on
the
slides
on
this
one.
So
if
you
don't
mind
if
you
can
go
back
to
the
terminology
slide,
so
I
can
ask
my
clarifying
question
while
you're
doing
that,
let
me
prefix
it
with
a
question
I
think
is
answered.
O
So
you
can
see
where
I'm
coming
from
on
my
clarifying
question
it
has
to
do
with
which
types
of
assets
are
in
scope
here,
right,
there's
three
types
of
assets:
you
might
think
pure
pure
digital
things
that
have
no
reality
in
the
real
world.
There
are
things
that
are
pure
physical
and
there
are
things
that
are
hybrid
and
hybrid.
O
I
mean
things
like
the
keys
to
the
physical
device
right,
the
digital
keys
for
there
and
I'm
interpreting
digital
asset
as
being
those
two
categories,
but
not
the
pure
physical
one,
but
it
could
be
the
digital
asset.
That
is
the
proof
of
possession
or
proof
of
ownership
of
something
okay.
So
that
would
then
lead
to
my
actual
terrifying
question.
Okay,
because
if
I
look
at
asset
transfer
here,
there
are
two
types
of
asset
transfer,
voluntary
and
involuntary
elliott.
Leader
gave
a
great
presentation.
O
I
think
it
was
in
one
of
the
iot
meetings
I
think
at
last
ietf
about
involuntary
ownership
transfer,
which
more
typically
shows
up
in
the
hybrid
case,
but
can
show
up
in
the
pure
digital
as
well,
and
we're
talking
about
lawful,
repossession,
lawful,
confiscation,
lawful,
foreclosure
things
like
that.
Okay,
so
my
question
is:
are
those
in
scope
because,
according
to
this
slide,
the
answer
would
be?
Yes
if
you
go
forward
a
slide?
O
The
answer
is:
where
you
talk
about,
you
know
the
the
notion
of
a
gateway,
or
whatever
tends
to
imply
that
the
answer
is
no,
but
maybe
that's
because
you're
jumping
into
solution
here,
but
this
is
the
problem
statement
thing.
So
my
question
is:
is
involuntary
asset
transfer
of
digital
assets?
Is
that
part
of
the
architecture
or
not
part
of
the
architecture?
This
is
the
problem
statement
I'm
asking
which
problem
you're
trying
to
solve
so.
L
To
because
the
first
question
is
the
type
of
asset-
yes,
the
the
digital
or
hybrid,
it's
out
of
scope
for
us
right
now,
because
we
assume
that
the
applications,
the
the
two
sides-
know
what
they
talk.
They
talk
about
the
same
asset,
and
so
we've
had
discussions
about.
How
do
you
identify
an
asset?
There
is
in
fact,
industry
efforts
to
provide
numbers.
There's
an
itin
number
for
digital
assets
is
being
developed,
so
gateway
g1
needs
to
tell
gateway
d2
it's
this
one.
It's
this
asset
g2
will
then
look
up.
Okay.
This
is
yes.
It's
valid.
L
O
Okay,
this
is
the
problem
statement.
Presentation
gateways
are
part
of
the
solution.
If
the
problem
you're
trying
to
solve
is
also
to
do
involuntary,
then
that
dictates
whether
and
how
you
use
gateways
right,
I'm
trying
to
not
use
the
word
gateway
in
my
question,
because
I
don't
consider
that
to
be
a
problem
statement.
I
consider
that
to
be
a
solution
to
a
problem
right.
The
problem
is,
do
you
want
to
do
involuntary
digital
asset
transfer
or
not,
and
it
sounds
like
your
answer?
Is
you
do
not?
L
J
J
So
what
I
mean
by
that
is
that
in
many
networks
and
many
networks
of
heights
will
use
bitcoins
or
nfts
they're
constructed
in
a
way
that
they're
they're
not
supposed
to
be
non-trusted,
which
is
to
say
that
there's
no
entity
which
can
involuntarily
transfer
your
asset
from
one
dollar
right,
and
so
my
question
is:
when
is
the
gateway?
In
this
case?
L
So
so
so
the
gateway
is
trusted
and
the
gateway
just
obeys
the
command.
So
I.
J
So
let
me
just
let
me
understand
this
correctly,
that
forget
about
forget
about
like
destination
network
two,
I
have
an
asset,
a
network,
one
you're,
saying
that
the
gateway
can
transfer
my
assets
to
dave
without
my
cooperation.
That
is.
J
L
B
J
L
B
Two
two
different
banks
using
swift
right
have
some
level
of
trust
when
they're
talking
to
each
other.
From
what
I
understand,
this
is
very
similar
right.
One.
C
B
Has
trust
that
another
one
is
going
to
act
in
a
they
have
an
agreement
in
place?
Maybe
it's
backed
by
legal,
that's
out
of
scope,
how
to
define
that
trust
is
out
of
scope,
so
my
you
know,
and
I'm
not
a
proponent
but
my
vision
of
when
they
started.
This
is
that
we
start
with
a
small
building
block
and
the
expansion
into.
How
do
you
define
what
each
network
looks
like,
and
how
do
you
define
what
all
that
is
you
get
the
protocol
between
the
gateways
first,
but.
J
That
that
doesn't
work,
because
because
because
the
structure
of
these
systems
requires
that
the
protocol
embody
those
those
other
transition
requirements
is
part
of
it.
That's
what
I'm
saying
it's
like.
If
you
have
systems
systems
which
are
like
zero
trust
systems,
which
the
system
with
some
systems
in
the
large
list
are
not,
then
the
asset
transfers
have
to
embody
the
as
this
protocol
is
asset
transfer,
the
whole
thing
falls
apart.
If
you
look
at
like
side
chains
on
like
you
know,
side
chains
and
bitcoin
right,
they
embody
them
by
the
president's,
and
so.
L
J
E
B
So
the
definition
of
what
is
what
is
a
digital
asset
or
what
is
transferable,
is
that
a
scope
right,
so
we're
not
talking
about
any
particular
form.
It
doesn't
say
that
it
couldn't
be
used
for
bitcoin
or
it
couldn't
be
used
to
transfer
my
the
thing
that
came
up
in
iab
open
just
a
bit
ago.
Right
was,
I
have
a
movie
that
I
bought
with
with
one
vendor.
J
B
P
Okay,
so
my
question
is
slightly
related
to
these
ones,
but
I
don't
think
it
quite
got
answer.
I
think
it
might
be
easier
and
fitting
to
think
can
you
can
you
just
define
a
little
bit
more
when
you
say
trusted
what
you
mean
by
trusted
and
I
don't
care
about
the
gateways
trusting
each
other.
I
don't
care
about
the
other
gateway,
the
the
other
network.
What
does
it
mean
on
this
diagram
that
network
one
trusts
gateway?
One
like
what?
What
do
you
mean
by
that
part
of
word?
Trust
in
that
context,
so.
P
P
So
what
would
voluntary
even
mean?
I
mean
it
seems
that
the
definition
you
provided
trusted,
which
I
think
is
fine
for
the
context
of
this.
It
removes
all
possibility
of
even
knowing
whether
it's
voluntary
or
involuntary
like
g1.
Just
does
it
and
that's
pretty
much
what
you
said
before
right.
It
takes.
L
C
L
R
Wilson,
cisco,
so
I
just
want
to
question
and
clarify
about
what
allowed
failure
modes
are
here.
So
is
it
acceptable
for
this
transaction
to
fail
and
require,
like
human
intervention,
sorted
out.
R
L
R
R
R
Yeah
and
it's
like
when
you
transfer
money
between
countries
and
things
it's
possible,
the
money
goes
from
a
to
b
or
just
gets
lost,
and
you
have
to
then
phone
people
up
and
say
where's
my
money.
So
it's
not
like
it's
guaranteed
to
either
be
in
your
source,
account
or
the
destination
account.
You
can
end
up
failing.
R
L
B
E
Hi
roman
de
niro,
carnegie
mellon
university,
just
to
clarify
what
I
think
is
a
simple
question
from
a
perspective
of
sat
when
we
say
asset,
it's
just
a
blob.
B
O
O
O
L
O
So
I
think,
that's
a
very
hard
problem
and
you
might
want
to
back
off
on
that
and
and
say
legally
recognized
across
the
entire
set
of
things
where
I
think
that
one
might
be
tractable.
G
M
M
M
B
H
Doesn't
work
but
I
think
I
agree
the
definition
like
rigorousness,
but
I
think
we
should
let
them
present
their
case
rather
than
beating
them
down
my
understanding,
good
stuff.
This
is
good
stuff,
but
I
appreciate
it
very
much.
My
understanding
is
that
the
two
gateways
actually
authenticated
each
other.
Let's
not
use
the
bigger
words.
If
the
smaller
one
can
work,
they
authenticate
it
and
they
agreed
to
transfer
this
block.
That's
it
and
let's
not
dive
deeper
than
that.
K
G
So
yeah,
let
me
make
a
presentation.
I
imagine
we'll
have
some
questions
from
the
same
lines
that
thomas
face
so
I'll
try
to
address
them
later:
okay,
next
ip
yeah.
Can
you
hear
me
now
yeah
all
right,
very
good?
G
Okay,
so
I
work
for
ibm
research
and
I'm
going
to
try
and
motivate
some
use
cases
for
you,
which
are
the
kinds
of
cases
we
have
been
encountering
for
the
past
few
years,
so
in
ibm,
research
and
ibm
in
general,
we've
been
building
over
the
past
few
years,
a
lot
of
solutions
around
the
business
networks,
especially
in
the
supply
chain
area.
So
what
you
see
here
is
again
the
first
bullet
point:
promise
of
open
network
monitoring
them
all.
Just
it's
just
an
allusion
to
ethereum
you
can.
G
G
You
can
build
networks
on
the
in
different
ways
and
those
networks
can
do
can
solve
different
purposes
like,
for
example,
you
see
here
there
are
networks
that
we
have
built
that
handle
trade
logistics
for
example.
So
one
example
I
give
you
is
a
trade
length,
it's
a
system
that
ibm
built
with
mask
and
involved
other
exporters
and
shipping
carriers
and
connected
all
of
them
to
manage
mainly
shipping
logistics.
G
That
is
all
the
documentation
involving
the
export
of
goods
from
one
country
to
another,
passing
through
a
carrier
and
making
sure
that
all
regulations
are
followed.
There's
there
are
networks
we
built
for
trade
finance.
That
is,
if
you
all
know
about
the
counterparty
risk
in
the
export
import
problem.
Refinance
is
the
way
to
to
mitigate
that
problem.
That
is,
does
the
exporters
transfer
the
goods
first
or
the?
G
Does
the
importer
pay
for
the
goods
first,
because,
whichever
side
links
first,
I
mean
the
other
side
stands
at
risk
right,
so
trade
finance
has
over
the
past
few
centuries.
People
have
come
up
with
solutions
such
as
the
letters
of
credit
to
make
sure
that
exporters
importers
are,
can
safely
conduct
trades
without
losing
out
in
the
process,
and
that
involves
banks.
So
we
have
built
some
networks
from
that
we
build
networks.
G
We
built
demonstrations
on
how
to
share
kyc
information
across
different
banks
who
are
part
of
a
single
network,
and
then
you
can
also
have
a
networks
built
to
track
the
provenance
of
food
like
there's
a
network
alive
in
food
trust.
So
all
of
these
networks
they
were
built
according
to
philosophy
which
involved
the
minimum
scope,
so
the
various
enterprises
that
already
had
processes
to
run
these
different
things
like
trade,
finance
or
logistics
or
provenance.
G
They
did
not
want
to
rush
headlong
into
building
big
networks,
solving
all
of
their
all
of
the
activities
that
their
organizations
did.
So
you
ended
up
building
networks
just
for
trade
logistics
just
to
trade
finance,
just
to
track
provenance
of
supply
chain
of
goods
and
supply
chain,
and
so
on.
So
that's
what
called
minimum
ecosystems!
So
that's
the
reality.
You
have
these
different
kinds
of
networks
that
are
that.
G
And
they
might
also
be
built
on
different
kinds
of
platforms
underneath
built
on
either
and
they
can
be
either
centralized
decentralized
or
what
have
you
so?
G
The
what
we
end
up
with
are
a
lot
of
these
different
networks
and
by
the
way,
these
networks,
many
of
them,
can
be
either
open
access
or
they
can
also
be
permissioned,
that
is,
they
have
a
gatekeeper,
usually
it's
a
consortium
which
guards
who
can
join
the
network
and
who's
privy
to
the
information
within
it.
So
you
have
these
processes
and
you
have
the
assets
that
we
possibly
manage
within
these
respective
networks.
What
kind
of
fragmentation
are
we
talking
about
here?
G
So
there's
a
horizontal
fragmentation,
for
example,
you
can
have
a
trade
finance
networks.
I'll
give
you
two
examples.
There's
a
network
called
bitrate.
There's
a
network
called
marco
polo
beat.
It
was
built
on
a
platform
called
hyperledger
fabric.
Markov
is
built
on
platform
for
corda
and
they
both
are
trying
to
solve
the
same
problem
that
they're,
bringing
together
banks
and
exporters
and
importers
and
managing
trade
finance,
but
they
have
different
types
of
banks,
different
frontal.
G
Now
what
happens?
If
you
have,
you
want
to,
let's
say,
move
from
one
network
to
the
other,
you
have
to
make
sure
that
the
letter,
credit
rate
in
one
network
is
understandable
or
it
can
be
moved
to
a
different
network.
G
It's
not
only
possible
in
this
fragmented
world,
so
the
horizontal
fragmentation
is
different
network
solving
similar
purposes
but
handling
a
different
clientele.
So
because
people
want
to
have
flexibility
in
terms
of
what
networks
the
affiliate
affiliate
with
you
need
to
be
able
to
bridge
that
gap.
Vertical
fragmentation
is
about
a
network
like
trade
lens
managing
trade
logistics
and
your
network,
like
bitrate,
managing
trade
finance.
G
These
both
of
these
processes
are
not
independent.
In
the
real
world,
you
have
read:
finance
drive
straight
logistics,
trade,
logistics,
right,
right,
finance
and
I'll
come
to
that
in
a
second.
But
but
you
have
these
networks
built
on.
They
can
be
built
on
this
same
kind
of
platform.
The
trade
lengths
and
meat
trade
are
both
built
on
hypoglycemic,
so
these
networks
need
to
connect
with
each
other
in
order
to
be
able
to
securely
transfer
assets
or
communicate
asset
states
between
each
other.
G
That's
that's
another
kind
of
fragmentation.
So
let
me
come
to
a
concrete
example
to
the
next
slide:
okay,
just
compartmentalize
your
vision
to
the
middle
part.
I
call
these
seller
and
buyer,
but
you
can
think
of
them
as
the
exporter
and
importer
as
well.
So,
let's
say
an
export
and
importer
agree
on
a
trade
and
they
come
up
with
a
purchase
order.
Let's
say
they
have
a
common
reference
referring
to
that
particular
trade
instance.
G
And
if
you
look
at
the
network
on
the
left,
that's
managing
trade
finance,
okay,
so
it's
bringing
together
your
seller,
your
buyer
and
their
respective
banks,
and
it
is
running
a
process
of
a
letter
of
credit.
What
happens
there?
A
buyer
requests
for
letter
credit
from
its
bank,
its
bank
proposes
one.
The
seller's
bank
then
approves
it,
and
the
seller
is
then
later
supposed
to
provide
evidence
that
it
should
be
good
and
that
creates
an
obligation
in
the
network
in
the
network
system
of
record
and
the
buyer
is
supposed
to
pay
it
back.
G
So
that's
really
the
workflow
that
is
being
run
by
these
different
parties
in
a
typically
in
a
decentralized
manner,
and
the
answers
is
managing
our
letters
of
credit.
That's
that's
really
what
you
have
on
the
books,
and
we
will
need
that
letter
credit
to
be
somewhat
removable.
We
cannot
afford
to
have
it
just
be
siloed
in
this
network,
as
we've
seen
the
next
slide,
the
network
on
the
right
is
managing
trade
logistics.
G
What
that
again,
these
are
very
simplified
forms
of
what
such
networks
do
in
the
real
world.
Okay,
so
the
network
on
the
right
it's
bringing
together
sellers
and
carriers
seller
is
trying
to
booking
consignment
on
the
network
dispatching
via
carrier
and
the
carrier
is
then
uploading.
This
document
called
bill
of
plating,
which
is
proof
that
the
seller,
shipped
the
goods
and
usually
in
the
state
financing
world
bills
of
lading,
are
provided
as
evidence
that
let
us
credit
that
the
seller's
part
is
fulfilled,
that
is,
the
exporter
exported
the
goods.
G
G
A
lot
of
slides
left,
so
you
might
want
to
start
yeah
okay.
So
what
you
see
here
is
sort
of
a
join
of
the
two
networks
you
saw
on
the
previous
slide
state
finance
network
on
the
left,
trade
logistics
on
the
right,
and
there
is
a
there's,
an
inherent
interdependency
between
these
two
processes.
You
can
see.
G
Let
me
go
first,
go
to
the
extra
step
that
was
added
at
the
top
optimal
plating,
so
there's
a
bill
of
lading
in
the
trade
logistics
network
on
the
right
that
is
needed
in
the
trade
finance
network
in
order
for
the
seller
to
prove
that
they
actually
exported
the
goods.
So
how
do
you
do
that?
That
is
really
what
we
want
to
enable
by
allowing
a
cross
network
protocol
between
the
date
finance
network
and
the
trade
logistics
network.
G
That
will
allow
the
tape
finance
network
to
obtain
a
view
of
the
bill
of
lading
from
the
data
six
network,
and
then
that
is
what
allows
the
business
to
to
force
an
obligation,
the
buyer,
to
pay
the
setup
step.
Six.
You
can
think
of
another.
The
dependency
on
the
other
side,
where
the
trade
logistics
network
may
want
to
use
to
find
out
whether
a
valid
letter
exists
before
it
allows
the
seller
to
create
a
confinement.
So
these
are
real
world
use
cases
and
the
state
finance
network.
G
As
I
give
examples
you
can,
I
can
think
of
bitrate
or
marco
polo
and
trade
logistics
network.
Prime
example
is
the
trade
lens
network,
that's
run,
media
maersk
and
ibm
builder
has
built
a
technology
for
that.
So
you
can
see
what's
happening
here
is
in
effect,
the
asset
transfer,
protocol
or
satp
is
allowing
two
different
networks
to
interoperate
with
each
other,
without
forcing
those
two
networks
to
join
together
in
one,
because
that's
much
harder
thing
to
do.
G
The
the
two
networks
exist
independently
for
a
reason,
and
they
do
not
want
to
necessarily
merge
their
processes
together
for
a
lot
of
different
reasons.
This
this
these
kind
of
protocols
enable
this
to
happen,
and
just
in
case
you're
wondering
this
is
just
an
example.
Don't
don't
just
fix
it
on
the
letter
of
credit
or
the
bill
of
trading?
This
can
be
really
anything
any
kind
of
asset.
The
asset
semantics
are
restricted
to
their
respective
networks.
The
protocol
itself
is
communicating.
Information
does
not
care
about
what
do
the
assets
are.
G
It
just
cares
about
the
fact
that
information
or
answers
are
communicated
securely
between
the
two
networks
and,
in
this
case,
between
the
two
gateways
that
acting
as
spokespersons
for
the
two
networks.
G
Okay,
I'm
not
going
to
go
into
this
in
detail
just
showing
that
we
can
extrapolate
the
example
I
gave
in
the
previous
slide
to
encompass
the
entire
supply
chain
lifecycle
you,
we
added
two
networks,
food
tracking
network
and
payments
network,
and
you
can
build
a
much
bigger
workflow
spanning
all
of
these
different
networks.
You
can
even
add
a
regulatory
compliance
network
which
manages
kyc
documents
next
slide:
okay,
moving
to
a
different
kind
of
example.
So,
the
past
few
years,
maybe
past
couple
of
years,
there's
been
a
lot
of
interest
in
cbtc
right.
G
So
I'm
not
a
finance
expert.
But
what
I
understand
from
the
whole
cbd
about
the
whole
cbdc
landscape
is
that
you
have.
You
can
have
two
different
kinds
of
central
bank,
digital
currency,
the
wholesale
central
bank,
digital
currency
and
the
retail
cpc.
Wholesale
cbdc
involves
the
central
or
the
reserve
bank
of
a
country
allowing
banks
to
maintain
reserve
currency
accounts
with
it.
So
and
you
can
you
can
transfer
reserve
currency.
G
To
the
commercial
banks,
or
it
can
redeem
it
at
some
point
as
well,
you
have
retail
cbdc
networks,
on
the
other
hand,
which
involves
various
commercial
banks
and
their
and
their
clients,
and
may
also
involve
a
reserve
bank
just
acting
as
sort
of
a
regulator
or
an
auditor.
So
what
can
you?
How
is
this
a
use
case
for
for
that
you,
you
have
a
commercial
banks
that
maintain
reserve
accounts
in
jose
cbdc
network.
G
G
You
want
to
make
sure
that,
if
the
an
asset,
if,
if
some
cptc
gets
burned
in
jose
cbdc
network,
it
has
to
get
recreated
in
the
retail
cbdc
network.
Otherwise
the
protocol
ought
to
fail.
The
atomicity
is
paramount
here.
Similarly,
you,
if
a
commercial
bank
is
part
of
two
different
radio
cbdc
networks,
you
should
be
able
to
transfer
cbtc
across
its
accounts
in
between
from
one
network
to
the
other.
So
this
provides
an
example
of
asset
transfer.
G
Okay,
I
think
I
kind
of
covered
the.
This
is
what
I
was
talking
about
in
the
previous
slide.
We
have
currency
getting
transferred
from
the
wholesale
to
a
retail
cpc
network
or
across
two
different
retail
cbd.
So
I'll
stop
here
and.
E
G
Okay,
so
here
look
at
the
the
steps
highlighted
in
green
step,
6
and
step
11.
Okay,
those
are
the
steps
that
were
not
there
in
the
previous
diagram,
where
I
had
the
two
networks
disconnected
from
each
other.
So
here
what
is
happening?
You
have
a
trade
finance
network
running
a
gateway.
You
have
a
logistics
network,
running
gateway
trade
in
step,
6
state
logistics
network
asks
for
the
state
of
the
state
of
a
letter
of
credit
referencing
a
particular
purchase
order,
which
is
sort
of
refers
to
the
that
particular
trade
instance.
G
The
train
finance
network
is
supposed
to
then
provide
information
for
that
along
with
some
proof
of
authenticity.
So
that's
really
it's
it's
a
how
what
the
information
is
and
what
the
proof
is.
That's
kind
of
beyond
the
scope
of
the
gateway
between
the
gateways.
You're
communicating
this
information,
the
query:
according
you
supply,
what's
called
a
view
address,
and
then
you
get
a
view
of
the
letter
predicting
response
and
other
stuff
is
happening
behind
behind
the
gateways.
Similarly,
for
the
bill
of
reading
this,
something
similar
is
happening,
but
in
the
opposite
direction.
G
Sure
yeah,
no,
that's
fine.
I
mean
what
we
are
concerned
about
here
is
not
how
one
particular
system
is
processing
these
transactions
in
a
secure
and
atomic
manner,
but
see
the
thing
is
we.
We
have
seen
a
lot
of
these
different
systems
coming
up
a
lot
of
different
networks
emerging
the
past
few
years,
and
we
don't
think
that
there
suddenly
mh1
killer
network
on
which
everybody
will
just
move
their
asses
and
run
their
workflows.
G
S
Yeah
thanks
so
much
definitely
recognize
the
complexity
of
the
problem
and
yeah.
Sorry,
michael
prorock
speaking,
you
know
I
work
pretty
deeply
in
supply
chain
and
especially
analyzing
data
crossing
supply
chain,
financial
networks,
et
cetera,
so,
like
I
said
very
aware
of
the
problem.
But
how
would
you
differ
this
from
say,
epcis
or
existing
protocols
for
exchange
of
this
information,
kind
of
cross,
different
parties
using
standardized
rest
or
other
network
channels
and
protocols.
G
I
think
the
main
thing
we're
concerned
about
here
is
that
the
two
networks
they
need
to
conclusively
communicate
an
asset
with
all
security
and
automatic
properties
guaranteed
between
the
between
the
two
networks.
So
if
the
the.
I
G
Yeah,
so
the
security
properties
are
are
paramount,
so
I'm
not
familiar
with
the
full
details
of
the
protocol
that
you
mentioned,
but
what
we
are
trying.
K
G
S
I
I
understand
that,
but
like
gs1,
epcis
and
other
similar
protocols
in
the
trade
networks
right
I
mean
I
mean
this
has
been
done
for
years.
It's
not
to
say
there
aren't
improvements
to
be
made
there
right,
but
you
know
you
could
still
pass
all
of
that
metadata
along
with
it.
So
that's
why
I'm
just
trying
to
get
some
clarity
about
what's
being
added,
because
it
feels
extremely
complex
to
be
discussing
what
it
inherently
is
just
saying,
network
a
has
a
piece
of
information.
They
want
to
share
with
network
b
right.
G
Yeah
yeah
look,
I
agree,
so
the
protocols
like
gs1
are
sort
of
orthogonal
to
what
we're
talking
about
here.
We
do
need
a
way
for
the
asset
information
to
be
interoperable
across
different
systems,
and
this
is
something
that
say
both
like
script
are
also
trying
to
do
we're
trying
to
have
standardized
message
formats
for
different
kinds
of
assets,
but
that's
really
beyond
the
scope
of
what
we
are
trying
to
do
here.
J
No,
I
know
so
I
guess
the
question
I'm
trying
to
get
at
is
what
does
this
look
like
for
the
network's
perspective?
And
what
I
mean
by
this
is
this,
so
we
have
like,
we
have
like
an
existing
network
and
gateway.
One
is
attached
to
the
existing
networking
gear,
which
is
actually
now
attached
to
that,
has
never
been
gateways
in
between
right
and
so
the
asset
left
the
network
one
and
move
the
network
too.
Yes,
what
does
that
look
like
for
network
one's
perspective?
J
G
Okay,
there
are
two
different
ways.
I
can
answer
this
one
short
way.
This
is
how
what
the
networks
are
doing
behind
the
gateways
is
out
of
scope
for
our
current
working
group.
But
if
you
want
the
longer
answer
like
how
we
can
do
this
across
two
different
networks,
we
have
we
have
built
pocs
and
we
have
built
written
papers
on
this
topic,
so
the
networks
are
supposed
to
be
able
to
confirm
the
transactions
that
the
gateway
carrying
out.
You
cannot
have
the
two
networks
just
yeah,
but
yeah.
J
I
guess
I
mentioned.
I
understand
the
answer
yeah.
I
guess,
but
my
point
is
like,
like
the
asset
was
floating
around
network
one
and
it
had
like
ledger
things
for
network
one,
and
now
it
has
moved
to
network
two.
Does
that
look
like
a?
Does
that
look
like
a
transfer
with
no
counter
like
like
what
does
it
look
like
it's
owned
by
the
gateway
and
then
the
gateway.
G
Okay,
so
let's
do
yeah,
I
think
yeah.
B
Yeah
we
are
time
for
this
presentation.
If,
if
you
feel
like
you're
explaining
enough
use
cases,
then
we
can
go
on
yeah,
we
can
always
come
back.
Okay,.
B
D
Okay,
that's
good
all
right,
okay,
so
I'll
just
start
by
saying:
apologies
that
martin
couldn't
be
here.
So
he
was
the
my
colleague
that
worked
on
this.
D
But
okay,
let's
move
on
so
the
secure
asset
transfer
protocol
defines
a
number
of
api,
endpoints
identifiers
and
message
flows
corresponding
to
the
asset
transfer
between
two
gateways.
You
can
move
on
to
the
next
slide.
D
Please
so
we'll
start
talking
about
resource
descriptors
resources
can
be
asset,
identifiers,
public
keys
block
identifiers
or
other
items
which
are
specific
to
the
network.
You
can
move
on
please,
and-
and
here
we
have
an
example
of
what
the
draft
proposes.
A
new
media
type
will
be
used
and
the
protocol
is
sat.
D
So
the
developer
or
application
may
validate
this
entity
with
the
issuing
authority,
and
you
can
see
in
this
example,
we
have
quant
the
gateway
id
is
the
fully
qualified
domain
name
of
the
sat
compliant
gateway.
This
must
resolve
to
a
valid
ip
address,
so
here
we
say
api.gateway1.com
it
could
be.
It
could
be
anything
as
long
as
it's
a
fully
qualified
domain
name.
D
D
One
yeah,
so
the
network
or
system
identifier
is
specific
to
the
gateway
which
corresponds
to
the
target
network.
It
is
used
to
direct
sat
interactions
to
the
correct
network.
D
For
example,
we
have
trade
lands
or
the
eu
supply
chain
or
marco
polo
et
cetera,
and
finally,
the
resource,
as
mentioned
specifies,
I
think
I
lost
the
slides
yeah.
B
D
Yeah
we
were
here
so,
as
I
was
saying.
Finally,
the
resource
specifies
an
item
held
on
the
underlying
network.
The
string
must
be
meaningful
to
the
underlying
network,
so
some
examples
can
be
a
network
address,
a
data
block,
a
transaction
id,
an
alias
or
a
future
object
type
not
yet
defined
in
the
draft,
and
now
we
can
move
on
to
the
next
slide.
D
D
On
this
slide,
we
can
see
the
proposed
mandatory
fields,
which
are
you
know,
the
sat
protocol
version,
a
unique
identifier
representing
a
session,
an
incrementing
sequence,
number,
the
sat
phase,
a
resource,
url,
a
developer
urn
for
the
assertion
of
the
developer,
application
identity.
We
have
an
action
or
a
response
which
is
a
get
or
post
and
arguments.
D
And
finally,
we
have
a
message:
signature,
which
is
the
gateways
ecdsa
signature
over
the
message
and
other
relevant
attributes
may
exist
that
need
to
be
captured
for
blogging
purposes.
But
the
draft
goes
over
these
in
more
detail.
We
can
move
on
to
the
next
guys
thanks
next,
we
would
like
to
summarize
the
flows
proposed
by
the
sat
draft,
so
these
are
the
transfer
initiation
flow.
So
after
the
identity
and
asset
verification
flow
with
our
scope,
a
sender
gateway
proposes
a
session
to
a
receiver
gateway.
D
So
these,
following
slides,
will
outline
the
flows,
but
we're
mostly
going
to
cover
the
purpose
of
the
messages.
So
for
the
first
flow,
we
have
two
messages
being
exchanged,
the
initialization
request
message
and
and
the
corresponding
response.
D
The
first
one,
the
initialization
request
message
is
sent
from
the
sender
gateway
to
the
recipient
gateway,
and
the
purpose
of
it
is
for
the
clients
to
initiate
an
asset
transfer
after
receiving
the
smart
message.
The
recipient
gateway
needs
to
validate
the
profiles.
D
If
everything
is
approved,
the
recipient
gateway
constructs
the
response
message,
and
the
purpose
of
this
is
for
the
shipping
gateway
to
indicate
agreement
to
actually
proceed
with
the
operations
we
can
move
on
to
next
slide.
Please
next,
we
have
this
lock
evidence
verification
flow.
So
this
is
comprised
of
four
message:
exchanges,
so
two
messages
and
acknowledgements
for
each
of
them.
The
transfer
commence
message
is
sent
from
the
sender
gateway
to
the
recipient
gateway
and
it
signals
that
the
sender
is
ready
for
to
start
the
transfer.
D
Next,
we
have
the
lock
evidence
message
and
this
is
sent
by
the
sender
gateway
to
the
recipient
and
it
conveys
claims
of
the
locking
or
escrow
of
the
asset
on
the
respective
network
and
the
receipting
gateway
must
validate
these
claims
and
then
proceed
to
respond
with
the
lock
evidence
response
message
in
this
message:
the
recipient
gateway
indicates
that
it
accepts
the
claims
of
the
escrow
of
the
assets.
From
the
previous
message,
we
can
move
on
to
the
next
slide.
D
Finally,
we
have
the
commit
establishment
flow,
so
the
commit
preparation
message
has
the
purpose
of
indicating
its
readiness
to
begin
the
commitment
of
the
transfer.
D
The
response
to
this
message
is
an
indication
from
the
recipient
gateway
that
it
is
also
ready
to
proceed
with
the
commitment
of
the
sorry
with
the
commitment,
finalization
step,
and
then
we
have
the
commit
final
message
message,
and
this
is
for
the
sender
gateway
to
signify
that
it
has
completed
the
local
extinguishment
of
the
asset
on
its
network
and
the
receiving
gateway
can
proceed
to
regenerate
the
asset
on
its
corresponding
network
or
system,
and
the
response
to
this
must
contain
claims
that
the
asset
was
regenerated
on
on
that
respective
network
in
this
case
for
gateway
2
on
the
picture.
D
And
finally,
we
have
the
transfer
complete
message,
which
is,
for
the
sender
gateway
to
indicate
that
the
transfer
has
been
completed
and
no
further
messages
are
to
be
expected.
D
So
that's
about
it
for
this
brief
overview.
Of
course,
the
draft
covers
all
of
these
topics
in
much
more
detail.
Yeah.
Thank
you
for
listening.
B
K
B
B
All
right
hearing
none,
then
we
will
go
on
to
the
the
real
brunt
of
the
day
and
and
we
will
go
back
to.
B
Let's
remind
ourselves
of
the
purpose
of
a
buff
right,
so
there's
there's
a
number
of
things
that
we've
already
sort
of
brought
up
and
I'm
going
to
reiterate
those
in
a
minute,
but
is
the
ietf
the
right
place?
Is
it
a
tractable
problem
all
those
types
of
things
I'll
note
that
so
far
you
know.
We've
certainly
talked
about
a
number
of,
I
think
critical
issues.
One
being
you
know,
is
voluntary
or
involuntary
transfers
which
of
those
two
are
in
scope.
B
Is
it
tractable
to
only
do
voluntary?
I
think
was
brought
up
at
one
point
when
it
is
voluntary
who's
actually
doing
the
request.
There
seemed
to
be
some
confusion
about
that.
Does
the
asset
owner.
B
Request
does
the
network
of
the
asset
owner
initiate
it
does
the
gateway
initiate
it?
How
is
that
you
know
locking
mechanism
to
some
extent,
that
is
in
the
out
of
scope
section
for
how
the
the
proponents
have
structured
their
their
initial
stab
at
a
working
group
charter.
B
The
other
big
one
that
was
brought
up
was
trust,
really
there's
sort
of
two
types
of
trust
that
I
I
heard
described.
One.
B
Client
to
gateway
again
that
sort
of
declared
out
of
scope
and
then
there's
trust
between
gateway
to
gateway,
and
you
know
it
sounded
like
there
was
some
sort
of
legal
frameworks
that
might
be
needed
to
actually
back
up.
You
know
trade,
memberships
and
things
like
that,
but
the
key
question
is:
can
a
protocol
actually
be
trusted
if
we
don't
define
complete
end-to-end
example
now,
so
that
was
sort
of
brought
up
that
if
you
only
define
the
gateway
to
gateway
now,
then
you
don't
get
the
binding
that
you
get
with.
B
You
know
trying
to
deal
with.
How
does
that
work
inside
of
a
particular
network?
So
with
that
I'll
open
the
floor
to
you
know
general
discussion
at
this
point.
Ideally
we'll
have
time
later
for
a
charter,
but
I
think
the
general
discussion
will
decide
whether
we
should
do
a
charter
or
not
so
who
wants
to
start
the
discussion.
I
T
So
I
think,
there's
a
few
topics
that
have
come
up
that
seem
to
me
to
indicate
that
we're,
maybe
implicitly
making
some
assumptions
about
the
nature
of
the
networks
at
the
source
and
recipient
ends.
In
terms
of
like
there's,
some
concept
of
being
able
to
lock
an
asset
or
to
be
able
to
support
an
asset
disappearing
from
the
network,
or
something
like
that,
and
so
it
might
be
useful
to
explore.
T
L
I
can
try
answer
that
then,
so
so
so
in
the
protocol
itself,
there's
a
thing
called
a
blob
called
a
lock.
We
just
is
called
lock
evidence
right
and
what
that
conveys
is
evidence
that
that
the
gateway
one
has
somehow
disabled
temporarily
this
asset
in
network
one
and
this
evidence
has
to
be
delivered
to
gateway
two
and
gateway.
Two
has
to
read
this
and
assess
whether
there's
an
acceptable
level
of
locking
or
disablement
and.
L
Yes,
so
so
in
phase
zero
gateway,
one
has
to
tell
gateway
to
this
is
my
type
of
network.
This
is
the
blob,
the
locking
blob
structure
that
I
will
use.
Do
you
understand
my
blob?
If
you
don't,
then
forget
about
the
transaction?
L
If
you
do
then
proceed,
and
so
the
question
for
us
right
now
is:
do
we
define
that
locking
block
we
haven't
decided?
Yet
maybe
it's
in
scope
right
now,
I
kind
of
feel
it's
out
of
scope,
because
it's
just
too
much
work
to
do,
but
but
absolutely
great
question,
because
that
is
one
of
the
core
sort
of
features
of
that
we
need
to
address.
What
is
this
evidence
blob?
Look
like.
K
O
Okay,
dave
faber.
The
bottom
question
on
this
slide
is:
is
the
working
group
likely
to
succeed,
and
I'm
gonna
answer
this
from
the
5218?
What
makes
for
a
successful
protocol
right,
because
you
can
argue
that
a
working
group
would
be
successful
if
it
publishes
an
rfc?
That's
not
my
definition.
O
My
definition
is
a
working
group
is
successful
if
it
makes
a
standard
that
actually
gets
deployed
and
used
okay,
and
so
that's
how
I'm
interpreting
that
question,
and
so
I'm
gonna
give
my
thoughts
as
to
how
to
make
it
not
fail.
I
mean
by
by
changing
the
scope,
but
I
can't
prove
that
it
will
succeed,
because
that
requires
information
I
don't
have
about
whether
people
would
actually
use
it
where,
presumably,
you
guys
know
that
better
than
I
do
in
terms
of
the
use
cases.
L
O
So
I'm
going
to
trust
you
that,
yes,
as
long
as
you
do
certain
things
that
people
will
deploy
it
but
yeah,
I'm
looking
to
yeah
yeah,
but
I
would
say
that
as
long
as
you
scoop
things
down,
I
think
it's
tractable
as
long
as
you
scoop
things
down
by
changing
some
terminology
or
maybe
changing
the
scope
or
whatever
wording,
to
still
tailor
it
specifically
to
the
use
cases
that
you
walk
us
through.
O
So,
for
example,
if
you
don't
mention
you
either
don't
mention
legal
recognition,
the
option
a
or
b
scope
it
to
where
there's
a
common
agreement
within
the
system
of
what
that
legal
framework
is,
and
that
could
be
a
treaty.
It
could
be
a
single
jurisdiction,
whatever
it
is.
As
long
as
there's
general
agreement
on
what
that
means,
then
you've
removed
the
part
that
makes
it
not
one
of
the
parts
that
makes
it
not
detractable.
O
Okay,
I
think,
within
the
examples
that
you
showed
in
the
use
cases
document,
I
think
that
assumption
holds,
but
you
know
better
than
I
do
okay.
So
as
long
as
you
scope
it
down,
it
says:
okay,
there's
a
common
agreement
on
what
legal
ownership
means.
Okay
and
then
you
scope
it
down
to
say
voluntary,
only
okay,
because
you
could
probably
do
that
in
the
context
of
a
particular
definition
of
what
the
legal
jurisdiction
is.
O
All
the
entities
are
equal
in
the
same
jurisdiction
whatever,
then
you,
then
I
think
that
the
architecture
and
the
protocol
can
probably
apply
okay.
What
I
don't
know
that
it
was
kind
of
a
question
that
phil
kind
of
begged
right
was
should
something
where
a
legal
thing
come
along
and
say
if
I
can't
do
involuntary
I'll
just
ban
it
right.
O
I
can't
comment
on
whether
that
will
happen,
because
I
don't
know
enough,
but
as
far
as
taking
the
architecture
make
it
be
tractable
from
an
engineering
point
of
view
right,
because
often
engineering
principles
and
legal
infrastructures
may
not
be
consistent
with
each
other
and
we're
engineers
here.
So
I
would
scoop
it
down
to
say:
okay
voluntary
only,
but
there
is
a
common
understanding
of
what
the
legal
framework
is.
Okay
and
within
that
narrow
scope,
then
I
think
it's
tractable
whether
it's
going
to
be
deployed.
I
leave
that
to
you
and
to
others
to
comment
down.
B
Closer
to
the
mic,
please-
and
certainly
you
know,
one
thing
that
I
would
love
to
hear
would
be.
Are
there
other.
B
Here
but
oh
two,
excellent
colin.
P
Actually,
those
two
hands
go
exactly
to
my.
I
I'm
actually
on
the
problem
means
solving.
So
I'm
trying
to
get
my
head
around
this
one,
a
little
bit
more.
I'm
thinking
about
that
aspect
of
of
the
buff,
and
I
I
get
how
this
would
connect
the
networks.
I
I
I
got
your
example
with
the
finance
network,
but
you
know
the
next
thing
we
heard
at
the
mic
is
yeah,
but
that's
already
a
well-solved
problem,
so
I'm
trying
to
get.
P
I
think
it
would
help
you
develop
energy
and
get
things
here
if
you
could
say
more
about
specific
use
cases
that
are
fairly
concrete,
where,
where
this
type
of
technology
could
help,
and
particularly
ones
that
weren't
already
solved
today
and
would
be
people
who
really
wanted
to
deploy
that.
So
I
don't
know
if
you
can
comment
anymore
on
that
now
or
whether
that
something
happens
later.
But
that's
that's
something
where
I
feel
like
I
I
I
could
learn
more
and
it
would
be
really
compelling
to
help
me
understand
like
yeah.
G
I
can
comment
on
what
we
have
done,
so
there
are
things
in
the
pipeline
that
I
can't
comment
about.
But
what
I
can
comment
about
is,
if
you
check
you'll,
find
a
couple
of
press
releases
from
last
december,
where
we
conducted
an
experimental
poc
involving
the
bank
of
france
and
hsbc,
where
we
they
both
have
networks
built
on
different
technologies.
One
of
them
has
network
managing
cptc
and
a
coupon
payment,
workflow
and
securities
and
stuff
like
that
on
a
network
based
on
hyperledger
fabric.
G
Another
one
had
something
similar
slightly
different
workflow,
but
still
managing
cpdc
and
security's
assets
on
the
corda
on
the
corder
platform,
and
we
built
several
different
use.
Cases
connecting
the
two
different
networks
be
able
to
atomically
exchange
assets
across
each
other
to
be
able
to
transfer
asset
from
one
network
to
the
other,
to
be
able
to
communicate
views
of,
let's
say
a
coupon
state
from
one
network
to
other
things
like
this.
So
we
did
this
and
it
was
perceived
to
be
a
pretty
success
as
an
experiment.
G
B
G
So
the
thing
that
we
are,
we
specifically
tried
to
solve
again,
giving
an
example
of
the
system
that
I
finally
maintained
wrong.
It's
called
vivo
and
it's
it's
about
enabling
two
different
decentralized
ledger
networks
to
be
able
to
interoperate
with
each
other,
so
part
of
that,
and
really
what
we're
focusing
on
with
satp
is:
how
do
you
enable
the
the
two,
the
the
proxy,
the
spokespersons
of
the
of
two
networks
to
be
able
to
securely
communicate
their
assets
or
asset
state
views
across
each
other?
That's
a
part
of
the
bigger
scheme.
G
Now,
of
course,
when
we
we're
building
something
for
for
a
client
we're
going
to
be
doing
the
whole
entering
process.
So
that's
really
not
part
of
the
scope
here,
but
definitely
a
key
part
of
that
is
the
ability
for
two
gateways
of
two
networks
to
be
able
to
communicate
with
each
other
and
gateway
is
really.
I
think
somebody
released
the
question
earlier.
What's
the
gateway
yeah,
how
how
do
what
how
the
network
structure
involves?
We
don't
have
to
go
into
that,
but
to
be
able
to
communicate
something
institutionally
between
two
networks.
G
Protocol,
we
are
talking
about
whatever
use
case
you're
talking
about.
G
This
is
not
really
a
solved
problem
in
the
in
the
decentralized
network
space.
You
have
two
different
decentralized
networks.
How
exactly
are
they
supposed
to
trust
what's
happening
in
other
networks,
because,
in
effect,
what
we're
trying
to
do
is
enable
a
cross-country
transaction.
That's
flying
to
different
networks,
if
you
think
of
it
absolutely
that
way,
you
need
something
like
this
and
it's
a
problem
that
you've
been
trying
to
solve
the
past
few
years
and
you
made
take
a
lot
of
steps
towards
it.
G
B
A
Okay,
is
it
all
better?
Yes,
that's
better!
Thank
you.
Okay.
Sorry.
So
my
question
is
in
the
area
of
like
end-to-end
and
gateway
to
gateway.
As
I
understood,
the
problem
protocol
will
be
developed
gateway
to
gateway.
Only
is
that
does
that.
B
B
Okay,
thank
you
in
the
meanwhile
I'll
read
that
off
in
a
minute,
roman.
E
Hi
roman
union
carnegie
mellon,
just
kind
of
thinking
that
you
presented
both
with
the
use
cases
and
the
protocols.
The
protocol
in
particular
flow
was
super
helpful.
If
we
think
about
this,
as
just
as
a
distributed
systems
problem-
and
we
made
the
assumption
that
I
think
we
talked
about
before,
which
is
the
asset,
is
opaque,
we're
assuming
nothing
about
what's
behind
the
gateway,
and
we
said
we
need
a
three
pc
and
some
security.
E
G
Acid:
that's
a
property
so
cross
network,
atomicity,
so
acid,
transfer
and
acid
exchange.
You
need
to
make
sure
that
if
an
acid
gets
burnt
in
one
network,
it
has
to
get
recreated
in
the
other
or
it
doesn't
get
burnt
and
acid
swap
is
either
it
changes
hands
in
both
ways
or
in
either
direction.
That's
the
problem
to
me.
G
L
B
Okay,
bernie
did
you
want
to
try
again,
I
think
you
came
back.
You
left
and
came
back.
A
J
So
I'm
concerned,
we
are
following
your
trap
here,
which
we
often
see
at
itf,
which
is
a
few
people
interested
in
something
come
to
us
for
something
which
is
embedded
in
a
much
larger
context
and
which
can't
be
understood
without
that
context,
and
we
don't
have
that
context
and
we're
asked
to
define
this
like
very
narrow
piece.
So
it's
like
we're
building
a
rocket
and
could
you
please
design
the
upholstery,
and
that's
concerns
me
so
maybe
they're
like
I
like
around
this
room
and
I
see
a
lot
of
people.
J
I
already
know-
and
I
know,
did
not
work
in
this
field
and
so
like
my
question
is
like
who
are
the
relevant
people.
I
have
in
this
room
for
us
to
actually
solve
this
problem
in
a
reasonable
way,
and
I'm
not
going
to
listen
to
you
when
you
tell
me
that
I
don't
have
to
worry
about
the
embedding.
B
Hold
on
sorry,
that's
it.
Can
I
ask
a
a
a
clarifying
question.
Yes,
question
totally
get
it
right
and
the
other
thing
that
people
do
all
the
time
is
they
come
to
the
ietf
and
they
bring
such
a
monstrosity
that
we
curve
it
down
into
a
very
minimum.
So
you
know
finding
that
middle
line
is
hard.
So
do
you
have
a
notion
in
your
head
of
what
you
think
the
bare
minimum
pieces
would
be
in
order
to
you
know
is
it
includes,
like
you
know,.
N
J
Have
no
idea,
because
I
was
completely
unable
to
understand
the
context
so
like,
and
it's
precisely
what
I'm
saying
is
that,
like
what
is
necessary
here
is
the
people
who
are
okay.
Consumers
of
technology
have
to
come
here
and
have
to
work
with
us
to
develop
it
if
we're
gonna
develop
it
because
otherwise
we're
developing
according
to
requirements
which
thought
they
don't
like
you.
Not
they
don't
trust
you,
but
they
came
from
you
and
did
not
come
from
that
community
as
well.
R
So
I'm
not
I'm
not
sure
what
the
problem's
tractable
from
the
way
I
look
at
it
and
it
may
be.
It
means
that
I
just
think
that
it
needs
to
have
more
constraints
as
to
under
what
scenarios
this
might
fail
and
and
exactly
I
think,
those
listed
out
and
what
happens
because
that's
the
thing
that
I'm
most
worried
about
is
what's
the
trust
relationship
between
the
two
gateways
effectively
and
if
one
of
them
says
actually
I've
failed
to
commit
this.
R
Can
you
roll
back,
for
example,
what
stops
that
gateway
from
actually
committing
that
asset
to
that
new
network
and
asking
another
gateway
to
to
effectively
keep
it
alive,
or
vice
versa?
Those
sorts
of
questions.
R
And
because
I
think
that
again
back
earlier
earlier,
I
think
three-phase
commit
guarantees
the
acid
properties
under
certain
boundary
conditions
and
in
the
real
world
those
sometimes
fail,
and
I
think
in
the
real
world
when
those
fail,
you
have
a
human
come
and
has
to
come
sort
things
out.
So
I
think
if
you
scope
those
things
down
and
say
it
works
in
the
mainline
cases
here.
These
are
the
cases
where
things
can
go
wrong
and
you'll
need
some
outer
bound
way
of
resolving
this.
R
K
L
B
Okay,
let's
put
it
another
way,
you're
sort
of
asking
where
does
like
the
legal
framework
have
to
take
over
you
know
to
cover
the
cases
where
the
protocol
can't
solve
right.
That
would
be
another
way
of
looking
at
it.
Yeah
bernie
wrote
his
question,
so
let
me
read
it.
The
first
presenter
mentioned
the
scope
to
be
only
gateway
to
gateway.
Is
it
meant
that
the
protocol
will
be
designed
to
only
work?
You
know
gateway
to
gateway,
or
will
there
be
room
for
a
possible
later
extension
to
make
this
work
end
to
end?
L
E
L
M
Hi
phil
handbaker,
I'm
not
sure
that
us
the
way
that
you're
scoping
it
at
the
moment
taking
the
networks
out
completely
depending
upon
the
properties
of
that
network.
M
You
know
those
people
are
using.
One
percent
of
the
global
energy
they've
constructed
a
ponzi
scheme
that
has
destroyed
at
least
a
hundred
billion
dollars
of
wealth
and
is
in
the
pr
and
those
are
relevant
when
it
comes
to.
If
the
itf
starts
a
working
group
that
may
be
take,
that
is
going
to
be
used
as
a
way
of
endorsing
that
whole
field,
and
I
think
that
the
itf
has
to
be
very
careful
before
it
takes
any
step
that
would
be
interpreted
in
that
way.
B
Certainly,
interpretation
brings
an
interesting
question,
but
I'll
note
that
we
already
have
invented
something
that
has
helped
bitcoin
and
other.
You
know
realms
work
right.
It's
called
tcp
they're
communicating
their
transactions
over
it.
So
the
you
know,
the
point
of
this
group
was
to
transfer
digital
assets
and
yes,
it
could
be
misused
for
something
that
we
don't
like
in
your
camp.
Otherwise,
but
michael.
D
Sorry,
just
to
add
to
philip's
question
just
for
an
answer.
I
think
during
our
conversations
for
the
drafts
and
everything
we
knew,
there
were
networks
like
bitcoin
or
ethereum,
which
have
this
sort
of
negative
impact
on
the
world
at
the
moment,
but
you'll
notice
that
in
our
use
cases,
and
mostly
in
our
conversation
as
well,
we
focus
on
networks
based
on
technologies
such
as
hyperledger
fabric
or
corda,
or
even
other
things
which
are
more
on
the
permission
side
of
networks
rather
than
permission
less,
but
let's
not
go
into
that
direction
necessarily.
D
G
S
Yeah,
my
michael
perak,
here
again,
I
just
wanted
to
get
some
clarity
because
I
think
you
know
for
sure
I
look
at
the
general,
both
goals
c
for
sure.
Yes,
problem
needs
solving
right,
it's
something
I
encounter
constantly
right,
but
you
know
one
of
the
clarity
things
for
me
is
every
example.
You've
given
has
been
either
hyper
ledger
or
some
you
know,
dlt
type
approach,
etc.
S
S
L
That's
a
good
question,
so
we
want
to
make
it
so
that
in
fact,
instead
of
a
network,
it
could
be
a
monolithic
database
on
one
side.
Okay,.
I
S
That
might
be
helpful
to
bring
into
the
examples,
because
if
it
and
and
if
we
are
talking
because
so
much
has
come
back
to
like
dlt
type,
you
know
or
ledger
type
examples,
and
things
like
that,
and
obviously
there
has
to
be
some
bridge
between
that
kind
of
stuff,
because
that
stuff
is
out
there
right.
It's
not
gonna.
It
may
not
go
away
overnight.
It
may
never
go
away.
You
know
we.
L
Don't
know
actually
actually
alex.
I
know
martin
has
explained
in
some
of
the
meetings
in
the
part
of
this-
that
in
fact,
these
guys
and
you
can
have
have
a
deployed
solution.
One
side
is
in
fact
a
monolithic
sort
of
banking
computer.
It's
not
even
a.
S
Okay,
that
that
that's
helpful,
have
you
thought,
through
things
like
these
centralized
identifiers,
which
are
now
a
spec,
and
things
like
that
right,
because
there
are
aspects
touching
on
this
and
other
standards,
orgs
also
aspects
I
mean
you
know
I'm
looking
at
this
and
going
well.
Why
wouldn't
I
just
make
a
bunch
of
jobs
or
cwts
right,
exchanging
these
claims
so.
L
G
L
N
So
mun
thompson,
I'm
looking
at
these
questions
here,
and
these
are
good
questions.
These
are
the
usual
questions.
Unsurprising
bernie
asked
a
question
about
end
to
end
and
that
was
very
quickly
waved
out
of
scope.
N
The
discussion
here
we
talked
about
locking
and
various
other
things
and
being
able
to
assert
to
the
your
counterparty
that
you
have
a
lock
on
the
thing,
that's
being
transferred.
That
to
me
suggests
that
the
the
scope
is
a
lie.
N
I
I
have
a
very
hard
time
believing
that
you
can
implement
something
like
this
by
putting
that
stuff
out
of
scope
when
it
is
so
fundamental
to
the
protocol
here
that
we're
talking
about
delivering.
N
But
I
don't
want
you
to
answer
that
question.
I
think
that
until
we've
dealt
with
the
question
of
constituency,
none
of
this
matters
very
much.
So
I
see
a
few
people
here
who
are
interested
in
pushing
this
forward,
but
I'm
not
seeing
evidence
that
the
people
who
would
be
deploying
it
here
in
the
room
and
that
that
concerns
me
so
the
point
I
can
made
earlier
until
that
is
answered.
I
think
that
the
third
question
and
the
sixth
question
have
answers
of
no
right
now,.
N
B
C
C
I'm
rebecca
I'm
with
14
bis,
we
deal
with
supply
tracking
and
we
mostly
deal
with
at
the
moment
hybrid
assets.
So
basically
like
actual
parts
that
have
like
you
know,
ids
associated
with
them
and
ownership
is
like
digitally
recorded
and
like
a
lot
of
times.
Currently
that,
with
the
clients
we
manage
like
it's,
everything
is
on
like
a
hyper
ledger
like
fabric
network,
and
everything
is
internal,
but
we
can
see
like
in
the
future.
C
If,
like,
for
example,
our
clients
are
transferring
assets
that
we
did
not
create
for
them,
for
example
like
off
another
client
of
theirs
like
network.
This
could
be
something
that's
like
interesting
to
us
and
we
would
want
to
establish
a
protocol.
Obviously
we
would
have
to
work
with.
However,
their
clients,
like
you
know,
are
dealing
with
how
they
handle
like
their
own
assets
and
data,
and
we
would
have
to
build
a
specific
set
of
rules
for
our
gateways.
But
this
is
definitely
something
we're
interested
in.
I
Hi,
I'm
john
robotham,
I'm
with
ernst
young,
and
we
deal
with
these
problems
of
either
permissioned
blockchains
talking
to
other
permission,
blockchains
or
permissionless
ones.
Talking
to
permissionless,
or
you
know
you
name
the
the
combinatorics,
it's
the
these
are
our
you
know
critical
problems.
I
would
also
point
out
that
a
fair
amount
of
the
threat,
the
theft
going
on
right
now
in
in
permissionless
blockchains,
happens
at
the
gateways.
I
Q
So
I
apologize,
I
also
have
an
authentication
challenge:
ori
steele
with
the
transmute
I
work
in
the
digital
identifier,
cryptographic
assertion
space
in
involved,
both
in
permission,
lists
and
permission,
blockchain
technology
agree
with
the
comments
john
made,
regarding
where
we
see
fraud
and
attacks
in
the
space.
You
know
on
the
end
points,
certainly
a
concern.
Q
Certainly
the
the
industry
will
solve
for
atomic
swap
blockchain
interoperability
between
networks
awkwardly
by
itself,
in
absence
of
guidance
from
a
group
of
folks
who've
organized
to
create
you
know,
clarity,
I
don't
know
how
much
representation
in
my
first
time
at
ietf,
I'm
not
sure
how
much
how
many
financial
institutions
come
to
these
kinds
of
meetings
to
provide
input,
but
they
are
relevant
to
the
questions
you
know
on
the
screen.
I
I
think
my
main
interest
in
in
this
work
is
sort
of
seeing
cyber
physical
supply
chain
cases.
Q
So
you
mentioned
some
of
the
hybrid
physical
digital
cases,
seeing
folks
coming
together.
Thinking
more
about
those
use.
Cases
is
important
to
my
business
and
you
know
we
intend
to
deploy
technologies
that
support,
providing
physical
supply
chain
traceability,
but
also
non-physical
digital
supply
chain
software
supply
chain
use
cases.
I
think,
there's
it's
a
good
amount
of
overlap
in
some
of
the
pieces
that
you've
proposed.
I
also
think
it
is
a
challenging
space
and
that
scoping
is
even
is
very
hard
in
this
space.
Q
S
S
This
stuff
is
going
on
like
we
see
this
actively,
I
do
a
lot
of
work
in
you
know
everything
from
human
trafficking
to
looking
at
forced,
labor
and
stuff
like
that,
and
we
see
illicit
transactions
over
these
networks
as
well
as
occasionally
legitimate
transactions
outside
of
the
you
know,
permission
type
stuff
right
in
private
network,
so
it
is
something
where
we
have
to
get
some
kind
of
structure
and
order
on
it
if
for
no
other
reason
than
to
understand,
what's
going
on
and
to
enable
some
kind
of
regulation
around
this
stuff,
so
it's
definitely
something
willing
to
participate
in
willing
to
make
sure
we're
coding
in
accordance
with
standards
should
they
emerge
out
of
this
stuff.
S
So
it's
so
definite
support,
understand
the
problem
set,
but
it's
a
dangerous
area
for
sure
for
a
lot
of
reasons.
So
thank
you.
J
B
S
Yeah
and
I'll
follow
up
on
that
and
eric,
and
I
think
that's
an
excellent
question
and
is
part
of
my
concern
like
I
think
the
change
actually
is
a
good
thing,
because
I
think
bringing
strong
technical
merit.
That's
internet,
interested
cross
boundary,
that's
important.
What
I'm
not
sure
on
is
how
we
actually
defined
the
scope
well
enough
to
actually
sit
down
and
write
about
this.
S
I
am
not
a
hundred
percent
convinced
that
ietf
putting
a
standard
on
this
will
be
very
helpful
unless
everyone
actually
follows
the
standard
right
or
we
see
a
large
degree
of
adoption,
and
this
is
one
of
the
things
that
we
see
supply
chain
wise
and
you
know
it's
come
up
in
other
conversations
at
ietf,
just
thinking
about
things
like
faa,
right,
an
exchange
of
information
around
drones
and
other
types
of
things
when
you
start
getting
into
these
international
boundary
type
questions
sure
it's
great
if
everyone's
respecting
you
know
some
north
american
model
or
north
america,
plus
europe
etc,
but
the
world
is
much
much
bigger
than
that
right,
and
so,
if
we're
only
seeing
a
partial
buy-in
from
this,
that's
concerning
this
goes
a
bit.
S
To
is
the
ietf
the
right
place
to
solve
it,
and
I
am
personally
not
100
sure
of
that
right.
There,
I
think,
from
the
technical
side
and
defining
these
protocols,
maybe
but
these
network
protocols
for
exchanging
information
in
an
asset
type
way
they
kind
of
exist
right.
So
that's
a
little
bit
of
the
concern
I
have
is.
Maybe
this
is
more
of
a
regulatory
and
a
policy
question.
Maybe
this
is
an
iso
thing
etc
and
I'm
not
100
sure
yet,
but
I'm
highly
interested
in
the.
L
S
S
We
see
that
in
agriculture,
all
the
time
like,
oh
yeah,
we've
got
such
and
such
going
on,
but
in
reality
all
you're
doing
is
charging
the
farmer
a
bunch
of
money
for
very
little
value.
So
if
we
can
break
that
with
a
totally
open,
non-proprietary
standard
cool,
but
if
that
just
defines
something
so
people
can
masquerade
and
say
they're
doing
interop.
That
concerns
me
as
well
right.
G
Just
want
to
comment
on
where
this
is
being
used
or
applied
so
at
least
under
the
hyperledger
umbrella,
the
flagship
there's,
a
interoperability
project
called
cactus,
and
then
we
have
our
project
under
hyperlogic
labs
for
weaver.
These
are
two
most
mature
interoperability
projects
for
permissioned
blockchains.
Really
because
that's
the
point
of
view
we
were
coming
from
and
within
cactus,
you
already
have
an
implementation
of
the
protocol
that
was
talked
about
here
and
we
were.
G
We
have
we've
been
building
this
concept
of
using
gateways
to
communicate
views
as
well
as
exchange
assets
or
transfer
assets,
and
we
intend
to
comply
with
whatever
protocol
is
agreed
upon
for
the
workings
so
and
in
the
recent
past
we've
had
basically
every
week
we
have
a
lot
of
meetings
with
cactus
weaver
with
the
eea
with
other
various
big
players
in
the
blockchain
world,
and
everybody
is
interested
in
that
everybody
wants
to
see
it.
B
Q
So
it's
or
still
from
transmute
I'm
queued
to
respond
to
eric,
I
think
you'd
know
better
than
I
would
er
how
the
ietf
might
handle
a
situation
like
like
this.
This
proposal,
the
only
thing
I
can
say
is
that,
given
the
nature
of
the
problem,
I
think
engagement,
international
engagement
from
regulators.
Q
You
know
banks,
financial
institutions
and
technology
developers
would
all
be
critical
and
it
should
be
observable
that
they're
there
to
participate
in
the
work
and
that
if
those
constituents
are
there
and
they're
engaged
they're
contributing
to
the
work
and
their
contributions
are
visible.
I
believe
the
ietf
can
produce
interoperable
standards
that
address
the
financial
industry's
needs
for
transport
protocols.
Q
However,
I'm
not
I'm
not
sure
we're
fully
seeing
them.
You
know
enough
of
them
them
here
today.
I
don't
know
what
the
criteria
for
observing
the
presence
of
those
kinds
of
entities,
but
I
think
right,
seeing
regulators
stepping
up
to
the
mic.
You
know
saying
things
like
you
know:
we'd
like
to
we'd
like
to
participate.
We'd
like
to
comment
we'd
like
to
see
you
know
my
naive
perspective
on
it,
as
I
would
like
to
see
their
their
comments
on
the
proposal.
U
Hi,
it's
ducard.
I've
got
lots
of
different
affiliations,
but
for
this
time
I'm
going
to
wear
the
one
that
will
paint
a
target
on
my
back
in
this
audience
a
little
start-up
called
quotidian
engineering
and
development,
which
is
explicitly
a
distributed
ledger
technology.
Startup
I've
been
involved
since
the
days
of
the
cypherpunks
in
1993..
U
I
think
before
this
work
can
effectively
be
addressed.
Oh
okay
use
case.
The
first
use
case
that
I
was
involved
in
is
I
have
the
the
sad
honor
of
consuming
the
most
expensive
bag
of
popcorn
in
history,
which
I
bought
with
bitcoin
from
a
vending
machine
that
was
prototyped
by
one
of
my
co-inventors.
U
U
To
see
this
to
see
this
addressed
in
the
itf,
we
have
to
first
off
take
on
the
visceral
response
that
it
generates
amongst
ietfers,
if
you
say
blockchain,
you
just
painted
a
big
honking
target
on
yourself
and
part
of
that
is
legitimate,
because
proof
of
work
is
evil.
Okay,
but
proof
of
scope.
I'm
not
sure
is
the
answer.
Proof
of
stake-
I'm
not
sure,
is,
is
you
know
the
answer
either
right?
U
L
G
L
Thomas
all
right
so
so,
no
absolutely
agree.
This
technology
is
here
to
stay
in
some
form,
another
it's
going
to
evolve
and
I
believe
the
itf
needs
to
step
to
the
plate
and
say
yeah.
We
can
address
some
parts
of
this
intro
problem
at
the
very
minimum
and
it's
almost
like
well,
if,
if
the
itf
is
not
going
to
do
it,
who
else
is
going
to
do
it?
Because
each
of
these
you
know
communities,
you
know
they're
still
stuck
on
like
designing
their
own
internal
networks
and
basically
promoting
their
own
tokens.
L
Okay-
and
this
is
you
know,
this
is
back
with
the
you
know,
expected
engineers,
the
speculators
right.
So
so
that's
why
we
came
to
the
itf,
because
I
think
this
is
the
home
of
interop.
I've
been
telling
now
just
to
give
you
a
picture,
there's
a
whole
generation
of
under,
I
don't
know,
25s,
who
don't
know
the
itf
produced
tcp
ip
tls,
ipsec
x509
they've,
like
never
heard
of
the
itf
before-
and
this
is
a
problem
not
not
for
me,
it's
like
everybody,
and
I
think
this
is
a
good
this.
L
B
B
We
have
some,
I
think,
options
about
where,
where
what
would
be
most
useful
to
you
two
to
make
a
decision
in
the
long
run
we
can
start
doing
polling
now
or
you
know
next
up
technically
on
the
agenda
was
a
charter
discussion,
but
it
feels
like
we
might
want
to
preempt
that
with
usability
and
other
stuff,
so
we'll
let
them
confer
and
they're
developed
to
markle
tree's
head
so
give
us
one
minute.
While
we
pause
for
a
moment,
I
will
note
that
I
did
pull
up
some
of
the
actually.
F
Thanks
davidskenazi
internet
architecture,
enthusiast
so
just
to
respond
to
some
of
the
points
here
before
the
the
main
question
that
I'm
noodling
on
is
whether
the
itf
is
the
right
place
for
this,
because
the
itf
has
some
advantages
in
that
yeah.
It's
free,
but
you
know
so
is
github
and
it
has
some.
You
know
ipr
policies
which
are
great
and
it
has
some
credibility.
F
But
then
the
question
is:
are
we
is
like,
like
we
were
saying
there
seems
to
be
some
use
cases
that
make
sense,
but
at
the
same
time,
the
optics
of
us
enabling
technologies
that
folks
have
concerns
about,
could
damage
the
credibility
of
the
ietf,
and
that
is
somewhat
of
a
concern
that
is
worth
taking
into
consideration.
So
I
I
don't
really
have
a
question
for
you.
It's
just.
I
think
that
should
be
part
of
our
decision
on
whether
we
want
to
adopt
this
work
on
what
the
consequences
would
be
long
term.
F
A
B
D
If
I
can
add
to
that
comment,
so
I've
been
working
with
blockchains
for
the
past
like
five
to
six
years,
and
I've
started
with
ethereum
and
bitcoin,
but
then
moved
on
to
more
refined.
You
could
say
networks,
and
each
of
them
is
evolving
on
top
of
the
past
one.
So
what
if
bitcoin
was
very
you
know,
disruptive,
but
also
destructive?
You
have
a
theorem
which
you
can
take
into
a
private
consortium
and
build
something
amazing
with
it
without
any
of
the
environmental
concerns.
D
Just
because
of
how
you
know
how
capable
the
technology
is
to
adapt,
but
the
problem
with
these
networks
developing
on
them.
I
had
to
learn,
like
you
know,
10
different
apis
to
be
able
to
interact
with
all
of
these,
and
none
of
them
have
any
like
standards
between
each
other.
It's
very
hard
to
take
what
you
learn
from
one
network
and
sort
of
say:
okay.
This
is
similar
to
to
this
other
one
and
the
problem
that
all
of
these
networks
face
is
that
they
sort
of
especially
the
permissionless
ones.
D
They
sort
of
like
have
this
ego
that
it
should
be
decentralized.
There's
no
governing
authority
in
this
blockchain
space,
especially
and
also
in
networks,
so
I
think
a
group
like
ietf,
which
is
you,
know
worldwide,
and
has
this
sort
of
it's
not.
You
know
a
single,
a
single
entity,
it's
multiple
companies
coming
together.
This
would
be
like
a
perfect
use
case.
I
think,
for
an
institution
like
this.
An
organization
like
this
yeah.
B
Thank
you
all
right,
roman,
are
you
records.
H
K
Fallout
city,
so
the
questions:
can
we
talk
a
bit
like
like
a
couple
of
minutes
about
the
charter
and
and
see
if,
if
that
gives
us
more
clarity
about
I'm
answering
one
of
the
first
two
questions
of
the?
If
the
buff
about
whether
this
is
the
right
place
and
whether
there's
a
problem,
we
could
solve
yep.
P
I'll
make
this
really
brief.
I
I
think
the
itf
should
be
very
hesitant
about
not
doing
work
because
of
a
value
judgment
about
how
it'll
be
used.
A
ton
of
the
use
cases
I
heard
were
permission
block
change,
which
probably
have
very
low
energy
impact
on
the
world.
Obviously,
there's
some
use
cases
that
very
high
energy
impact,
but
if
we
start
going
down
that
path,
there
is
no
into
it,
and
you
know
it
sure
is
a
good
thing.
P
We
didn't
invent
the
shaw
hash,
I
mean
you
know
think
about
that
for
a
second
right,
so
I
I
just.
I
really
think
that
that's
a
slippery
slope
that
we
have
no
way
as
an
organization
of
picking
the
right
place
to
be
down
that
slippery
slope.
So
I
I
I
I
expect
individuals
to
decide
whether
they
contribute
and
work
in
the
area
or
not
based
on
their
own
decision
point
about
that.
But
I
don't
think
that
that
is
one
of
our
criteria
for
choosing
working
groups,
particularly
and
yeah.
We
should
agree
with
david
completely.
P
We
should
think
about
that.
Maybe
I'm
totally
wrong,
but
my
point
of
view
is:
that's
probably
not
that's
a
very
dangerous
criteria
for
us
to
go
down
because
tons
of
our
stuff
is
used
for
all
kinds
of
off
military
operations
and
everything
else
right
like
you
look
at
what
our
stuff
is
used
for,
you
get
a
much
more
dicey.
B
Than
you
think
I
agree
all
right
so
in
order
to
dive
down
the
scope
thing
a
little
bit
I
have,
I
do
have
a
couple
of
slides
on
the
draft
scope
that
was
floated
around
by
proponents
and
on
the
mailing
list,
and
so
I'm
going
to
let
you
read
those,
but
basically
you
know,
one
of
the
things
is
that
it
is
supposed
to.
You
know,
make
use
of
existing
ihf
technologies
that
you
know
we
are
familiar.
B
B
With
respect
to
what
is
the
scope
right,
so
api,
endpoint
definitions,
they
didn't
define
that
super
well
in
the
in
the
presentations,
but
it
is
in
the
documents,
in
terms
of
the
gateways,
have
an
api
to
interact
between
the
gateways
and
there's
a
gateway
api
inside
that
is
sort
of
at
the
outer
scope
side,
the
resource
identifiers.
B
You
know,
as
people
noted
that,
may
people
may
want
to
change
how
the
identifier
actually
looks.
That's
for
the
working
group
to
do
the
message,
flows
and
payloads
again.
Those
are
very
high
level
presentations
that
the
documentation
is
actually
quite
a
bit
longer
and
then,
of
course,
you
know,
there's
some
terminology
work
and
the
milestones
that
were
listed
were
you
know.
A
use
cases
document
sounds
like
six
months
because
they
already
have
done
a
lot
of
work
in
that
space
for
anybody
that
is
going
to
do
deployments
or
would
consider
it.
B
Your
particular
use
cases
spelled
out
too
so
that
we
can
make
that
stronger,
there's
an
architecture
document
and
a
protocol
document.
B
J
I
was
hoping
to
go
back
to
the
first
line
that
one
no,
the
ones
the
ones
that
said
verifiable
asset
transfer.
I
believe.
J
J
Two
three
hundred
three,
four
fantastic,
four
pages.
This
is
why
I
watch
okay.
I
don't
know
what
this
means,
and
this
is
actually
really
important.
I
I
just
think
I'm
nitpicking,
but
it's
really
important,
because
one
of
the
key
descriptions
this
goes
to
the
scope
of
the
question
is
one
of
the
key
to
not.
Descriptions
of
the
dlts
is
by
double
spending
protection,
and
so
is
double
spending
protection
in
scope
or
out
of
scope
for
this
problem.
L
L
J
Okay,
this
is
gonna
require
some
more
fleshing
out
here
because
okay-
well,
yes,
so
I
I
do
think
this
is
too
small,
but
I
think
I've
said
much
that
I
can
say
about
that.
B
One
of
the
interesting
things
to
me
is
that
you
know
we've
yeah
I'll.
Let
you
go
ahead
and
I'll
rephrase.
Q
L
B
K
B
B
You
know
from
these
two
of
the
most
important
lists
and-
and
you
know,
text
about
is
the
background
description
right
and
things
like
that
is
certainly
subject
to
debate.
I'll
note
that
you
know
the
one
oddity
for
me
in
this
whole
endeavor
is
that
we
do
have
a
case
of
a
a
potential
ietf
protocol
that
has
some
integrated
legal
frameworks
that
are
sort
of
necessary
for
the
protocol
to
work.
Well,
I
actually
don't
think
this
is
the
only
case
where
that
might
be,
but
I
was
struggling
to
find
the
other
ones.
I
think
the.
L
I
can
give
you
another
example,
so
x509
certs
issued
by
a
ca,
there's
a
there's,
a
long
csp.
That's
it's
a
legal
document.
It's
got
warranties
and
if
the
ca
screws
up,
you
sue
the
ca,
that's
illegal!
If
you
send
in
the
lawyers,
I
mean
that's
what
they
do
in
the
real
world.
B
K
S
B
Yeah,
I
think
the
proponents
are
very
open
to
help
all
right,
so
I
don't
see
any
more
questions
which
brings
us
to
given
our
understanding
of
the
problem
space
today
and
giving
our
understanding
of
the
proposed
scope
and
deliverables.
B
We
should
go
back
to
ask
sort
of
the
more
standard.
You
know
buff
questions.
B
B
We're
good
at
objections
all
right,
so
I'm
going
to
start
the
session.
Please
raise
your
hand
if
the
problem
space
is
well
understood.
Do
not
raise
your
hand
if
the
problem
space
is
not
if
you're
not
going
to
use
the
tool.
You're
welcome
to
raise
your
hand
in
the
room.
If
you
think
the
problem
space
is
well
understood,.
B
We're
still
seeing
a
few
more
votes
creep
in,
but
it's
looking
pretty
clear
at
this
point.
B
All
right
so,
for
the
note
taker
we
have
11
hands
raised
and
34
not
raised,
which
indicates
that
the
problem
space
35
not
raised
the
problem.
Space
is
not
super
well
understood,
which
does
not
mean
that
the
work
couldn't
happen,
but
it
does
make
it
more
challenging
if
we
don't
come
to.
B
On
what
the
problem
space
actually
is,
I
think
that
the
other
big
one
that
came
out
of
this
discussion
is
is
the
ietf,
the
right
standards
body
to
take
on
this
work.
B
H
A
B
Right,
I
think
we'll
call
that
good.
So
for
the
note
taker
that
was
24,
hands
raised
and
20
knot
rays
with
a
slight
leaning.
B
The
ietf
might
be
the
right
place
to
actually
conduct
this
work
25
to
19.
Last
minute
change.
Okay,
the
queue
has
been
cut.
So
the
next
question
I
think,
would
be.
B
B
B
B
M
B
So
that
to
me
sees
that
there
is
probably
a
scope
question
that
would
require
some
work
with
the
the
ads
in
order
to
come
up
with
a
scope
that
was
actually
better
or
possibly
another
buff,
roman
and
paul.
What
other
questions
would
you
like
to
have
answered.
K
So
it
seems
like
we
need
some
more
discussion
and
maybe
another
above
or
maybe
some
more
communication
between
the
parties
before
we
move
on.
Does
that
seem.
B
It
does
so
I
mean
the
one
thing
to
to
recognize.
Is
I
had
the
mailing
list
up
there
earlier
it's
standard
ietf
mailing
list,
mailman
sat.
If
you
want
to
get
involved
and
be
part
of
the
discussions.
I
suspect
that
the
authors
have
gotten
a
lot
of
good
feedback
today,
regardless
of
the
outcome
of
the
boss
and
whether
a
working
group
is
formed,
you
can
see
where
some
of
the
holes
in
terms
of
knowledge
that
that
people
need
in
order
to
find
it
a
tractable
problem.
B
All
right,
so
I
think
with
that
we
will
declare
this
buff
done
and
thank
you
to
the
presenters
for
bringing
your
problem
space
to
us
and
it's
an
interesting
one.
So.