►
From YouTube: IETF105-I2NSF-20190725-1000
Description
I2NSF meeting session at IETF105
2019/07/25 1000
https://datatracker.ietf.org/meeting/105/proceedings/
A
C
D
I'm
sure
you've
seen
this
thing
plenty
of
times
this
week.
So
if
you
haven't
read
it
yet
now's
the
time,
okay,
I
suppose
you
finished
reading
it.
So
here's
a
bunch
of
stuff.
We
have
our
charter
mailing
list.
We
need
a
minutes
taker
and
a
jabber
scribe
now,
I'm
also
on
the
in
the
jabber
room,
but
still
need
somebody
described
and
but
first
somebody
to
take
minutes.
D
D
D
Okay,
so
from
the
from
Prague
to
now
a
few
things
that
happen.
If
you
haven't
been
following
closely
on
the
mailing
list,
we
had
a
IPR
scare,
where
the
University
of
Farrell
authors
published
an
IPR
statement
that
he
had
patents
on
the
yank
models
in
the
documents
and
we're
willing
to
provide
them
provide
licenses
on
a
ran
basis,
but
perhaps
for
a
fee
that
kind
of
thing
kind
of
scares,
people
in
the
security
area
and
so
we're
having
a
bit
of
a
scare.
D
So
anybody
wants
to
weigh
in
on
this
now.
Okay,
by
the
way,
the
room
mic
is
right
there
by
the
door,
it's
kind
of
a
strange
place
to
put
the
only
room
mic,
but
somebody
wants
to
say
everything:
you're
doing
here
is
stupid
and
then
stomp
out,
that's
that's
what
the
mic
is
for.
Well,
you
can
do
some
other
things
so
milestones.
D
A
Ietf
well
for
this
applicability
drop
is
finally
finished,
all
the
review
and
these
mean
to
move
into
iesg,
and
we
have
requested
young
doctor
review
for
all
our
yeah
modules.
Those
are
the
important
deliverables
by
our
working
group.
We
get
lots
of
comments
and
the
authors
are
working
to
address
them.
There's
a
capability
when
we
have
completed
working
group
last
call
yes,
I.
E
A
You
thank
you
so
much.
Thank
you
so
much
for
moving
this
forward.
It
has
been
a
long
process.
The
capability
one
is
also
moving
forward.
We
have
requested
a
yes
G
to
move
forward
good.
Thank
you.
Some
creative
issues
that
young
modules
we
have
lots
of
comments
from
the
young
doctors.
We
want
to
resolve
all
those
issues
before
we
move
forward
for
the
working
group
last
call
and
there's
I
think
that's
the
main
progress
for
those
and
for
the
ipsec
flow
protection.
Maybe
you
wanna
say
a
few
words.
A
We
actually
had
a
joint
meeting
with
IDR
working
group
in
the
routing,
because
there's
several
proposals
of
IPSec
control
simply
because
more
and
more
sd1
and
that
were
coming
to
place
and
eating
the
controller
based
IPSec
control.
So
we
had
a
joint
meeting
that
was
yesterday
right
and
FEMA
said.
Okay,
and
so
maybe
you
can
give
a
few
more
well.
D
D
Secondly,
he
comes
up
with
a
new
algorithm
or
absurd
a
5-second
me
I
mean
any
anybody
from
Russia
might
come
up
with
the
document
saying
how
to
use
when
one
of
their
ghost
protocols
in
IP,
sir
and
then
it
t,
really
gets
publishing
they
get
in
Vienna
number
that
without
any
I.
To
second
me,
involvement
and
we
don't
want
to
have
to
leave
to
revision
our
document
every
time.
Something
like
that
happens.
A
Common
to
it,
so
this
Sdn
IPSec
flow
protection
draft,
only
copper,
the
initial
configuration
for
the
IPSec,
the
policies
and
there's
a
two
to
two
cases
right
and
there's
another
scenario
which
is
for
the
rekey
mechanism.
That
is
also
that
is
covered
by
a
different
draft,
which
was
discussed
in
The
Dispatch
right,
the
the
control
Ikey
controller,
a
key,
that's
a
different
aspect:
right:
okay,.
D
D
Our
document
has
two
modes
of
operation
or
cases
as
they
call
them
the
Ike
and
I
class,
and
the
controller
Ike
is
trying
to
do
something
that's
kind
of
in
the
middle.
Yes,
so
we
could
are
the
documents
that
we
have
do
rekeying
just
fine
in
the
case,
then
the
I
complementation,
the
Ricci
and
in
the
IKE
list,
and
the
controller
has
to
periodically
replace
the
keys,
and
the
document
supports
that.
So
we
don't
really
need
controller
Ike
for
that
controller.
D
A
If
not,
we
will
go
move
forward.
Here's
basic
summary
of
all
the
data
models
to
be
created
generated
by
this
working
group,
the
capability,
one,
that's
the
base
data
model
and
that's
like
a
base
information
model
which
is
moving
into
is
G
requesting,
and
then
we
have
the
data
model
itself,
the
data
model.
We
have
one
base
model
and
then
we
have
detailed
at
the
interface.
We
have
individual
ones
to
be
more
detailed,
build
on
top
of
the
base
model,
and
that's
it
from
here.
F
This
is
from
SKT
you
so
basically,
last
week
and
my
SKT,
you,
a
student
and
a
tiara
in
Korea
researchers,
participate
in
I
to
NSF
hackathon.
This
is
a
nice
hackathon,
so,
amazingly
our
three
years,
nine
consecutively.
We
did
a
custom.
So
this
time
we
try
to
to
demonstrate
I
to
NSF
framework
and
interfaces
can
be
used
by
the
public
cloud
platform
and
with
commercial
power
system,
as
well
as
open
source
ricotta
for
our
repeater.
F
So
this
is
the
build
system.
This
hackathon,
we
implemented
those
our
draft.
Okay,
except
the
security
policy
translation
document
you
can
see
a
full
draft
are
working
rotor
document.
So
at
the
beginning
of
this
working
group
session,
chair,
Linda
and
Yarbrough
mentioned
those
are,
there
are
model
draft
are
already
laid,
but
it
might
escape
you
a
team
and
modified.
F
There
are
my
mother
dropped
the
plot
team,
young
doctors,
reviews
acceptor
and
as
a
facing
interface
there
a
model
dropped.
Two
things
are
left,
so
I
will
explain
layer
the
next
yeah
there
are
model
slides,
ok,
so
this
is
the
poster.
This
is
my
team
and
including
it
ETA,
researchers,
ok
and
also
my
student
Chiho.
F
So
this
is
the
the
I
to
NSF
system
having
an
asset
framework.
Ok,
so
this
time
we
have
commercial
power,
it
is
called
a
wins
with
providing
fiber
service
in
South,
Korea
and
also
those
platform
is
based
on
EGR
I
seek
as
Oh
a
cloud
system
is
standard
for
on
security
on
earth.
Social
Security
is
everywhere,
yeah
like
air.
Ok,
that's
the
name
is
a
secret
it
on
air.
So
basically
you
can
see
in
the
middle.
Are
we
have
I
to
inaudible
security
controller
in
the
left
hand,
side
I
to
Alaska
user?
F
F
You
can
see
graphical
interface,
specify
the
talk
security
function
facility,
app
ability,
so
we
specify
and
then
using
rest
large
duration
interface.
It
can
be
used
to
register
an
SF
into
a
security
controller.
Also
consumer
face
interface
is
specified
operator.
Highlander
security
policy
is
specified,
for
example,
during
our
time,
employee
cannot
access
as
NS
services
such
as
YouTube
and
Facebook,
etc,
and
then
that
our
policy
will
be
translated
by
security
controller
is
fastly
a
security
policy
translator
and
then
that
policy
is
delivered
the
over
an
SF
a
facing
interface.
F
So
this
figure
shows
this
is
an
entry,
a
router,
and
then
you
can
see.
This
is
a
ricotta.
This
is
a
fiber,
so
they
are
linked
each
other
using
a
service
function,
chaining,
okay,
and
that
case
one
we
are
tested
the
two
cases,
the
first
one
is,
you
can
see
as
soon
as
it
Web
Filter.
During
what
time
employee
cannot
access
s
an
SNS.
The
social
network
service
can
access
this.
Second
one
is
also
just
a
block
or
access
to
website
using
this
firewall
system.
F
F
F
So
this
is
a
video
script,
so
you
can
visit
and
then
you
can
see
the
nice
video
clip
to
show
whole
process
a
registered
NSF
and
also
specify
user
on
high-level
policy.
Also,
we
can
show
the
process
of
translation
and
also
we
can
demonstrate
the
two
cases-
the
Web
Filter
and
the
fiber.
So
this
is
open
source
project,
so
the
benders
and
operator
also
other
universities.
It
can
take
advantage
of
our
work
project
because
we
share
all
sorts
crota.
Ok,
that's
it!
Thank
you.
Yeah
thank.
A
You
very
much
that's
a
really
good
was
hi
Kazan
demonstration.
Actually
in
I
talked
to
Paul
yesterday
we
probably
will
introduce
this
heck
sound
to
the
Owen
ug.
They
have
us
security,
Software,
Defined
security
services.
I
think
this
will
be
a
good
reference.
Thank
you.
Ok,.
D
F
D
You
have
mirror
from
the
floor.
No
hats,
so
I
think
it's
very
encouraging
that
you're
working
with
the
commercial
firewall
and
I
you
think
that
usually
be
able
to
work
with
other
commercial,
fair,
especially
the
more
popular
ones
like
from
Cisco
and
Palo,
Alto
and
former
employer.
That
I
shall
not
mentioned,
or
is
this
ok.
F
Company-
yes,
oh,
maybe
not
international
right
so,
but
it
is
a
famous
in
Korea
yeah,
so
I
think
the
Cisco
or
some
juniper
orderly
offenders
case
in
order
to
integrate
their
fiber.
They
needed
to
implement
an
SFO
facing
interface,
also
restoration
interface,
so
if
they
are
willing
to
implement
that,
we
can
definitely
test
yeah
together,
yeah,
okay,
thank
you.
Okay,
the
echo.
G
Developers
is
about
because
I
this
morning,
I
was
discussing
with
the
chair
of
the
energy
about
the
intent
base
and
relax.
One
of
the
things
that
we
are
looking
for
is
precisely
way
of
somehow
formalizing
I
mean
intent.
Language
is
no
intent,
because
in
formalizing,
internists
is
an
endless
task,
but
the
VA
was
precisely
to
try
to
find
some
common
patterns
and
come
on
semantics
in
the
do
you
have
the
formal
definition
of
the
language
you
use,
the
the
ideology
I
mean
for
the
definition
of
the
policies
alike,
so.
F
That
is
a
question,
so
you
can
see
the
let's
see
consumer
patient
interface,
so
you
can
see
consumer
facing
interface.
The
basic
important
thing
is
the
EC,
a
policy
you
right
so
according
to
that
issue,
a
policy
rule
and
our
consumer
facing
interfaces
they're
a
young
model,
so
this
interface
is
just
one
tool
map,
so
it
was
actually
according
to
that
Tara
young
model,
so
in
order
to
provide
the
intent
to
base
the
security
service.
Currently,
my
illiteracy
lab
also
has
interest
in
it.
F
G
H
H
H
Okay,
let's
go
next.
This
is
developer
to
the
University
of
bottom
and
I'm,
going
to
present
the
last
updates
of
the
SDM
protection,
a
basic
prohibition,
yeah
well,
first,
just
a
quick
summary
of
the
last
changes
now
from
version
3
to
version
4.
Now
we
have
two
names
for
cases
we
have
the
ID
case
and
the
eyeglass
case,
and
perhaps
the
main
changes
that
now
we
have
divided
the
model
in
three
parts.
H
H
H
B
H
H
H
H
H
B
H
I
H
D
H
H
Well,
we
are
waiting
to
have
more
stable
draft
version,
because
the
changes
from
the
last
version
to
this
one
are
are
very
intensive
changes.
So
already
I
list
you
to
adapt
the
source
code.
The
link
to
the
Egypt
repository
was
presented
in
the
last
meter,
but
if
you
need
them,
I
can
send
an
it
again
to
the
Middle
East.
The
problem.
G
G
Well,
an
IPSec
distribution
or
set
up
schema
for
anything.
So
what
does
that
make
a
deployment
using
our
the
mobile
say?
We
have
a
using
own
or
network
street
or
SM
to
deploy
a
vnfs
that
that
communicate
can
use
IPSec
as
defining
the
model.
So
what
we
try
to
measure
is
a
demonstrating
that
it
is
applicable
in
in
the
case
of
any
V
deployment
and
based
on
control
service
models.
F
Hello,
everyone
again,
this
is
a
pearl
from
skq,
so
basically
the
five
there
are
young
mothers
are
updated,
also
I
submitted
during
this
alarm.
Okay
question
I
submitted
a
libyan
letter
with
the
device,
the
attractor
linked
to
each
of
a
young
doctors
who
gave
us
the
lip
you
okay,
so
basically
our
data
models
are
come
to
the
information
model
tractable
and
as
a
cop
abilities.
F
So
the
first
three,
the
base
young
model
is
the
sous-sous
capability
data
model.
Based
on
that,
we
providing
full
interfaces,
especially
consumer
pays
interface
and
another
facing
laceration
interfaces.
Also
NSF
were
monitoring,
therapist
is
Lionel,
the
you
can
see
the
first
cop
ability,
data
model
and
second
third
force
consumer
NSF
facing
laceration
interfaces.
Error
models
are
devised
according
to
a
young
doctors
are
comment,
so
we
had
a
nine
I
kept
a
hackathon
project
last
three
years
and
we
already
deport
times
during
a
zone,
so
my
students
are
encouraged
with
these
kinds.
Overs.
Okay,
thank
you.
F
So
we
updated
according
to
a
young
doctor.
So
we
are
the
most
over
here
comment
our
address,
except
and
as
a
facing
interface.
There
are
motor
case
to
comment.
The
first
one
is
so:
the
access
control
list,
young
or
module,
has
the
overlap
test,
some
definition
or
with
our
Ellison
facing
so
we
needed
to
import
the
use,
the
SEL
definition
so,
but
it
takes
time
so
we
reflected
that
one
last
probably
next
our
Wednesday,
we
the
plug
the
tap
replacement
or
second
one
also.
F
F
Think
is
that
the
power
that
we
can
leap
like
one
week
enough
and
then
we
can
divide
estimate
and
then
we
can
ask
for
a
what
your
escort,
but
other
consumer
facing
and
registration
interface
are
ready
for
what
were
last
call
I
believe
so
I
agree
much
I
explained
so
the
capability
theorem
our
model
case.
The
major
comment
is
about
the
security
considerations.
We
tried
to
use
recommend
in
the
temple
and
also
we
added
the
two
things
so
possible
attack
related
to
mobility
data
module.
So
we
are
added
to
our
sentences.
Okay,.
F
Next
data
and
time
case
already
are
built
in
time,
so
we
just
use
the
disability
and
terror
and
time
letter
that
we
defined
again
and
also
or
we
had
some
misleading
to
finish
on
time
zone.
So
our
intention
is
the
time
interval.
Some
interval
start
time
starting
time
and
end
time,
so
we'd
apply
step
Tom,
so
we'll
place
time
zone
with
the
time
intervals,
so
a
CL
number
is
not
used,
so
we
delete
and
descriptions
are
improved
according
to
young
doctor
is
a
comment
again.
F
The
security
consideration
we
added
okay,
I
previous
on
the
I
so
to
write,
and
then
another
facing
this
second
one:
okay
and
that's
a
facing
case,
other
comment
or
yet
addressed
the.
But
you
can
see
the
ACL,
especially
some
definition.
This
is
a
so
ACL,
so
pack
imagine
case
we
use
the
ACL
definition
and
secondly,
I
mentioned
the
co-operative
data
model
and
as
a
facing
the
other
model
are
there
have
they
have
a
common
definition?
We
just
defined
cavalli
data
model
and
MS
effacing
just
will
be
used
in
pi
in
protein.
F
Okay,
so
those
are
I
think
we
can
do
next
several
days
and
then
we
can
be
eyes
and
SME
the
consumer
patient
case
really
devised
according
to
young,
but
the
major
thing
is
accessor
control,
so
policy
management
was
authentication
method
we
can
using
list.
So
we
can
accommodate
a
possible
awesome
case
of
Macerata.
F
So
the
time
is
it
yeah
now
list
the
letter,
then
I
think
a
container
and
then
laceration
interface,
so
research,
an
interface.
The
case
really
buys
a
young
doctor,
so
we
revise
the
young
module
structure
according
to
the
guideline.
So
this
is
young,
the
developers
title
line,
so
we
leave
iced
and
also
we
using
it
more.
You
can
see
a
natural
isolation,
prefix
yeah,
previously
a
leader
job
come
confusing.
Some
prefixes
are
we
using
this
griffix
minor
issue
and
and
then
previously
we
didn't
ask
for
the
ipv6
address.
F
Now
we
support
IP
version
6
address
as
well
as
I
keep
Adam
for
us,
and
then
we
revise
the
example.
Ok,
IP
version
6
addresses,
allowed,
ok
and
also
a
nest
reporter
at
the
recipe
and
change
the
NSF
report,
because
it
is
misleading
because
we
have
already
addressed
so.
We
change
the
NSA
book
port
that
toxic,
we
function
port
and
we
revised
security
considerations
according
to
Caroline
and
I
said
for
monitoring
at
era.
Model
case
is
not
lady
for
working
or
last
call,
so
we
are
planning
to
implement
and
show
pro
concept
next
Singapore
meeting.
F
So
my
student
and
other
university
social
University
I
believe
that
two
universities
are
working
for
the
NSF
monitoring
terror
model.
So
next
step
is
capability.
There
are
model
of
this
money,
our
chair
of
already
the
cast
the
iced
tea
evaluation
so
and
then
other
document,
the
case
consumer
facing
reservations,
page
I,
believe
our
lady
for
walking,
guarulhos
coal
and
and
as
a
facing
interface
case.
We
can
wait
until
two,
a
major
issue,
just
some
detector
ring
and
the
the
use
of
a
structure.
F
So
after
that,
I
will
let
terrorists
to
working
or,
let's
call
a
monitoring
case.
I
mentioned
that
we
preparing
for
testing
and
for
next
the
Singapore
IT
have
a
meeting
okay.
This
is
appendices,
I
believe
you
guys
can
take
a
look,
so
we
can
show
command
old
and
new
tech.
Something
like
that.
So
you
can
take
a
look
if
you
question
Frisian
em,
you
know
I.
F
A
F
G
Developers
again
is
about
the
registration
interface,
so
the
document
I
sent
you
an
email.
Yesterday,
probably
you
have
not
seen
it
yet.
I
have
today.
I
have
checked
that
has
700,
ok,
emails
waiting
for
me
and
I.
Imagine
that
they
save
the
case
of
most
of
us
in
the
case
of
the
registration
interface
Annie's,
discussing
in
Pratt,
about
pointing
from
the
when
the
registration
of
a
capabilities
made
a
pointer
to
the
the
Jan
Mobile
of
the
gen
model,
fragment
that
can
be
used
precisely
to
control
that
capability
in
another
model,
so
is
a
link.
G
Well,
did
you
remember?
We
were
talking
about
the
idea
abuse
in
this
jang
jae
myung
mount
scheme
amount
to
make
that
reference.
I
would
like
to
see
if
this
can
be
used,
because
at
the
end
would
support
the
translation
process
like
the
one.
You
are
working
yeah
that
if
you
want
to
touch
this
capability,
go
do
to
this
more
specific
one,
that
it
would
be
the
let's
say,
the
part
of
the
BNSF
facing
interface,
damn
useful
for
managing
the
capability.
Simply
okay,.
F
G
Just
remember
precisely
one
dimension
is
a
bit
second
row.
It
would
be
probably
interesting
to
show
that
I
mean
the
most
detail.
A
schema,
Jiang
schema
we
have
right
now
is
precisely
from
the
FOV
IPSec
well
Gabi
was
presenting,
probably
would
be
nice
to
have
in
the
registration
interface
to
that.
To
add
an
example
to
say
well
for
these
capability,
we
can
point
to
these
other
scheme
and
we
were
showing
how
we
are
relating
both
documents.
G
B
G
G
F
Okay,
so
I
think
the
security
policy
translation
in
I
to
NSF
is
important
part
of
our
security
controller,
because
the
high
level
e
Young
and
the
low
level
young
are
specified
in
our
working
group.
The
problem
is
how
to
link
to
each
other
also
how
to
translate
so
this
document
we
specify
the
link
is
between
two
young
data
model
and
also
be
providing
implementation
high,
trying
for
architectural
translation
and
also
process
of
translation.
So
the
motivation
of
our
draft
years
you
can
see
in
the
middle,
you
can
see
plot.
F
My
son's
computers
were
a
smartphone
from
a
malicious
website,
so
this
is
a
kind
of
intent.
Diego
mentioned
the
right,
so
this
is
a
high
level
and
intent.
So
our
hackathon
case,
we
provide
a
graphical
interface
to
put
some
this
kind
of
my
son's
a
computer
and
malicious
website.
So
this
my
son's
and
malicious
websites
are
called
the
endpoint
information.
F
Okay,
so
somehow
we
need
to
so
right
now
we
are
using
proprietary,
some
method
to
populate
this
information
into
security,
our
controllers
NSF
database,
okay,
but
secondly,
given
this
policy,
our
core
is
we
in
horse
into
the
truck
security
function?
Such
as
a
fire
and
repeater
so
drop
packets.
From
my
son's,
this
IP
addresses
used
by
my
son's
computer,
so
the
malicious
websites
on
Sunni
militia
websites
are
hung,
calm
and
illegal
calm.
Maybe
we
can
have
more
if
we
specify
using
and
the
find
information
in
the
PC
interface.
F
So
the
goal
is
given
this
high
level
policy.
We
translate
into
this
low
level
policy
right
most
of
the
file
filter
case
this
IP
address
and
this
URL
are
specified
rather
than
my
son's
computers
and
malicious
websites
right.
So
the
core
is
how
to
translate
so
so
we
providing
the
some
compiler
Sharia
translation.
F
So
we
these
are
protocol,
which
we
provide
easier
security
policy
construction.
We
don't
need
care
or
selection
or
Apple
appropriate
NSF.
What
kind
of
NSF
are
are
required
to
provided
this
holiday
policy
and
the
secondly
efficient
maintenance?
Okay,
in
the
past
that
we
can
use
it,
XSLT
is
a
user
to
map
her
labor
yang,
examiner
AR.
This
is
a
kinda,
hard
coded,
but
our
approach
case
we
provide
more
flexible
or
link,
however
young
and
reliever
young,
so
previously
some
our
ad
alone
yeah
long
right.
F
You
mentioned
this
draft
of
the
case
apparently
looks
like
a
soft
engineers
of
the
judgment
like
that
data.
Why
we
try
to
we
adding
some
new
section,
explain
the
mapping
list
orbit
era
and
motor
peers
to
share
that
consumer-facing
interface
and
an
SFA's
interface
okay.
So
this
is
a
disarm
a
thing.
So
this
is
an
NSF
populate
the
information
entity.
This
is
enter
point
such
as
a
my
son's
computer
or
some
malicious
website.
Okay,
oh
yeah,.
F
So
because
without
the
mapping,
the
designer
overstocked
engineer
the
figuring
out
so
our
sometimes
the
mapping
is
not
one-to-one,
sometimes
I
want
too
many
something
like
that.
So
in
the
case
we
specify
the
consumer-facing
and
NSF
Amy
I'm,
not
I'm.
Aware
of
that,
but
developer
can
lead
our
or
tracked
case.
It
will
be
the
harder
time
to
figure
out
how
can
map
or
how
to
link
to
each
other.
So
that's
why
we
come
up
with
this
draft.
F
E
I
I
think
I
understand
the
gap
as
in
what
the
software
engineer
doesn't
understand
and
designer
the
way
you
can
explain
it,
but
why?
But
that
would
that
all
seems
like
that
would
happen
in
the
same
organization
building
the
controller
right.
So
why
do
we
need
to
standardize
this?
Typically,
when
we
do
that,
because
there's
lots
of
different
parties
or
vendors,
who
we
want
you
into
operating
with?
Not
that
we
want
to
standardize
the
different
kind
of
roles
inside
the
organization,
because.
F
You
can
see
this
er
entity
relation
therapist
case,
so
this
information
should
to
be
populated
by
some
operator
side
and
also
this
NSF
capability
case
provided
the
vendors
right.
So
they
need
the
tools.
This
data
structure
should
be
shared
by
each
other
and
what
kind
of
fields
are
should
be
are
populated
without
that
it
is
a
difficult
to
translate.
So
somehow
we
need
to
define
the
relation
between
capability
base
and
some
endpoint
information,
so
this
is
the
first.
The
second
one
is
this
mapping
you
can
see.
F
The
first
line
is
a
consumer
facing,
for
example,
if
I
were
a
red
filter
case,
my
son's
source,
eventually
I
feel
there
should
be
treasure
light
to
write
this
okay.
So
now
this
is
a
consumer
facing
entity.
This
is
another
face
entity,
so
these
guys
are
mapping.
Information
is
blissfully
specified
that
is
documented,
so
developer
can
easily
understand
how
to
implement
their
yet
translator.
Okay,
so
that
yeah.
B
So
Roman
I,
all
to
try
to
I
was
encouraging
him
to
try
to
put
this
actually
out
as
a
standard.
You
remember
that
earlier
this
week,
I
mentioned
we're
trying
to
import
some
of
the
work
from
I
to
us
or
some
of
the
concepts
into
this
secure.
Vpn
I
was
hoping
an
example
like
this
might
give
them
an
idea.
Now,
I,
don't
know
if
this
goes
on
in
the
security
realm
to
be
a
published
standard
or
for
just
getting
information
to
help
us
bootstrap.
B
My
my
concern
is
to
see
some
examples
so
that,
as
we
get
into
a
bgp,
secure
VPNs
that
I
have
some
examples
to
point
to.
So
I'm
pleased
to
see
this
model
and
this
work,
because
it
gives
an
example
of
how
things
might
work
for
translation
and
that's
a
difficult
part.
Okay.
So
I'm
I'm
I'm
happy
to
see
this
for
a
different
reason
than
the
security
vendors.
I
you
know
you're
straight
security
vendors,
I'm
happy
to
see
it
because
I'm
working
with
routing
security
VPNs-
and
this
may
give
some
clue
whether
it
is
or
not.
B
A
G
J
Hello,
friends,
AHA,
so
sorry,
I
haven't
a
real,
very
clear.
It
heard
about
this
talk,
but
my
current
feeling
is
that
if,
of
course,
this
document
we're
happy
with
where
for
this
work
but
I
think
only
when
the
document
does
not
only
cover
the
mapping
between
the
you
know,
within
the
customer-facing
interface
motor
to
the
nacelle
facing
interface,
latemodel
I
think
the
more
important
the
part
of
the
better
way
to
write
this
document
should
be.
J
You
should
write
something
about
your
your
internal
architecture,
design
and
you
know
the
process
how
to
do
the
magic
that
the
general
way
to
do
the
mapping,
but
not
the
man.
You
deserve
okay,
so
that
that
that
that
we
are
wearable
contribution
to
the
industry,
because
I
can
see
that
not
only
you
nitro
and
safe
in
a
lot
of
service
provision
motor.
We
have
two
two
levels
of
service
murderer
or
you
know
that
data
model.
We
need
this
kind
of
general
mapping.
J
We
can
either
in
any
controller
right
so
I'm
more
hoping
to
see
these
kind
of
things,
but
not
only
the
mapping.
You
know
and
another
another
point
I
want
to
see
in
this
document
is,
if
you,
of
course,
you
can
give
some
example
from
the
item
assess
perspective
for
how
to
translate
the
security
service
model
to
the
security
configuration
motor,
but
which
gives
some
more
advanced
examples
because
you
know
for
Cepeda.
J
You
just
keep
a
very,
very
simple
example,
but
actually
we
want
to
see
that,
for
example,
if
you
from
the
end-to-end
the
way
to
see
a
security
service,
maybe
in
in
the
in
the
in
the
past,
there
are
several
security
functions
and
how
we
turn
them
together
and
how
we
map
this
kind
of
very,
very
huge
remoter
to
the
you
know,
to
the
to
the
several
part
of
models
on
the
on
the
sale
facing
motor.
So
maybe
I
think
that
know
that
example
will
be
be
more
youthful
bit
more
representative.
F
Currently
we
explain
a
simple
web
filter,
or
so
we
can
combine
fire
and
repeater
logging
together,
so
yeah.
Definitely
the
solid
function,
cheney
in
case
we
needed
two
more
improve
the
example,
but
you
mentioned
so
you
can
see
this
is
the
architecture
of
security
policy
translator.
So
we
improve
this
example,
but
given
this
kind
of
architecture
process,
vendors
can
easily
implement
this
morning.
Somehow
engineer
I
forgot
his
name,
but
he
working
for
an
AC
pn
that
they
are
doing
right,
given
some
requirement
that
they
try
to
compete,
rollable
and
transport
network
right.
F
He
has
also
lots
of
interest
in
this
rock,
but
in
my
opinion
we
tried
to
focus
on
this
concrete
example,
I
told
as
a
first
and
then
we
can
generalize
if
we
broaden
at
this
moment
I
think
maybe
it
takes
two
or
three
years
who
knows
because
this
is
not
easy
job,
but
we
tried
to
publish
this
information
or
alpha
C
and
then
TC
will
be
starting
a
point
for
other
area.
That's
my
key.
J
When
you
focus
on
the
alternative,
a
specific
mode
or
something
but
I,
think
you
also,
you
still
need
to
give
us
a
general
general
way
or
general
system
how
it'll
work
to
do
this
mapping,
so
actually,
yes,
so
that
the
map
on
there
that
maybe,
is
not
important,
but
the
way
how
to
do
this
mapping
seem
hot
that
is
youthful
and
that
can
be
the
Ojibwa
for
future.
Your
skates
do.
F
Ya
clarify
how
to
map
means
some
given
the
higher
level
yeah
Laura
boyoung,
so
how
to
map
some
some
leaf
node
to
some
other
leaf
node
link
automatically.
So
what
do
you
mean
by
your
yeah
matching?
So
I
can
understand
that
your?
What
do
you
mean
by
so
currently
we
we
are.
This
is
not
a
machine
learning,
risers
operator
or
populate
the
mapping
information
manually
into
a
database-
okay,
so
maybe
in
the
future.
So
given
her
live,
a
young
low-level
young
machine
learning
case
can
figure
out,
but
I
think
it
takes
several
years.
F
B
You
so
much
Paul
I'm
gonna,
ask
Franco
a
clarifying
question:
that's
okay,
okay,
so
Frank!
My
understanding
is
that
you
treasure
the
the
ability
to
have
the
architecture
and
the
methodology
of
doing
the
translation
between
the
the
high
level
yang
data
extractor
and
the
policy
detail.
But
what
you
really
are
looking
is
for
the
draft
to
specify
the
work
is
simply
an
example.
Did
I
understand,
you're
coming
I.
I
J
You're
right,
you
understand
my
point
I.
My
point
is
that
the
the
example
is
not
our
goal
right.
The
example
is
just
to
illustrate
or
clarify
how
we
do
this
kind
of
formatting
and
I
hope
that
this
week
and
we
have
pickable
to
other
use
cases
for
other.
You
know
policy
making,
so
so
that
so
that
can
that
way,
are
we
increase?
J
B
Again
from
the
use
case
that
I
have
with
another
set
of
work,
I'm
doing
and
routing
this
general
use
case
is
helpful
for
understanding,
high-level
capability
policy,
high
level
policy
that
we've
developed
for
security,
especially
in
the
some
of
the
IPSec
stuff,
and
how
to
translate
it.
So
I
appreciated
this
document
where
the
working
group
takes
it,
what
the
ad
does
with
it
it.
The
fact
that
it's
currently
existing
is
helpful.
B
E
Okay,
go
ahead.
Yeah
live
room
engineer,
pulling
the
thread
that
I
think
Frank
keyed
up
around
the
architecture.
I
just
wanted
to
ask
a
clarifying
question:
the
documents
I've
seen
to
date
for
my
to
NSF
and
as
I
kind
of
read
the
Charter.
It
was
primarily
first
enumerate
the
architecture
which
was
drop,
the
boxes.
You
know
conceptually
on
a
piece
of
paper
and
then
a
lot
of
the
work
was
then
describing.
E
What
is
the
role,
those
boxes
and
then
what
goes
in
and
out
of
those
boxes
so
describe
kind
of
the
information
flow,
the
api's
and
then
describe
the
data
models.
Is
that
accurate?
Yes,
okay?
So
as
I
read
this,
this
strikes
me
as
something
different
because
for
the
first
time
we're
going
inside
one
of
those
boxes.
So
it's
now
not
the
API
in
and
out
of
the
box.
It's
about
what's
happening
inside
the
box.
F
Yeah,
okay,
so
the
the
thing
is
alone,
I
think
left
the
three
years,
a
hackathon
project,
so
I
think
this
translation
is
I
think
at
a
very
usable
part,
I
want
to
share
to
ITF
our
community,
so
this
can
be
expanded
for
the
general
for
some
Network
comparation.
There
is
a
disparity,
so
I
believe
we
need
to
decide.
We
can
adopt
this
or
work
for
AI
12s,
the
security
first
and
they
expanded
to
other
area.
So
that
is
a
mind
yeah.
So
that's
chant,
yeah
yeah.
E
F
C
I
A
F
Okay,
this
is
a
poor
from
s
Kiki,
so
the
line
now
our
long
mention
under
the
chatter.
Maybe
we
can
take
a
look
chatter.
So
basically,
as
I
understand,
our
I
to
NSF
is
focus
on
framework
and
the
inter
pieces,
and
those
are
our
primary
our
course
may
be.
After
our
the
main
primary
tasks
such
as
we
are
polishing
or
in
two
pieces,
there
are
models,
and
then
maybe
we
can.
F
The
chatter
consider
some
year
of
deployment
over
I
to
NSF
I
think
at
the
security
policy
translated
up
one
thing:
the
other
thing
is
during
our
Harrison
project,
so
the
interaction
with
other
systems
such
as
the
US
embassy,
sub
function,
chaining
also
STM
controller,
the
kind
of
interface
that
are
not
specified
so
to
realize
I
to
MSF,
maybe
a
we
can
add
new
item
there.
Our
model
for
those
SF
PCI
to
an
SF
interaction,
I,
told
us
ASTM
interaction,
yeah.
That
is
a
my
suggestion.
Yeah
I.
D
Have
never
just
not
to
mention
that
I've
a
couple
of
days
ago
has
submitted
a
new
individual
graft.
That's
a
profile
of
the
Sdn
IPSec
work
for
a
specific
use
case
within
the
data
center,
and
this
is
not
a
working
with
draft
and
not
in
se
for
adoption
right
now.
I
might
do
that
after
the
Sdn
IPSec
draft
goes
through
the
is
Jian
well,
when
it's
real
really
done,
I
might
ask
for
a
doctrine.
A
F
Our
Linda
has
a
question,
so
we
have
another
document
about
amicably,
electrons
architecture
based
on
I
to
NSF.
Probably
you
can
click,
and
so
that
one
is
for
how
to
develop
I
to
NSF
on
top
of
cloud.
The
system
such
as
an
appellee
Laplace
architecture,
so
I
think
at
the
bottom
yeah
young,
a
second,
yet
one
so
total
are
explaining
how
to
implement
I
to
NSF
on
top
of
the
ETS
I
and
I
probably
reference
your
architecture,
so
we
explain
how
to
build,
build.
F
I
turn
I
have
a
component
how
to
link
a
torus,
a
framework
into
an
appropriate
lab
France
model.
So
if
the
working
group
think
this
is
useful,
especially
implementation
and
deployment,
I
believe
it
is
valuable,
so
we
targeted
information
or
artificial
working
group.
If
Okamura
power
have
interested
in
it,
that
is
a
year
possible.
Yeah,
yeah
item
yeah
right
now
are
so
we
need.
We
didn't
discover,
yeah,
okay,.