►
From YouTube: IETF92-DMM-20150324-1730
Description
DMM meeting session at IETF92
2015/03/24 1730
A
Alright
welcome
everybody
IDF
92
in
Dallas,
and
this
is
the
TMM
working
group
meeting
first
couple
of
things
of
vb
meet
this
week
price
so
today
and
then
thursday.
Again
today,
we
only
have
this
quick
slot
for
one
hour
with
two
presentations
and
the
10
will
continue
with
the
rest
of
the
stuff.
One
on
Thursday.
A
I
already
put
the
blue
sheets
going
around,
so
put
your
names
there
and
the
so
for
today
we
have
basically
two
things
about
the
kind
of
reports
from
the
working
teams
and
working
dream
team
leaders
describing
to
status
what
they
have
achieved
in
last
I'd
f91,
so
exposing
mobility
state
is
working
to
him
one
and
then
the
enhanced
anchoring-
and
this
is
basically
all
that
we
have
today
and
then
on
Thursday.
We
will
continue
with
work
in
teams
three
and
four
after
that
we
have
the
ones
of
TM
of
the.
A
Ok,
all
right,
ok,
trying,
okay,
good,
alright!
So
let's
get
started
so
the
just
a
bit
of
background
so
that
the
those
who
haven't
been
following
the
so
we
have
our
base.
Our
offices
done
now
all
are
out
and
the
multi
working
teams
are
kind
of
working
towards
their
solutions
and
most
likely
these
working
groups
are
also
at
some
point
of
time,
going
to
propose
some
some
individual
drafts
that
then
we
handle
like
any
other
individual
ID
and
out
of
those
like
a
relative
working
group
designs.
A
C
D
D
The
presentations
are
already
sent
out
in
the
in
the
bit
in
the
meanest,
with
the
files
of
representation
in
Costa,
the
movie
network,
again
a
function,
a
movie
night
the
address
and
then
there's
one
more
presentation
to
be
arranged.
Data
and
the
next
slide
are
after
seeing
this
and
after
reading,
all
these
are,
and
in
addition
to
all
these,
they
were
also
a
number
of
jobs
proposal
ever
presented
prior
to
the
formation
of
his
team,
and
so
we
should
also
take
them
in
on
day,
and
they
were
quite.
D
A
number
of
different
proposals
are,
and
so
it
when
look
at
them.
The
artists
is
the
charter
of
the
working
team
contest
each
other
and
from
the
Charter.
Yet
it
is
supposed
to
the
final
solution
for
for
the
gaming
assignment
and
for
the
recession,
change
of
direction.
So
during
the
week
or
the
next
slide
on
head
upon
means,
we
lose
the
top
night
but
disappear
yep
I
we
went
to
the
you
question
of
what
do,
what?
D
There's
a
function
which
is
already
defined
is
the
penguin
function
in
the
which
was
fighting
in
749,
and
that's
the
only
one
defined
an
ending
function
is
basically
to
enter
the
IP
address
to
broadcast
address
and
it
makes
life.
However,
okay,
however,
there
are
many
other
functions
of
the
anchor
that
are.
There
are
some
of
them
are
there?
Some
of
them
are
not
there
and
in
order
to
try
to
take
them
into
account,
the
next
life
is
trying
to
distinguish.
C
D
The
distinguished
of
the
angling
function
of
the
IP
address
anchor
IP
addresses
versus
the
bottom.
One
is
the
section
the
tub
on
entering
the
IP
address
is
suffering
from
in
almost
every
chooses
your
ankle.
So
the
question
is:
what
are
you
angry?
What
I
do
at
your
engine
and
internet
IP
addresses
obvious,
but
the
other
one
is
to
anchor
the
section
and
we
normally
put
them
together
as
a
single
move,
pretty
anchor
or
separate,
and
on
and
and
a
difficult
situation,
you
can
ask
more
deeply
what
are
they
angling
rhetoric,
brothers,
etc?
D
Now
you
can
separate
them
and
an
idea,
justice,
obvious,
ok
and
then
Evelyn
section
we
will
talk
about
section
will
use
the
thing
of
it
end
to
end
car
as
that
when,
however,
if
you
put
an
anchor,
so
this
is
something
that
you
like
you're
trying
to
get
it
passed
through
and
they
are
scenery
which
take
their
this
Englishman
each
other,
so
I'm
going
to
go
to
some
of
them
next,
one
suppose
you
are
entering
the
IP
address.
Only
ok,
so
its
own
have
an
IP
address.
Then
they
are
true.
D
They
are
sionary
in
which
you
use
the
IP
address
of
the
network,
so
today,
so
no
power,
no
reason
you
have
to
network
that
one
on
one
side
that
you
on
side
the
person,
the
11
Angus
idea,
just
one
and
anger
level,
two
angles
idea:
just
you
and
you
can
are
the
mobile
node
with
an
IP
address
or
I
beat.
You
could
be
running
section
which
is
using
deception
has
likely
to
so
it's
using
it
in
the
next
life.
D
Is
that
exactly
the
same
thing,
except
that
you
are
using
and
idea
just
not
of
the
network,
so
the
section
is
IP
one.
On
the
right
hand,
side,
everything
is
the
same,
except
that
they're
I
p3
is
using
a
section
identified
by
IP
one,
which
is
not
the
United
to
lose
money
in
that
section.
That
would
take
care
of
a
lot
of
some
herbs.
/
example
alpha.
I
was
looking
at
scenario
in
which
you
are
using
address,
which
does
not
belong
to
the
network.
D
Okay,
so
I'll
chop,
and
then
the
next
line
is
that
they're
suicidal
mobility.
You
want
to
keep
IP
address
in
the
mid
section,
so
I
IP,
IP
one
and
you
move
to
a
different
network
as
you
move
you're
keeping
that
IP
address.
But
if
you
keep
that
I
PHS,
you
need
to
have
some
of
it
if
Politico
to
do
it.
So
this
is
not
human
scope.
Yet
because
this
is
my
be
addressed
in
the
next
slide
is
that
you
are
releasing
a
tough
time
and
we
are
changing
the
I,
the
ending
up
IP
address.
D
D
So
if
you
don't
want
to
change
IP
address,
one
classical
thing
is
to
actually
move
the
entering
of
octave
entering
functioning
from
the
number
one
to
number
two.
Okay,
you
move
that
the
amount
of
water
that
you
but
the
anger
also
look
too
natural.
So
as
you
move
into
so,
the
welding
will
be
taking
hell.
So
as
you
move
okay
on.
D
E
D
D
C
A
D
They
are
in
the
same
their
co-located,
no,
not
a
beautiful
okay
that
they
will
see
that
later.
So
it
goes
on
this
slide
in
this
side
are,
is
basically
the
role
of
demise
ation.
What
people
know
it
talkable
optimization.
D
C
E
One
quick
question:
I'm
having
a
tough
time
in
following
this
right.
So
what
is
the
session
anchor?
What
is
an
IP
anchor
is
IP,
you
know,
so
what
I
understood
is
you're,
putting
your
host
route
for
it
or
it
not
even
a
host
or
your
routing,
a
specific
IP
flow
through
one
path
and
you're
still
keeping
the
mobility
session
on
a
different
on
a
different
anchor.
Is
that
what
it
is?
Are
you
talking
about
a
if
this
intercom
of.
D
D
There
was
a
sly
some
somewhere
before
on
the
definition
of
the
second
anchor
the
session.
The
end
of
the
session
means
you
try
to
cause
the
chocolate
to
go,
for
it
is
one
of
the
section.
Ellis
was
not
not
Oh,
for
example
you,
if
you're
using
at
any
cost,
then
it
will.
We
go
for
one
of
them,
but
not
out
there.
Okay,.
E
Be
very
simple
question
right:
how
is
the
forwarding
plane
in
the
IP
network
able
to
steer
one
particular
IP
flow
on
one
path
and
all
the
rest
of
the
flows
on
a
different
path?
Are
you
assuming
that
there
is
a
flow
based
stitching
in
the
IP
network?
Are
you
making
flow-based?
You
know
forwarding
decisions.
D
D
D
F
Michael,
it's
just
a
clarifying
question:
can
you
go
back
to
the
previous
slightly
just
for
this
case
and
the
previous
case?
So
in
this
case
that's
the
most
obvious
and
pragmatic
scenario,
and
here
the
mobile
node
keeps
ip1,
even
though
it
moves
to
the
excess
router
anchor
but
excess
water
to
is
not
an
anchor
at
all
this
scenario
right,
so
the
anchor
one
remains
the
anchor
and
forwards
packets
to
the
milanos
current
location
on.
D
D
C
D
F
D
F
D
F
D
E
Know
I
think
just
to
add
to
their
clerk
I'm,
not
following
this
in
the
past.
What
we
talked
about
was
in
everybody
agreed.
We
talked
the
word
session
rankin
right
essentially
starting
point
is
a
client
gets
an
IP
address
or
prefix
whatever
that
subscribers
session
is
anchored
on
a
node,
how
we
move
that
subscriber
session,
along
with
the
prefixes,
our
transistor
different
anchor
point
right.
We
talked
about
that
interactions
and
that
we
also
talked
about
you
know
prior
work,
which
was
the
global,
aha
work
which
was
done.
E
You
know
many
years
back,
we
I
thought,
or
at
least
many
people
thought
we
are
going
to.
You
know,
bring
that
work
back
into
life
and
you
know,
reuse
leverage
all
of
that
work,
but
this
is
completely
going
in
a
different
direction.
I
am
NOT
following
how
it
aligns
to
what
we
discussed
in
the
past.
We.
D
Had
a
definition
of
angling
function
which
which
is
defined
in
RFC
749
I
mean
we
do
not
have
any
definition
of
over
the
anchor
in
any
our
see.
The
only
definition
is
an
anchoring
function
and
we
usually
top
of
it
as
a
knife,
and
that
function
is
on
an
IP
address.
Is
they
simply
to
to
advertise
the
prefix
for
that
I
PHS?
That's,
that's!
That's
what
we
have
met.
E
D
E
D
C
Okay
cattleya,
so
I
clarify
of
your
session.
I
think
that's
the
point
of
confusion
right
now.
It
sounds
like
to
me
what
you're
saying
is
that
you
have
an
IP
address
right
and
then
you
have
some
cup
session
/
flow,
which
is
not
specific
related
to
the
IP
of
the
mobile
device.
Right.
You
have
a
session
almost
by
itself
that
can
move.
That's
independent
of
the
IP
address
that
the
mobile
device
own.
D
C
I
think
I
think
there's
a
lot
of
confusion
or
issue
with
that
right.
One
is
that
under
mobile
side
does
the
mobile
still
only
IP
address
where
the
session
is
targeted
for
that's
one
question
and
the
other
parts,
the
transport
of
that
flow
to
your
anchor?
Is
it
through
the
routing
topology
or
what
boarding
topology
allows
that
to
work
right.
So
maybe
that's
to
point
that
is
unclear.
Maybe
I
don't
know,
maybe
it's
in
the
draft,
but
I
think
that
is
if
you
can
clarify
that
would
be
helpful.
That's.
D
So
many
previous
life
go
to
up
to
cycle
poco.
Basically,
there
are
lots
of
different
ways
to
do
with
the
mobility
us
you
can
and
some
people
call
the
anger.
Is
it
the
way
that
the
packet
is
going
for
it
or
whether
you
can
use
the
aggression
coming
or
or
some
of
that
on
or
to
some
of
these
detection,
so
they're
different
method
that
the
sex
that
be
that
the
section
that
the
end
of
the
section
and
you
such
as
the
redirected,
inductor,
indirection,
Stewart,
hungry
and
cashew
ation?
C
C
Yeah
yeah:
that's
why
I'm
saying
that
for
tunneling
I
think
you
can
do
that,
but
for
routing
you're,
not
routing
a
flow
level.
All
right
I
mean
you're,
saying
all
these
different
four
ways
are
ways
of
doing
it,
but
anyways
it's
okay.
I
am
I'm
just
trying
to
understand.
It
doesn't
quite
make
sense
to
me.
I
guess
at
the
end
of
day,
so
now.
D
We
go
Jeong
Jeong
for
about
five
or
six
teaspoons
nine.
Now
this
is
I.
Think
this
one
is.
This
is
the
end
result
of
the
IP
adjust
only
there's
nothing
to
do
with
this
action.
Okay,
it
is
entered
the
idea
just
only
okay,
it
doesn't
have
none
of
those
function
and
it's
basically
moved
to
anger
to
the
other
side.
Ok,
so
the
chocolate
will
float
okay.
So
this
is
a
this
email
didn't
touch
on
the
second
anger
and
then
go
to
the
next
line
and
yeah.
D
This
one
is
a
typical
mobile,
IP
scenario:
okay,
we
added
a
section,
the
end
of
the
session,
the
red
box
on
the
left,
hand,
side
and
the
right
box
on
the
other
side.
This
is
a
typical
premium,
c-,
okay,
because
this
one
is
thus
that's
the
tendering
it
keeps
the
binding
and
so
on.
So
it
has
a
binding
of
the
session
using
IP
address
one
over.
D
D
Because
the
anger,
the
session
authority
in
the
second
you
are
in
Hungary
and
then
it
go
directory
okay.
So
this
is
a
typical
row,
optimization
scenario
now
in
a
relative,
my
sessions
in
aerial,
the
anger
of
the
address
doesn't
didn't
change,
but
the
packet
doesn't
go
to
that
anymore.
I
mean
it's
do
advertising.
It
doesn't
go
there
anymore,
but
the
packet
book
directly
go
to
the
other
side.
Okay,.
E
Optimization
is
different
because
the
way,
if
you
look
at
the
way
route
optimization
is
working,
the
the
real
the
care
of
address
and
the
home
address
are
still
present
in
the
packet
headers.
It
is
the
way
use
of
in
a
destination
option
cider,
and
you
know
with
other
logic,
it's
happening.
It's
not
that
you
know
you
cannot
take
route.
Optimization
is
a
great
example,
and
you
know
relate
to
this
I'm.
D
Just
keeping
it
as
high
level,
okay,
because
this
is
that
this
is
not
doing
that
exact,
my
exactly
bollinger,
but
it's
keeping
it
as
an
allergy
right,
you're
right
it
doesn't
use
because
they're
different
photo
shows
I'm,
trying
to
harmonize
all
of
them
by
an
EP.
If
you
try
that
it
seems
to
me
the
Deacon
proposal,
can
we
catch?
You
see.
E
D
A
D
A
Definitely,
and
bring
bring
this
up
like
like
the
more
clarity
what
you
mean,
but
about
this
session
and
then
and
when
you
actually
see
it
move
to
anchor
function.
But
you
are
still
advertising
the
extreme
prefix
ho.
If
you
can
describe
that
with
existing
technologies
or
similar
technisat,
we
have
today
that
would
basically
make
it
easier
to
understand
what.
A
A
D
A
A
D
D
D
D
C
G
So
this
is
the
report
from
both
exposure
and
selection
working
team.
Next,
like
please
so
senseless
idea.
If
we
held
two
conference,
calls
in
conference
call
number
three:
you
see
the
date
the
attendees
the
agenda
was
to
discuss
the
item
number
one
on
our
plate,
which
involved
defining
at
the
API
between
the
applications
and
the
IP
stack
on
the
mobile
node,
and
for
that
we
had
two
drops
on
the
table.
G
You
can
see
the
notes
on
that
link
which
I
will
quickly
capture
in
this
presentation.
Next
slide,
please!
So,
basically
for
item
number
one
in
earlier
meetings,
we
had
identified
a
number
of
principles.
Basically,
we
had
agreed
that
we
gonna
define
an
extension
to
RFC
32
14,
which
is
the
source
of
this
selection
algorithm
and
the.
G
Next
slide,
please
the
first
draft
the
check
marks
well
anyways.
It
does
work
with
the
RFC
50
40
framework.
It
does
define
those
three
different,
three
new
types
of
flags
and
also
covers
the
case
when
the
requested
time
I
Gladys's
not
only
configured
it,
has
the
hooks
to
go
and
configure
them
and
make
them
available
applications.
G
So
this
draft
satisfied
those
three
principles
we
had
identified
already
on
next
one,
please
the
next
draft
it
it's
also
based
on
extending
our
CP
214,
but
it
only
defines
two
flags,
not
three,
and
we
couldn't
fully
understand
the
semantics
of
those
flags.
The
authors
were
not
going
to
call.
We
welcome
further
elaboration
on
that.
Nevertheless,
it's
obviously
missing
the
third
type,
even
with
the
right
semantics
and
there's
also
no
notion
of
dynamically
configuring.
G
So
the
working
team
is
based
on
proposing
to
proposing
working
group
to
adapt
this
draft
as
the
base
lab
and
I
have.
While
we
have
captured
that
on
the
meeting
minutes
and
also
last
week
or
we
can
help
ago,
I
sent
an
email
on
the
mailing
list
indicating
our
intention
to
propose
that
so
either
rights
now
or
eight
end
of
the
presentation,
I'm
hoping
the
chairs
will
ask
that
chin
to
the
working
group
for
adoption.
B
A
B
G
B
B
B
G
B
G
Again,
I
think
you're
talking
about
something
that
we
have
this.
We
had
gone
over
at
the
previous
idea.
We
have
shown
how
this
item
fits
into
the
charger.
So
if
you're,
if
you're
asking
if
this
fits
into
the
charger,
we
had
already
addressed
that
in
the
previous
ITF
meeting
sure,
maybe
you
pull
up
these
slides
from
the
archive
and
then
we
talk
about
that
now
I
mean
we
have
passed
it
I.
B
B
B
E
As
my
suggestion,
maybe
you
know
I
largely
champion
break
so
doping.
You
know
this
is
an
able
account
that
you
know
we
made
some
assumptions
on
that
for
the
BM
solutions
right.
We
made
certain
assumptions
right.
You
know,
for
you
know,
you
know,
avoiding
you
know,
essentially
to
avoiding
host
halls
or
whatever
reasons
right.
We
thought
for
awarding
tunneling
or
other
approaches.
Without
demon
is
a
vote.
You
know
smaller
cells.
They
will
lead
to
further
client.
You
know
wherever
it
anchors
that
becomes
0.
You
know,
local,
the
local
gateway
becomes
anchor
right
later
it
moves.
E
You
know
when
it
moves
from
that
anchor
to
a
different
location.
The
idea
was
to
enable
the
client
to
you
know,
get
a
you
know,
obtain
a
local
IP
address
from
that
network
and
over
time
phase
out
the
address
from
the
that
that
it
was
using
the
previous
network
or
previously
very
close
anchored
right
from
the
so.
The
client
at
a
point
of
time
has
multiplied
right,
but
he
doesn't
know
which
IP
address
is
born
right
now,
with
this
world,
we
are
enabling
the
client
to
be
more
intelligent
right.
E
B
G
H
G
Guys,
I,
don't
think
you
guys
are
following
these
discussions.
Careful,
I
gotta
tell
you
this.
We
have
breaking.
We
have
broken
down
this
problem
into
pieces.
Okay,
one
piece
is
how
the
application
indicates.
Its
intention
is
that
this
is
that
another
piece.
Another
piece
which
is
not
in
this
draft
is
how
the
ipsec
on
the
horse
excellent
yeah,.
G
I
A
So,
just
to
clarify
so
what
opera
is
presenting
is
tested,
one
enable
of
the
whole
set,
so
we
need
the
network
to
host
part.
This
is
d
from
the
application
to
stack
part.
So
it's
just
like
likely
and,
and
then
these
world,
whatever
is
presenting,
is
the
independent
of
the
actual
network
to
host
communication.
A
How
you
realize
that,
whether
you
use
DHCP
slack
or
whether
it's
a
mobile,
IP
or
something
else
like
whatever
singling
there
is
I
person
to
pick
it,
but
you
still
need
for
applications
and
tool
to
indicate
like
I
want
to
have
cars
in
the
instance
of
bolsa
whatever,
and
we
currently
5040.
No,
she
already
has
the
concept
of
CoA
and
home
address,
but
that's
purely
from
the
mobile
ipv6
point
of
view,
and
now
in
the
dmm
we
have
the
kind
of
thing
like
this.
A
Some
addresses
are
okay
to
call
and
we
want
to
know
a
bit
more
about
what
kind
of
address
the
application
is
going
to
ask
for.
So
that's.
Where
is
heading
to
I.
I
B
Think
it's
still
open
because
I
I
don't
think
this.
This
reflex
is,
can
solve
all
the
problem
of
the
dynamic
anger
such.
I
I
Let
me
join
in
and
answer.
First
of
all,
it's
not
supposed
to
answer
to
answer
all
the
problems
that
we
have.
What
we
talked
in
the
past-
and
we
identified-
is
that
if
you
think
about
supporting
mobility
and
anchoring,
there
is
some
overhead
in
the
network
in
order
to
support
anchoring,
either
via
host
routing
or
via
some
tunneling,
and
things
like
that,
and
another
thing
that
we
detected
was
that
some
applications
don't
require
this
over
heading.
For
example,
some
applications
won't
break
if
the
source
IP
address
changes
while
they
are
in
session.
I
So
what
this
draft
proposes
is
to
identify
various
types
of
services
that
are
required
from
the
network,
the
three
types
of
services
and
proposes
that
the
application
convey
to
the
IP
stack,
what
type
of
service
they
require,
and
then
the
IP
stack
through
other
means
will
communicate
that
to
the
network,
and
then
the
network
will
know
to
allocate
the
appropriate
resources
for
those
specific
applications.
This
is
what
this
draft
is
about.
Yeah.
B
I
A
A
F
A
H
H
G
G
G
H
A
G
H
A
H
A
You
don't
really
need
to
request
a
kind,
for
example,
when
the
network
provides
an
address.
It
can
say
that
this
address.
Has
these
properties
like
it's
going
to
be
anchored
and
then
then,
to
stack
when
you
asked
if
you
asked
for
an
address
that
hasn't
Ankara
and
the
address
that
was
given,
you
from
network
doesn't
have
that
information,
then
tis,
chlorophyll
binding
to
that
kind
of
address
it
fails
in
a
host.
Can.
H
G
H
A
H
A
G
Are
like
LEGO
pieces?
Okay,
they
can
all
clicking
together,
but
that's
not
a
gating
factor
for
working
group
adoption,
but
that
would
be
a
gating
factor
for
like
publication,
I
mean
these
things
don't
hold
up.
We
can
always
drop
things,
but
for
working
of
adoption
we
don't
have
to
wait
for
all
pieces
to
be
lined
up.
H
G
H
Did
the
two
options
that
you
shown
are
already
specified
in
RFC
and
are
not
implemented
in
operating
systems?
Okay,
for
example.
You
say,
if
convicted
a
you,
don't
see
that
others
being
a
home
address
or
a
care
of
others
or
particular
type
of
other.
So
you
know
it's
it's
you
have
the
address,
that's
all,
and
maybe
it's
prefect.
That's
all
so.
A
G
E
On
point
number,
one
right
between
the
you
know:
I
think
that
two
parts
to
it
I
think
I
just
want
to
repeat
that
what
alpha
is
saying
right,
essentially
allowing
the
application
to
be
aware
of
that
dress.
Property
is
one
thing
and
then
allowing
the
horse
to
you
know
obtain
a
particular
type
of
address
from
the
network
is
another
problem.
Another
aspect
that
is
both
on
dhcpv6
anybody
scoring
right.
We
proposed
extensions
right.
There
is
some
work
that
happened
in
with
I
agree.
E
There
was
you
know
we
publish
this
sometime
back
it
did
it
made
it.
You
know
there
were
some
because
of
some
other
work
in
dhcp
took
a
slow
track.
Now
that
we
have
proper
use
case
and
a
motivation,
we
should
be
able
to
that
those
documents
and
amateur
shape.
We
should
be
able
to
move
them
on
a
fast
track,
so
the
linkage
is
there
now.
With
respect
your
other
comment,
if
5014
saw
you
know,
some
socket
option
is
not
implemented
in
some
linux
kernel.
That
has
nothing
to
do
with
this
discussion.
E
F
So
nowadays,
when
you
program
applications
you
can
you
have
the
means
to
bind
the
traffic
particular
interface
or
IP
address
so
I
see
the
value
here
in
that
you
simplify
this,
that
the
application
just
says:
I
need
this
kind
of
IP
address
and
the
colonel
takes
care
of
the
rest
right.
So
you
mentioned
here
if,
for
example,
an
application
requests
matting
address,
the
colonel
takes
care
about
configuring
one
if
there
is
no
no
matta
good
friends,
but
I'm
not
sure
if
this
overcomplicates
poor
thing.
F
G
G
G
G
A
G
Call
the
agenda
was
to
discuss
principles
for
item
number
two
and
three.
Basically,
these
are
the
items
for
our
working
team
to
describe
how
the
IP
configuration
protocols
are
extended
in
order
to
support
this
IP
address
types,
okay
and
basically
network
at
the
time
of
configuring,
an
IP
address
or
prefix
on
the
Hajj
should
have
a
way
to
convey
the
type
of
the
IP
address
or
the
prefix.
Similarly,
the
host
at
the
time
it's
requesting
I
cut
it.
You
should
have
a
way
to
convey
the
type
it
needs.
G
So
we've
gone
over
a
bunch
of
iPads
configuration
techniques
and
we
recognize
well.
We
basically
identified
the
need
for
extending
ECB
v6
step
sizes
of
the
configuration
mobile,
ipv6
and
Ike
v2
in
order
to
plug
in
the
bottom
part
of
this
API.
Ok
right
now,
the
next
step
is
to
we
open
the
floor
for
people
to
propose
solutions
for
that
those
items,
and
we
have
one
that
we
will
see
on
thursday.
Ok,.
A
G
A
G
F
Simple
way,
my
first
concern
is
that
you
will
like
lie,
cause
you
to
define
three
cancel,
IPA,
fridge,
perfect
and
further,
and
the
application
developer
may
be
confused
about
the
prefix
and
they
they
will
sell.
Seldom
you
use
these
candle
prefix,
and
my
second
concern
is
that
is:
is
the
definition
of
the
IP
free
products
and
sake?
Option
should
be
done
in
the
MMO
workgroup.
How
in
other
IPO
related
a
loop
work
group.
G
F
G
F
G
Yeah,
well,
that's
why
we
documented
here
so
the
applications
that
wants
to
have
efficient
networking
through
the
techniques
would
adopt
this
API,
which
we're
going
to
document.
There
are
the
white
papers,
presentations
and
education,
all
that,
so
that's
what
an
API
takes
to
get
adopted
in
the
field,
so
hopefully
people
will
read
and
understand
them
correctly.
So,
ok,
that's.
F
G
Mean
this
item
number
23
well,
this
is
something
that
we
will
need
to
discuss
them
in
proposals
probably
will
be
considered
here
first
and
then
they
may
be
taken
to
respect
to
working
groups.
But
again
I
miss
my
personal
opinion.
Chairs
and
80
may
have
different
opinion
on
that
server,
but
at
least
they
will
first
discuss
here
to
make
sure
we
got
to
semantics.
A
G
A
G
A
C
A
So
let
me
point
some
things
out,
so
these
are
working
teams.
They
have
been
working
openly.
Anyone
could
join
these
working
teams,
they
were
fully
open
and
all
these
documents
are
in
part
of
any
other
individual
document.
These
are
just
interest
group
of
people
working
together
and
they
put
together
a
document,
so
they
are
not
really
a
design
team
product,
stale
bunch
of
people
with
the
same
interest
working
on
something
and
posting
them.
Yes,.
H
I
agree:
it
is
good
to
have
open
working
teams
and
open
all
your
conferences
and
the
phone
calls
were
announced
on
the
email
list,
and
this
is
very
good
on
another
hand.
It
is
at
the
meeting
at
IDF
that
decision
should
be
made
right
and
we
should
not
be
patient.
Somebody
who
has
not
participated
to
the
audio
conference.
C
H
A
So
it's
an
open
process.
If
someone
doesn't
participate,
open
process,
it's
his
own
or
her
own
fault,
and
and
then
all
these
you
can.
Anyone
can
propose
the
same
in
exactly
the
same
document
on
the
same
area
and
ask
for
an
adoption.
They
have
been
meeting
like
four
times
posted
two
minutes
on
the
mailing
list
at
this
is
what
we
discuss
and
so
on
and
da
da
da
and
now
they
just
went
and
asked
for
the
adoption
of
the
document.
Anyone
else
could
have
done
the
same
proposal
document
and
ask
for
representation
talents
on
it.
A
Of
course
not,
and
that's
why
everything
will
be
handled
here
on
the
mailing
list
and
enter
meeting
that
we
have
so
it
will
be
run
through,
like
any
other
individual
document.
There's
no
difference
on
that.
The
reason
why
we
actually
ended
up
with
this
process
is
like
that
the
is
to
reduce
the
amount
of
documents
on
the
same
topic.