►
From YouTube: CORE WG Interim Meeting, 2020-06-24
Description
CORE WG Interim Meeting, 2020-06-24
A
A
So
your
chairs
today,
as
usual,
are
Marco
De,
Luca
and
me
I'm,
a
humanist
I
didn't
get
any
mini
takers
to
volunteer
before
yeah
I
will
give
it
another
try
now!
So
please,
if
you
would
like
to
come
to
you,
even
if
it
is
partially
anything
really
that
will
help
the
discussion.
I
would
expect
today.
So
you'll
be
really
nice
to
have
the
minutes
of
the
meeting.
It
is
recording
anyways
but
I.
A
Well,
I
mean
we
will
make
with
current
minutes
and
anyway
it's
recording
so
he's
recorded.
So
that's
not
so
the
note
well
I
just
need
to
remind
you
that
no
dwell
applies
for
IPR
and
for
they
working
group
processes
and
the
contact,
and
but
in
general
I
mean
as
far
as
corner
goes
I
guess
you
know,
try
to
be
nice,
we
kind
of
know
each
other
anyway.
So
should
we
find
a
few
updates?
A
Well,
first,
you
see
here
the
Korean
be
that
I'm
hopefully
sharing
for
those
of
you
who
are
not
familiar
with
it.
It
seems
to
be
the
way
to
go
in
the
ITF
now,
so
we
will
probably
replace
etherpad.
He
has
authorized
features.
You
can
see.
The
the
output
of
the
markdown
in
a
nice
format
allows
for
plant
UML
and
other
diagrams
I'm
not
going
to
do
right
now,
so
very
convenient.
You
don't
need
an
account.
This
is
currently
free
access,
but
eventually
you
may
want
to
have
a
data
to
perk
up
on
the
core
updates.
A
Then
Oscar
group
come
so
a
theme.
Marco
samy
did
a
new
version
yesterday
or
today.
So
it
might
be
time
to
do
work.
Yes,
sorry,
yesterday,
yes,
yet
yeah,
so
it
might
be
time
to
working
on
last
call
so
might
be
a
good
time
to
start
looking
into
it.
Hopefully
I
guess
two
or
three
weeks.
Is
there
any
urgency
by
the
way?
Can
we
do
three.
A
Yeah
I
mean
that
he
will
do
another
update
today.
That
does
the
current
Barrett.
Basically,
so
we
will
probably
do
also
wrap
up
this.
This
drop-
and
that's
so
much
for
us-
update
high-level
updates
that
I
have
right
now
that
we
have
right
now
for
core,
so
we
can
get
into
the
actual
meeting
discussion
and
I
think
that
would
be
now
Christian
right.
F
Okay,
okay,
I'm
just
pasting
the
two,
so
I,
don't
really
have
much
kind
of
slide
dish
things
to
go
with
this
I'm
I'd
like
to
just
give
you
a
brief
update
on
on
where
we
are
with
resource
directory
and
I
have
to
apologize.
I
wanted
to
have
this
in
a
much
better
state
by
today,
but
I
just
didn't,
manage
and
hacked
up
with
a
few
things
would.
A
F
I
you
can
either
could
you
could
you
basically
follow
the
link
that
I'm
yeah
that
I'm
that
I'm
pasting
there?
So
let
me
just
update
this
so
so
with
that,
if
that
fragment
identifier
that
would
get
us
to
basically
what
has
happened
so
far
in
resource
directory,
which
is
basically
running
through
the
comments
that
we
got
from
from
the
from
the
feedback.
This
is
from
the
station's
currently,
so
that's
area
director
review
and
also
the
other,
the
other
areas
input.
Most
of
this
is
really
editorial.
I
mean
there
was
boilerplate
text
that
was
wrong.
F
There
were
things
about
the
Ino
registrations
that
that
could
be
improved,
so
so
really,
the
the
kind
of
the
small
stuff
is
represented
in
here
you
yeah,
okay,
in
in
here
the
most
invasive
changes,
probably
judge,
which
I
would
just
did
a
few
minutes
ago,
replacing
kind
of
consistently
referring
to
the
thing
as
the
Rd
as
opposed
to
Rd,
sometimes
and
sometimes
resourcedirectory
capital.
Is
that
capitalized
not
capitalized,
so
this
is
kind
of
consistent.
Now,
the
larger
the
larger
chunk
of
things.
F
That's
still
that's
kind
of
on
my
on
my
desk
right
now
is
the
next
link.
If
you
could
follow
that
as
well,
which
is
fro
saying
what
the
feedback
was
on
the
on
the
security
consideration.
So
my
view
of
the
development
of
the
security
considerations
was
that
we
didn't
have
much
initially
and
then,
with
each
round
of
feedback
there
God
in
cases
of
you
should
consider
this.
You
should
consider
that
and
describe.
How
is
this
done?
How
is
that
done?
F
We
had
to
talk
about
this,
and
my
takeaway
from
that
was
that
we
should
rather
leave
those
things
to
the
application
and
say
more
about
how
does
an
application
arrive
at
a
decision
of
what
to
protect
as
opposed
to
you
must
protect
this
and
this
and
that
and
then
everything
will
be
safe,
which
won't
be
because
your
application
requirements
will
be
different.
So
what
I've
started
to
Sketch
up
was
so
it's.
This
is
in
a
pull
request.
F
What
you're
seeing
here
is
the
rendered
version
of
that
it's
an
updated
version
of
the
security
policies
which
I
hope
to
finish
today
or
tomorrow,
but
this
shows
the
general
direction,
that
is
to
say
what
are
the
things
that
can
be
in
a
resource
directory
that
you
may
want
to
protect.
And
what
do
you
tap?
What
what
do
you
need
to
consider
when
you
do
that?
F
So
one
thing,
basically,
basically
the
four
big
topics
are,
you
may
or
may
not
want
to
protect
the
endpoint
name,
in
which
case,
if
you're
going
passage,
if
it
gets,
then
you
would
probably
have
that
in
this
in
the
common
name.
But
let's
give
one
very
particular
example,
and
we
are
not
giving
normative
specific.
We
don't
specify
how
this
is
done,
but
just
give
kinds
of
you.
You
will
have
to
find
some
way
to
pick
that
information
out
of
there.
If
you
want
to
protect
that,
if
you
want
to
protect.
F
On
the
other
hand,
what
is
in
the
resource
links
that
is,
for
example,
if
you
have,
if
you
advertise
firmware,
update
service,
and
you
don't
want
your
clients
to
just
ping.
Everyone
who
claims
to
have
a
firmer
server,
then
the
and
the
claims
that
would
be
in
there
would
be,
would
look
quite
different
if
you
could
scroll
down
a
page
or
so.
F
The
the
section
on
link
confidentiality
is
kind
of
very,
very
incomplete,
as
is
the
rest,
but
basically
that
says
that
will
say
if
you
post
anything
to
the
resource
directory.
That
is
not
what
you
advertised
in
your
well
incur
anyway,
then
think
twice
before
you
do
that,
and
on
the
other
hand,
if
you,
if
you're
in
a
situation
where,
as
things
like
that
happen-
and
it
will
be
up
to
the
resource
directory,
to
authenticate
the
lookup
clients
as
well.
F
If
everything
in
there
is
intended
to
be
public
anyway,
there's
no
need
to
authenticate
look
at
progress
and,
in
the
end,
all
those
things
could
be
combined
and
different
sectors
of
a
resource
structure
might
have
different
policies.
So
I
think
this
can
largely
replace
the
section
7
that
used
to
be
there,
with
kind
of
taking
in
snippets
of
that
which
are
which
are
still
relevant
and
would
then
also
slim
down
a
bit
the
security
considerations
section
just
because
it's
it
will
become
morphing.
F
F
F
He
said,
there's
there's,
there's
one
concern
that
we
that
we
that
came
up
that
we
might
want
to
talk
about
briefly
in
in
this
group
here.
That
is,
that
a
resource
directory,
if
it
doesn't
do
source,
authentication
or
client
loudness
authentication,
if
it
doesn't
it
into
a
plant
aliveness,
then
it
could
be
used
to
read
data
from
devices
on
local
network,
which
is
something
that
is
generally
would
so,
is.
F
F
I
yeah,
that's
I'll,
just
put
the
link
in
there.
C
G
C
F
F
Barely
enough
to
you,
I
don't
know
about
the
others.
I.
D
B
F
So
I
don't
know
whether
custom
is
whether
you're
typing
somewhere
or
not,
but
basically
I'll.
Take
this
as
okay
yeah.
We
still
want
simple
registration,
even
if
it
means
that
there
will
needs
to
be
an
echo
round
trip
to
verify
client
aliveness,
that's
fine
with
me
as
long
as
there's
yeah,
so
I
think
the
this
the
last
the
this
big
point
about.
What
do
we
want
to
say
about?
F
What
do
we
want
to
say
about
how
authentication
is
done
for
the
resource
directory
leads
us
to
the
I
mean
on
one
hand,
I'd
like
to
have
a
bit
of
feedback
of
whether
this
is
actually
the
direction?
The
working
group
wants
this
development
to
happen
because
I'm
not
working
with
much
here
so
I'm
just
trying
to
get
a
direction
that
will
get
this
on
and
I.
Don't
know
what
that's
right
wrong
and
I
think
that
discussion
will
also
lead
us
to
the
next
point
of
how
will
those
details
then
be
done?
F
So,
just
on
the
procedural
side,
what
I
will
do
is
Kentucky
finished.
This
pull
request
send
out
a
million
through
the
list
for
a
day
or
two
whether
anyone
wants
to
put
something
in
right
in
there
and
otherwise
make
this.
A
new
draft
version
ask
around
on
the
list
again
and
claim
that
this
is
addressing
the
reviews.
Comments,
I'm,
based
on
that
yep.
H
C
H
A
H
So,
basically
in
February
the
risk
of
territory
draft
at
the
sentence.
That
basically
said
if
you
need
authorization
and
then
please
use
ace
and
what
we
didn't
at
that
meeting
was
try
to
figure
out
how
the
generic
flow
would
be.
If
you
use
ace
I
have
the
note
somewhere
on
my
computer,
I
didn't
find
any
online
have
used
etherpad
I,
don't
remember.
A
G
E
A
Just
like
we
have
the
simple
registration
in
our
game
play,
something
else
will
be
done
as
price.
You
know
the
tool,
I
don't
know,
I,
don't
know
what
it
is.
If
it
is
there
having
the
resource
directory
as
a
rotation
server
and
is
having
one
single
entity
or
that's
better
or
not,
but
I
was
hoping
to
have
that
discussion
today.
Here
now.
F
F
But
what
is
I
think
the
most
like,
or
the
more
likely
situation
is
that
they,
the
registrant,
will
try
to
perform
its
perform.
The
the
registration
and
we
will
fail
with
your
credentials,
are
not
good
enough
forbidden
message
that
will
then
redirect
it
to
the
resource
or
to
the
authorization
server
from
where
it
will
be
able
to
get
its
crashes
or
not.
If
it's
not
authorized
to
and
combining
that
with
what
I've
been
starting
on
the
topic
of
the
of
the
various
ways,
a
resource
directory
can
can
define
its
policies.
F
I
think
this
is
the
more
realistic
scenario,
because
the
client
may
not
know
whether
the
resource
directory
will
limit
it.
Based
on
based
on
the
links
it's
trying
to
put
in
their
work
place
on
its
endpoint
name
or
once
or
not.
But
when
it
tries
to
do
the
registration,
then
the
resource
directory
will
not.
You
are
trying
to
do
this,
but
you
need
to
be
depending
on
the
set
up
authorized
to
use
the
name.
Ep
equals
that
name
in
the
sector.
F
G
F
In
look
up,
this
can
behave
the
same
way.
So
if
the
client
performs
a
lookup,
then
it
will
it
may
it
will
probably
hit
a
401
on
the
first
lookup
at
this
and
then
go
to
the
authorization
server
and
get
a
token
that's
good
for
that
particular
lookup,
provided
lookups
are
provided
lookups
native
indication,
I
think.
D
The
important
observation
is
that
the
resource
directory
is
agnostic
with
respect
to
the
onboarding
process
that
you
have.
We
are
not
trying
to
impose
a
specific
onboarding
process
here.
The
onboarding
process
has
to
come
from
somewhere
else,
and
the
resource
directory
is
just
an
application
that
helps
in
disseminating
information
that
is
so
useful
in
that
process.
F
D
There
are
different
ways
for
for
setting
up
system
system,
components,
components
finding
each
other
and
so
on,
and
we
are
not
trying
to
prescribe
a
single
way
for
for
the
components
to
find
each
other
and
and
obtain
our
authorizations
to
work
with
each
other
and
so
on.
We're
just
trying
to
provide
one
component
that
can
be
used
in
such
a
process.
D
Yeah
I'm
not
sure
we
even
need
another
case,
because
we
essentially
posting
to
a
certain
well-known
resource
and
should
be
described
described
by
AF.
The
interesting
more
interesting
question
is
who
can
actually
give
that
authorization
in
such
a
way
that
the
resource
directory
listens
to
it,
but
that's
again
outside
the
scope
of
the
research
we
got
so
much
mm-hmm.
F
I
agree
that
AI
F
should
be
able
to
do
this
without
any
any
special
casing.
We
might
need
to
have
a
way
to
include
query
parameters
in
an
AI
F,
but
then
again
whether
the
scope
is
to
two
posts
on
the
Reg
on
the
on
the
on
the
directory
resource
or
to
do
any
operation
on
the
registration
resource.
I,
don't
I,
don't
care
too
much
and
that's
something.
Basically,
the
resource
directory
and
the
authorization
server
need
to
agree
on
nothing
more
and
possibly
the
registrant.
I
F
Would
be
the
registration
resources,
so
I
think
that
this
this
so
basically,
whoever
is
allowed
to
post
to
say,
/rd
question
mark
EP
equals
foo
would
also
be
allowed
to
put
and
delete
slash
registration,
slash
whatever
it
takes
for
foo
and
either
of
those
can
be
used
to
describe
the
set
of
permissions
that
need.
Is
we
share
between
them?
Okay,.
A
David
and
Thomas
and
others
I
have
been
working
on
the
on
the
standard.
There
is
no
specific
authorization
required
for
registration
and,
as
far
as
I
do
there
is
no
lookup
even
present
I
wonder
if
there
is
some
some
platform
out
there,
using
our
be
an
ace
already
that
we
could
take
as
an
example,
because,
as
Carson
said,
we
are
just
building
components
and
but
without
having
the
whole
picture
is,
is
a
bit
hard,
at
least
for
me,
to
do
single,
who
optimize
things.
I
A
I'm
looking
for
someone
using
educational
look
up
in
which
two
different
parties
do
the
registration,
lookups
authorization
is
required
so
in
the
language,
including
sense,
and
please
David
or
Thomas
coram
me
from
wrong
when
Levitan
tone.
So
when
and
when
an
in-point
registers
on
what
is
called
observer,
which
is
running
a
resource
directory.
There
is
no
authorization
needed
because
there
is
some
either
pre
shared
key
or
some
synchronization
already
done.
A
A
G
F
So
to
understand
this,
that
lightweight,
m2m
server
would
have
original,
would
have
configured
knowledge
of
which
credentials
which
device
has
and
then
accept,
or
reject
registration
based
on
the
endpoint
name
when
it
comes
again
accepted
only
when
it
comes
from
the
device
that
matches
that
iPod
name.
This.
B
G
G
F
I,
wouldn't
really
put
text
about
concrete
age
usage
in
there
because
it
with
so
in
the
in
the
pull
request
that
I
showed
before
I
do
have
kind
of
some
examples
on
a
very
high
level,
where
I
do
mention
ace,
in
the
sense
that
this
will
basically
information
about.
This
will
end
up
in
the
scope
of
of
what
of
the
ace
token,
but
not
know
details
because
I
mean
the
details
will
depend
on
the
application
and
we
can
certainly
have
something,
like
example,
application
somewhere,
but,
for
example,
my
nose
will
always
be
suitable.
F
E
D
Snmp
version
3,
you
have
to
have
security
play
because
there's
there
is
no
way
that
nobody
is
going
to
define
that
for
you,
but
this
is
not
a
sending
snmpv3.
This
is
a
component
in
inner
space
where
people
already
are
building
security
solutions
and
whatever
they
are
providing.
We
be
able
to
provide
the
the
authorization
information
for
the
resource
directory.
A
Yeah,
if
I
actually
like
this
Co
police,
the
entering
was
not
to
actually
make
more
changes
to
the
resource
vector
a
draft,
but
rather
back
some
weeks
ago
he
was
found
out
or
when
we
were
having
a
meeting.
We
discussed
that
the
actually
ace
the
usage
of
ace
in
the
context
of
Rd
was
maybe
a
bit
complicated
and
we
were
wondering
if
he
will
be
work.
So
this
this
interim
was
about
that
about
how
to
simplify
the
usage
of
ace
for
early,
not
to
make
changes
to
Rd.
Okay,.
D
A
D
So
the
one
thing
that
we
haven't
really
discussed
in
ace
is
how
to
handle
dynamic
resources,
so
AF
can
describe
static
resources,
but
if
there
is
some
process
that
creates
resources
on
someone's
behalf,
then
you
would
have
to
define
how
the
authorization
that
was
initially
expressed
in
terms
of
access
to
the
resource
that
creates
these
dynamic
resources
would
confer
or
continue
to
the
dynamically
created
resources
and
mapping
ace
+
AF,
so
the
resource
droid
free,
probably
would
need
this
very
tiny
piece
of
glue,
essentially
saying.
If
you
created
it,
you
have
it
it's
yours.
F
D
D
E
E
F
D
D
D
Yeah,
so
the
the
client
should
be
a
weir
that
is
using
some
authorization
that
is
time-limited,
so
everything
that
the
client
is
doing
based
on
that
authorization
will
have
to
inherit
that
time
limit
both
on
the
server
side
and
on
the
client
side.
So
the
client
should
should
know
that
it
needs
to
re-register
yeah
before
that
authorization
actually
ends
yeah.
D
F
E
I
I
F
A
G
G
G
G
D
G
No,
there
is
nothing
at
the
protocol
level.
I
would
say,
but
indicate
that
just
that,
maybe
when
my
constants
in
all
case
when
we
want
to
make
a
key
a
key
update,
we
basically
invited
as
the
previous
key
at
the
lightweight
m2
observer
level.
So
when
the
device
cannot
well
as
receive
no
notarized
answer
from
the
server
because
he
is
using
the
are
not
up-to-date,
it
will
then
contact
contact
the
boots
web
server
and
the
boots
or
server
will
provide
him
with
the
updated
keys.
A
I
mean
I,
guess
the
lifetime
is
a
priori
infinite
unless
the
booster
observer
does
something
or
decides
due
to
the
implementation
characteristics
to
change
it,
but
what
I'm
thinking
is
not
if
I
eat,
one
company
has
one
light
within
them
server
and
another
hostile
evident
observer,
and
they
are
bound
to
the
first
ones.
Booster
observer
I'm,
not
so
sure
how
the
second
one
can
actually
change.
Is
that
case?
How
do
you
signal
the
keys
have
expired,
no.
D
A
F
Think
a
good
step
here
would
we
have
basically
have
some
text,
possibly
document
that
just
describes
an
example
setup
of
how
a
resource
directory
with
an
authorization
server
and
all
those
other
components
could
play
together
like
an
example
set
up
and
I
think
I
think
I
could
even
fit
that
in
at
least
preliminarily
in
the
in
the
resource
directory
extensions
in
the
sense
of
yeah.
This
is
this
is
one.
This
is
one
of
those
ways
how
all
those
things
could
be
combined.
A
A
You
have
any
input
because
we
have
had
discussions
on
RB
and
usability,
and
you
know
this
could
be
a
good
time
to
say
something
if
you
want
to
because
we
have
all
these
other
people
in
the
point
of
view
on
another
ace
experts
and
so
on.
H
I
don't
know
so.
The
original
idea
of
that
side
meeting
a
long
time
ago
was
to
take
the
components
that
we
develop
in
our
working
groups
and
just
put
them
together
once
to
see
if
they
can
actually
combined.
As
we
always
say,
we
build
the
components
and
then
let
other
people
combine
them,
but
maybe
there
are
some
gaps.
Maybe
some
clarifications
need
it
in
and
so
on
so
doing
this
non
very
idea.
Fee
activity
of
combining
components
seemed
like
a
useful
thing
to
do,
and
now
I
think
we
have
identified
some
areas.
H
I
H
And
in
theory,
everything
that
needs
to
be
implemented
should
be
there
right.
So
you
would
do
the
first
step
of
discovering
your
resource
directory
as
described
in
the
ID
draft.
Then
you
would
probably
try
to
do
a
registration.
You
would
fail
because
you
don't
have
the
authorizations
and
then
you
get
bounced
to
an
authorization
server.
For
that
we
should
have
a
draft
that
says
how
this
redirect
works
and
and
and
so
on.
So
in
theory,
all
the
components
should
be
there
and
it
should
be
easy
to
go
to
put
them
together
in
a
hackathon.
Also.
I
C
I
I
H
F
Also
from
from
the
authorization
server,
so
if
the
commissioning
tool
does
not
know
the
structure
of
the
scopes,
it
will
just
be
upgrading
its
token
time
and
again
and
again
and
again
for
every
registration
it
does.
If,
of
course,
the
commissioning
tools
to
me
gets
a
bit
more
knowledge
about
how
those
things
look
like
it
might,
it
might
get
it
might
ask
the
authorization
server
for
a
token
that
actually
represents
all
its
might,
and
let's
be
good
for
for
everything
that
the
then
it
can
you
yeah.
H
F
F
I
would
still
need
to
glue
those
together
in
the
sense
of
me.
I
could
use
your
AAS,
but,
as
my
RT
currently
needs
can
pre-configured
hard-coded,
Oscar
keys.
I
A
F
A
A
B
F
Just
a
question
about
the
lightweight
and
plan
setup:
can
life
a
tantrum
server
and
accepts
key
material
from
different
bootstrap
servers?
So
could
my
device
that
have
that
has
its
bootstrap
server
and
we
already
heard
that's
basically,
the
owner
of
the
device
go
to
some
wipe
and
m2m
server
and
say:
hey,
Here,
I
am
and
I
know.
You
won't
accept
my
credentials,
but
here's
my
bootstrap
server
ask
him
and
if
you
take
it
basically,
if
you
accept
that
bootstrap
server,
you
might
accept
my
credentials
as
well.
G
G
So,
basically,
all
the
devices
are
known
by
their
values
servers
and
it's
the
bootstrap
server
that
will
the
devices
when
the
first
boot
they
will
contact
the
bootstrap
server.
The
boot
shop
server
will
configure
them
with
the
information
on
how
to
connect
to
the
lightweight
and
through
the
servers
with
the
UI
and
the
credential.
And
then
the
devices
will
connect
to
the
lightweight
and
twin
server
will
register
to
them
using
a
kind
of
whistles
directory
format,
and
then
the
lightweight
m2m
server
will
perform
values
or
co-op
method
on
the
device
resources
to
configure
it.
So.
F
Basically,
that
is
saying
that,
at
the
time,
device
would
connect
to
a
lightweight
EM
tram
server.
Its
bootstraps
I
will
already
have
some
pre-existing
chat
with
that
server
because
it
configured
it
to
use
that
service
or
that
bootstrap
server
will,
in
some
out
of
scope
way
already
have
talked
to
the
actually
exactly
yeah.
F
A
I
A
A
E
It's
becoming
a
bit
more
generic
now,
so
what
you
mentioned
is
the
the
rest
case.
That
was
the
only
case
to
yesterday.
Now
it's
one
of
the
possible
cases
as
an
instance
of
a
template,
then
you
can
use
to
describe
more
generally
what
to
do,
even
in
terms
of
roles
on
resource
not
necessarily
identified
by
a
uri,
and
in
fact
we
are
creating
a
second
instance
to
cover
the
group
con
documents
in
ace.
A
A
We
have
person
discussion
today
we
a
theme
Alan
Solow.
It
was
presumably
were
a
bunch
of
us
and
a
meaning
in
language
and
to
him
as
far
as
I
also
understand,
there
is
no
actual
proxies,
it's
always
one
level
to
observer,
to
one
level
and
inclined,
and
vice
versa,
and
there
is
no
expectation
on
even
though
every
endpoint
has
co-opted
ID
and
server,
and
some
of
them
are
on
the
same
subnet.
There
is
no
expectation
of
direct
communication
between
two
levels
and.
F
I'm
not
so
much
talking
about
direct
communication,
so
the
way
the
way
lightweight
m2m
routes.
Communication
is
always
truly
live
right
after
observer,
which
is
which,
which
is
fine
but
thing
is
you
can't
express,
oh
so,
I'm
able
to
think
you
mentioned
that
there's
kind
of
this
IOT
side
and
then
there's
the
whole
back-end
and
there's
just
as
far
as
I
understand
currently
no
way
for
the
his
side
to
interact
with
other
things
on
the
IOT
side.
G
Currently,
the
light
with
him
to
him,
we
are
finishing
the
version
1.2,
but
the
enabler
then
do
it
will
be
a
requirement
phase
that
we
start
I,
guess
in
September,
together
the
requirements
from
from
the
values
company
that
are
involved
in
the
OMA
regarding
nitro
tantrum.
But
judging
from
the
past,
this
kind
of
requirement
that
one
IOT
device
would
like
to
access
resources
present
on
another.
The
IOT
device
authorized
by
the
right
went
onto
a
server
has
not
come
up
yet.
G
I,
don't
think
well,
I,
don't
know
if
will
be
part
of
next
version
of
light
within
twin
beds.
Currently,
no
one
has
expressed
this
kind
of
requirements.
As
I
said,
the
the
use
cases
are
various
centuries.
It's
really
a
device
management,
mainly
by
open
network
operators
on
of
the
devices
that
are
connected
to
their
network.
A
G
G
But
do
I
see
it
light-weighted
to
him
is
more
guttering
BIOS
standards,
mainly
IETF
RFC,
and
explaining
all
to
put
them
together
more
than
defining
its
own
processes
and
so
on.
So
everyone
at
the
OMA
is
very
keen
of
on
relying
of
on
publish
services
rather
than
reinventing
the
wheel
and
defining
new
stuff.
So
don't
know
if
if
the
need
arise,
the
width
of
that
it
will
be
the
domain
provider
for
the
future.
C
A
There
is
yeah
on
top
of
my
head.
There
was
this.
As
I
said
yesterday,
there
was
this
interesting
discussion
on
link
and
on
the
use
of
proxies
with
co-op
I
think
that
could
be
maybe
share
with
a
larger
audience,
but
the
main
participants
are
happy
with
the
conclusions,
so
it
might
be
that
we
have
another
ad
hoc
meeting
coming,
not
a
proper
entry,
so
we
just
you
know
just
follow
many
lists.
It
will
be
in
there
and
I
think
that's
all
I
have
Marco.
Please
feel
you
have
something
else
as
well.
I
just.