►
From YouTube: IETF-LPWAN-20220104-1500
Description
LPWAN meeting session at IETF
2022/01/04 1500
https://datatracker.ietf.org/meeting//proceedings/
B
D
D
D
What
happened
is
juan
carlos
published
or
two,
but
it
did
not
really
reply
to
you
until
now,
so
I
pinged
him
to
effectively
point
on
on
the
changes
that
he
made.
You
know
two
to
actually
answer
your
questions,
and
so
basically,
what
I
wanted
to
sort
out
today
is
if
we
were
all
happy
with
with
the
answers
and
the
version
in
o2,
so
we
could
effectively
call
for
publication.
A
B
D
D
A
Okay,
let
me
let
me
do
it
this
time
and
then
I
will
give
the
hand
to
you,
given
that
you've
prepared
the
hot
topics
afterwards,
so
hello,
everyone.
So
this
is
an
interim
meeting
of
the
lp1
working
group,
so
it
is
an
official
itf
meeting
and
as
such,
all
policies
of
the
itf
are,
in
effect
and
by
participating.
You
agree
to
follow
these
policies
and
processes,
so
you
should
read
all
the
bcps
and
all
the
documents
that
are
listed
here.
A
A
B
A
Move
no!
No!
It
was
a
real
inception
moment
like
like
okay.
It
works
so
yes,
anyway,
so
the
the
so
we'll
be
taking
minutes
and
please
participate
also
in
the
code
dmd.
So
sorry,
that's
the
notes,
itf.org
a
minute
taking
minute
taking
link.
So
you
have
the
link
just
here.
I'm
not
sure
if
you
can
click
it,
so
I'm
going
to
post
it
in
the
chat
at
this
time.
A
A
A
D
Actually
it
was
not
really
it's
just
that
there
was
no
line,
break
no
intermediate
lines
in
in
the
source
and
and
we
are
using
md
so
so,
apparently
that
that
that
was
the
prime,
the
transformation
from
ng
to
xml
to
rfc
lost
that
line,
inserting
a
blank
line,
just
restore
the
a
in
a
section.
So
I
did
that.
D
Then
there
was
the
review
for
the
of
the
data
model,
which
dominic
did
and
we'll
come
back
to
that,
and
then
there
was
the
compound
publication
process
which
we
hoped
we
could
start
before
christmas.
Actually,
we
could
not,
because
the
discussion
was
still
active,
but
that's
why
we
have
a
long
time
today
to
to
see
if
we
can
progress
on
that.
D
So
our
group
status-
we
we
have
the.
We
still
have
this,
this
question
of
whether
we
are
we
will
be
working
on
oem
and
that's
the
only
time
today
where
we
can
discuss.
You
know
where
we
go
with
how
I
am
so
dominique.
If
you
want
to
to
discuss
it
a
little
bit
today.
Just
now
tell
us
what
you
want
to
do
with
it.
E
D
D
D
D
D
So
I
understand
it
was
a
bit
late
and
I
don't
know
so
that
that
is
your
time
actually.
No,
I
have
another
slide,
but
basically
that's
that's
otto
that
was
published
and
the
data
model
was
published
at
the
end
of
november
as
well,
and
we'll
have
we'll
have
some
discussion
around
it,
because
there
was
an
interesting
thread
again
between
dominican
and
laurel
next
slide.
D
Okay
and
now
now
we
are
we're
we're
on
for
the
the
company,
so
so
the
the
different
questions
we
had
was.
One
of
them
was
the
the
data
model.
So
laura
you
removed
the
data
model
from
your
original
data
model
document,
so
there
is
no
no
trace
of
component
in
the
latest
database.
Correct,
yes,
and
so
normally
the
the
version
2
of
this
draft
of
copper
dac
should
have
the
data
model
piece
that
it
needs,
and
did
you
look
at
it?
G
G
D
H
D
E
D
H
D
A
D
B
G
No,
no,
no!
It's
good!
It's
good,
because
we
don't
need
to
modify
it
because
it's
something
that
is
systematic,
so
you
don't
have
to
say
you
disable
the
bitmap
compression,
because
it's
something
that
is
specified
in
component.
So
there
is
no
no
switch
between
one
behavior
and
the
other.
So
we
don't
have
to
add
something
in
the
data
model.
B
G
C
A
G
G
E
G
A
D
But
I
understood
that
alexander
was
telling
us
that
none
of
the
parameters
by
the
way
in
section
four
are
really
right.
Now
we
have
a
young
model
extension,
but
there
is
nothing
in
it
pretty
much
the
basically
whether
the
company
is
on
a
half
whether
we
compress
the
last
bit
map.
None
of
that.
None
of
that
is
indicated.
D
J
D
D
C
D
H
G
H
D
H
D
Okay,
besides
and
you
say,
besides
the
economic
behavior,
so
what
you
mean
by
besides
the
acanera
behavior,
you
mean
that
you're
augmenting
the
iconic
behavior
right.
I'm
not
sure
that
besides,
it's
actually
very
clear
that
it's
actually
that.
But
I
guess
that's
what
you
mean.
You
mean
achondra,
plus
death,
right,
yeah,.
H
C
D
Besides
the
archanera
behavior
and
it's
not
technically
telling
that
we
are
mounting
a
corner,
we
are
building
an
icon
or
a
documenting
it.
So
if,
if
that
sentence
could
be
slightly
rewarded
to
say
hey,
we
do
have
an
error
plus
what
we
specify
below
you
know
rewarding
like
that
would
be
a
little
bit
clearer.
E
H
D
Yes,
you
know
that's
when
it's
important
to
be
clear
when
you
extend
and
when
you
update
and
if
you
effectively
update,
then
you
you
need
in
the
documents
in
the
the
header
set
of
documents,
you
need
to
say,
update
8,
7
24,
and
so,
if
you
change
the
behavior
in
a
way
that
an
implementation
should
know,
but
I
don't
think
you
do,
then
you
need
to
say,
update
87
24.,
it's
very
touchy.
Eric
was
with
us
right
erica.
Are
you
still
there?
D
D
E
D
And
in
sections
3.2
you
need
to
basically
say
that
this
is
how
you
extend
or
update
depending
on.
If,
if
you
update
or
share
expert,
you
need
to
be
very
specific,
we
extend
it,
so
the
section
should
even
be
you
know,
extending
87,
24
or
something,
and
then
your
text
should
not
be
besides,
like
it
is
right
now
it
should
be.
J
D
D
H
How
you
use
this
act
is
a
bit
different,
but
in
the
other
sense,
after
the
old
one
or
the
old
zero,
you
will
be
receiving
the
compound
act.
So
I
think
that
it
will
be
an
update
in
this
sense
that
you
will
need
to
send
the
new
windows
and
it
will
extend
the
same
behavior
as
the
old
one
and
the
all
serial
messages.
Well,
in
this
case,
the
old
one
message
I'm
correct.
E
D
D
D
I
D
D
So
at
least
once
once
somebody
just
tells
you
something
you
can
refer
to
him
as
having
told
you
that,
but
I
think
it's
a
good
start
to
say
that
the
section
3.2
should
be
extending
should
be
called
extending
rfc
8724
and
then
in
the
text
itself.
You
should
not
say
besides,
like
an
error,
you
should
say
this
specification
extends
the
icon
error
behavior.
E
And
and
right,
and
so
that's
the
second
point
I
wanted
to
touch
upon,
which
is
to
me
the
the
delta,
was
when
I
asked
about
a
delta.
It
was
really
an
editorial
delta,
like
after
this
line,
insert
that
line
or
replace
that
paragraph.
With
that
paragraph,
the
way
it
is
written
right
now,
I'm
not
sure
what
the
final
text
is.
E
I
unders,
I
kind
of
guess
I
should
insert
new
paragraphs,
but
I'm
guessing.
I
should
also
remove
some
of
them
and
I
don't
know
where,
and
the
order
might
be
important,
so
I
was
really
expecting.
You
know
something
that
you
can
put
into
a
machine
to
edit
and
automatically
replace
text
in
8724
in
boom.
You
get
the
new
8724.
D
J
D
I
D
D
D
C
C
The
place
I
mean
I've
even
read
this
document
right,
but
at
some
point
of
time
you
need
to
explain
what
you
are
doing
in
section
two,
three
or
maybe
a
three
or
two
that
you
are
talking
about.
This
is
where
you
provide
a
diff
now,
if
the
diff,
it's
basically
whether
you
prefer
to
have
a
five
or
six
old
text,
new
text
or
text,
new
text
in
the
thing
which
makes
very
difficult
to
read-
or
you
completely
replace
a
section,
this
section
replace
section
blah
blah
of
this
rfc.
C
H
D
D
A
I
think
that
would
be
the
most
readable
way
to
do
it,
and
also
thanks
dominic
for
actually
pointing
to
this
point,
because
yes,
it's
the
original
rfc,
so
that
is
the
akon
air
mode
like
I
think
there
was
the
whole
automation
default,
finished
state
automation
that
was
included
into
that
into
that
consideration.
So
actually,
I
think
that
after
doing
that,
maybe
you
will
think
of
okay.
We
didn't
handle
that
wouldn't
handle
that
case
right.
So
it's
probably
going
to
you
know
it
will
help
a
lot
of
readability.
A
Of
the
rfc
8724,
so
there
is
no
mention
whatsoever
for
chic
compound
pack,
so
which
is
normal
right
because
it
didn't
exist
for
the
tanks,
so
should,
in
your
opinion,
should
there
be
like
mentioning
of
the
previous
parts
of
the
of
of
the
sheet
compound
ack
draft
that
actually
define?
What's
the
compound
ack
to
be
also
put
in
section
3.3
that
does
the
delta
or
or
maybe
there
is
section
3.3-
that's
like
generic
generics.
D
D
D
D
D
C
Yeah
no,
I
was
simply
repeating
that
it's
not
mandatory
that
the
chairs
are
the
document
shepherd.
It
would
be
really
good
to
get
somebody
else,
which
is
not
an
author,
of
course,
also
not
a
chair.
Maybe
a
call
on
the
mailing
list
would
be
nice,
except,
if
I
don't
know
anna,
for
instance,
wants
forward
because
there's
an
opportunity
to
learn
more
stuff.
C
I
D
D
So
please
go
through
whatever
ipa
you
publish
recently,
all
the
others
and
everybody
at
this
call,
and
if
you
have
something
related
to
this
document,
then
please
be
prepared
to
answer
about
that.
There
is
april,
and
with
this
we
move
to
the
next
topic,
which
is
the
one
group:
let's
go
for
the
chic
data
model.
G
D
G
Okay,
that's
good
okay,
so
I
will
skip
the
first
question.
So
just
tell
you
what
happened
so
before
the
vacations
dominic
sent
some
comments.
Then
we
have
some
discussion
offline
discussion
together
and
then
there
was
a
vacation
and
we
we
didn't
talk
about
much
during
the
vacation.
So
we
we
have
to
come
back
to
to
update
things.
G
So
I
will
go
only
to
one
question,
but
I
think
it's
very
important
to
to
discuss
here
the
other
we
can
do
offline
with
the
mailing
list,
and
so
I
think
one
more
important
question
is
about
interleaved
chic
fragment,
because
we
can
have
implementation,
for
example,
that
can
support
two
fragments
only
because
they
don't
have
enough
memory
for
three,
but
we
we
have
a
d
tag
that
has,
for
example,
two
bits.
So
it's
possibly
we
can
have
four
frag
for
packet
to
be
sent
at
the
same
time.
G
So
the
question
is:
what
do
we
do
for?
For
that?
Do
we
introduce
a
new
element
in
the
young
model
that
say:
okay,
theoretically,
you
can
send
four
packets,
but
practically
we
have
only
as
a
memory
for
two
because
it's
not
specified
in
the
in
the
rfcs
or
do
we
skip
it
and
say
the
implementation
has
to
to
do
the
maximum
each
time.
D
G
D
Do
you
think
there
are
other
parameters
like
that
which
would
make
sense,
because
you
have
to
think
about
not
the
rfc
but
the
implementation
yeah
at
some
point,
the
implementation
has
to
to
create
buffers
to
support
the
packets,
and
so
it
can
say.
Oh
I,
if
I
support
maximum
of
two
meaning
I
can
have
four
packets,
but
it
should
basically
check
if
it
can
do
what
the
profile
says.
So,
the
more
clear
you
are
about
how
many
packets
you,
your
only
buffers
you
need
to
to
have
the
best
for
the
implementation.
C
J
G
A
J
D
I
would
think
so
if
it
impacts
the
behavior
on
both
sides.
This
timer
impacts
the
behavior
on
both
sides
and
they
might
be
this
out
of
sync
if,
if
they
don't
have
the
same
value,
so
if
you
have
any
any
knowledge
that
needs
to
be
on
both
sides,
it
needs
to
be
on
the
data
model,
because,
even
if
it
doesn't
show
on
the
protocol,
it
impacts
the
way
the
two
sides
operate.
The
protocol.
I
D
A
D
B
D
C
E
E
C
J
G
A
A
E
J
J
D
D
D
Oh,
it
will
be,
I
mean
the
all.
This
will
continue,
so
I
hope
by
by
the
next
interim,
will
progress
on
on
those
issues.
So
we
we
have
something
to
discuss,
but
that
was
a
great
great
great
intern,
because
you
reprogressed
on
a
good
number
of
issues
for
those
documents,
I'm
very
happy
for
today,
just
before
we.