►
From YouTube: IETF-LPWAN-20220517-1400
Description
LPWAN meeting session at IETF
2022/05/17 1400
https://datatracker.ietf.org/meeting//proceedings/
B
A
B
A
Nice
to
see
you
guys
on
video,
so
carlos
said
that
he
would
not
be
joining
us
today
and
juan
carlos
said
that
it
will
be
a
stretch,
call
because
he
has
an
appointment.
A
A
E
Okay,
yes,
so
hello,
everyone,
so
this
is
an
official
itf
meeting
and
as
such,
all
the
itf
policies
are
in
effect.
So
please
read
the
note
well
and
all
the
bcps
that
are
related,
that
that
are
cited
here
so
by
participating
in
it,
you
agree
to
follow
all
itf
processes
and
policies,
and
if
you
are
aware
of
any
itf
contribution
which
is
covered
by
patents
or
patent
applications,
you
must
disclose
that
fact
or
not
participate
in
a
discussion
right.
E
E
Okay,
so
so
we
are
wasting
gigabytes
of
also.
We
are
not
wasting
gigabytes,
we
are
saving
the
knowledge
for
future
generations
and
everything
will
be
recorded
and
europe
and
the
presence
is
locked
and
julie.
Bird
is
pascal
tuber,
okay,
for
for
for
for
for
the
noting
for
the
well.
A
E
So
agenda
bashing,
so
for
today
we
have
several
presentations.
E
The
first
one
is
about
the
the
data
model,
so
the
the
engineering
data
model,
the
shepherding
the
reviews-
and
you
know
the
advancement-
then
a
short
presentation
of
the
compound
ack,
so
the
advancement
of
the
shepherding
and
the
chic
oversig
fox
shepherding
shepherding
so
quite
a
lot
on
these
advancements.
So
does
anyone
want
to
add
a
point
for
today's
presentation
or
for
today's
discussion.
E
E
Okay,
let
me
cut
the
camera
so,
as
I
was
saying,
the
first
two
points
will
be
discussed
later
during
the
call
and
the
working
group
will
ask
all
for
a
shift
over
sick
fox.
So
did
we
announce?
I
don't
remember
if
we
announced
the
the
end
of
the
working
group
last
call
pascal.
A
Yes,
we
did,
but
we
did
not
have
much
discussion
at
the
time,
so
I
kind
of
pulled
dominic
into
the
games
so
his
he
makes
his
his
review,
which
happened,
and
there
have
been
two
rounds
and
we'll
discuss.
A
Slides
later,
but
basically,
yes,
so
the
worker
plus
call
is
complete
for
both
compound,
which
is
waiting
for
your
ship,
complete
of
shepherd
right
up
and
the
young
data
model
and
we'll
see
that
we're
already
there
as
well,
very
much
very
close
to
press
the
button
and
then
probably
today,
we'll
we'll
kick
off
the
worker
plus
call
for
chicago.
It
will
depend
on
the
authors
if
they
feel
ready
for
for
the
plus
call,
then
we'll
probably
kick
it
off
today.
That's
the
intention.
A
Okay,
do
you
want
me
to
continue
and
the
only
the
only
real
new
document
that
we
got
is
the
ship
data
model,
which
is
now
nine,
so
there
are
two
iterations
since
the
last
call
and
it's
basically
the
two
steps
that
it
took
to
to
handle
the
linux
review.
A
And
so
let's,
let's
start
and
then
we
start
with
a
data
model.
That's
most
of
the
discussion
for
today
and
I'm
shepherd
for
that
one.
So
I
did
not
enter
all
the
steps,
but
so
there
are
a
number
of
issues
underneath
if
you
want
to
to
speak
about
your
review
and
where
we
are
and
if
you're
happy
with
the
resolution
in
the
mind.
A
That
that
was
quick
just
for
everybody
who
did
not
fully
follow
the
discussion
on
the
mailing
list.
There
were
a
number
of
items-
consistency
for
instance,
between
things
which
started
at
zero
and
things
which
started
at
one,
and
so
there
were
two
rounds
to
basically
agree
between
lauren,
dominic
and,
and
the
only
point
of
contention
was
the
way
we
express
the
timers,
and
so
we
have
a
special
slide
for
that.
So
we'll
come
back
to
the
tigers.
A
Next
slide
is
me
so
so
I
did
the
shepherd
review
it
was
just
published
for
all.
I
know
it's
not
answered
yet,
but
maybe
lauren
you.
You
did-
and
I
did
not
realize
so
in
my
review.
I
remembered
that
there
was
the
time
thing
so
that
time
thing
I
guess
is
is
fixed
and
then
there
were
a
number
of
needs
which
are
typical
from
a
shepherd
discussion.
A
So
if
you
didn't
see
it,
we
can
rapidly
go
through
yeah.
We
can
go
through,
but
I
think
so.
The
first
thing
is
that
there
is
this
rfc
7942,
which
says:
if
there
are
implementations
there
should
be
a
little
section
in
the
end,
where
you
say
are
the
implementations
and
since
there
are
implementations,
I
think
it's
good
to
know
for
the
isu
reviewers
that
these
things
are
effectively
implemented.
D
A
Okay,
so
that
I
mean,
if
you
can
have
some
words
on
that,
that
would
be
really
really
nice.
I
mean
that
will
help.
So
second
thing,
you
have
an
acknowledgement
section,
but
you
did
not
mention
your
young
doctor
and
since
you
know
right
now,
oh
you
fixed
it.
Okay.
The
other
thing
I
found
was
that
seriously
enough.
The
bcp14
text
appears
somewhere
in
section
seven
in
the
middle
of
the
yarn
module
and
since
you
start
using
uppercase
mast
in
section
two,
I
suggest
that
you
have
a
terminology
section
or
something
after
the
introduction.
A
Where
that
you
call
requirement
language,
it's
kind
of
the
way
we
do
it
now
and
with
the
usual
bcp14
text,
and
I
just
gave
you
an
example
of
how
that
was
done
in
one
rfc.
Just
if
you
should
not
have
the
that
text
nd,
I
just
gave
you
a
link,
which
is
three
one,
because
I
have
the
help
of
of
the
area
the
director
to
put
it
up
to
date.
A
So
the
then
there
are
id
needs.
You
know
we,
we
must
pass
all
the
needs
and
if
you
pass
the
needs
you
you'll
find.
There
are
five
instances
where
this
is
the
the
lines
are
too
long,
and
I
thought
you
said
you
were
using
some
lint,
but
but
when
we
were
in
ran
together,
you
basically
found
that
you
know
I
mean
in
the
middle
of
the
young
tours
of
the
atf.
There
is
one
which
will
ensure
there
are
not
too
many
lines
too
many
characters
per
line.
So.
C
D
It's
the
young
model,
we
have
description,
we
have
things
and
when
we
include
the
young
model
into
the
the
draft,
then
it's
very
tough
to
not
access
this
72
limit.
D
No,
but
it's
normally,
it
has
been
formatted.
D
The
most
complex
one
was
the
must
and
the
when
in
the
young
model,
because
it's
x-path
expression
and-
and
you
have
a
very
limited
room
to
to
write
it
and
make
it
clear.
But
I
try
to
to
do
it.
A
Okay
and
and
then
I
was
looking
at
the
references,
you
have
very
few
references.
You
only
have
two,
I
guess
on
on
the
the
shake
and
chico
co-op
specs
and
since
you're
defending
yank.
I
thought
at
least
the
first
time
you
used
the
dam
young.
You
should
put
an
informational
reference
to
er
nourishing
about
young.
I
mean
the
basic
young
thing
that
could
and
that's
that's
pretty
much
it
and
dominic
wants
to
speak.
Yes
dominique.
B
A
B
A
Okay,
so
that's
pretty
much
it
for
me,
so
lauren
do
you
sh?
Will
you
publish
a
10
with
accepting
the
tuple
requests
from
dominique
and
then
checking
those
items
they
have
here?
If
you
do,
then
I
will
press
the
button.
Okay.
A
Okay,
one
and
laura
will
give
you
the
ball
on
this
one,
to
explain
what
the
mantis
and
exponent
is
and
how
the
time
is
formatted.
But
this
is
basically
the
resolution
of
the
the
format
of
time
in
the
young
data
models.
So
maybe
you
can
explain
what
it
was
at
the
time
of
the
next
review
and
what
the
contention
was
and
then
what
the
solution
was.
D
Okay,
so
we
we
discussed
so
I
when
we
create
the
first
young
model
before
dominic
reviews
we
use
in
64
to
to
specify
the
uni,
the
timer
and
the
unit
was
in
seconds,
and
so
it
appears
that
we
may
have
some
very
reactive,
chic
implementation,
where
we
could
be
near
the
microsecond
and
we
can
have
very,
very
slow
chic
session.
D
So
what
we
propose
is
to
to
define
a
tick.
What
we
call
in
the
model,
tick,
duration
and
the
tick
duration-
is
2
power,
a
number
divided
by
10
power
6.
So
it
means
that
the
unit
is
microsecond
and
when
we
increment
this
number,
we
can
have
some
very
nice
parameter
like
the
one
with
20,
where
the
tick
is
about
one
second,
and
then
we
have
another
unit.
That
is
the
tick
numbers
and
we
say
a
timer
will
be
immediate
yourself.
A
D
And
so
here
there
is
a
small
python
program
in
the
repository
that
is
called
tick.pi
and
give
you
these
numbers.
So
the
first
column
is
the
value
for
exponential.
So,
and
the
second
number
is
the
step,
so
the
smallest
step
you
you
can
have
and
the
next
one
is
the
highest
value
you
can
have.
So
if
you
take
the
exponential
as
zero,
so
you
have
a
step
of
a
tick
duration
of
one
microsecond.
D
D
So
this
way
with
a
value
of
20,
we
cover
the
usual
range
of
chic,
but
we
have
some
way
to
either
to
re-reduce
the
timers
for
very
interactive
application
or
increase
the
timer
for
very
long
propagation
delays.
For
example,
if
we
take
25,
then
we
can
have
a
timer
of
25
days
and
if
we
take
35,
it
will
be
a
timer
of
71
year
maximum.
A
Yes,
and-
and
I
did
not
hear
it,
but
you
probably
want
to
say
that
if
the
value
that
somebody
uses
is
20,
then
it's
the
default
and
it's
implicit
right.
You
don't
have
to
pass
it
in
the
air.
D
B
Yes,
so
I'm
happy
with
the
this
algorithm
and
the
dynamic
and
resolution
it
provides
when
we
talked
about
last
friday,
I
forgot
to
ask
if
there's
any
comparable
mechanism,
that's
already
implemented
in
what
one
of
the
protocols
we
expect
to
have
on
board
our
devices
so
that
we
don't
add,
you
know
extra
code.
Is
there
anything
like
you
crop
or
maybe
some
other
product?
A
It's
close
to
what
triple
has,
but
I
think
what
we
so
so
in
the
in
the
dow
configuration
option
that
there
is
the
unit,
basically
it's
being
passed,
which
would
be
the
equivalent
of
the
expert
and
then
in
the
protocol
itself.
You
just
pass
the
timer
express
in
that
unit.
It's
just
that
by
memory.
The
unit
is
not
exponential,
but
I
can
check
it,
but
otherwise
it's
very
very
close
to
what
we
do
in
ripple.
D
D
D
And
the
other
point
I
we
have
also
discussion
with
dominic
and
there
was
a
confusion
in
position
because
position
was,
we
have
filled
position
that
vary
from
one
to
something
and
in
the
target
value
we
had
also
a
leaf.
That
was
called
position,
and
now
I
rename
it
as
indensia
index
here
or
something
like
this
and
this
here
and
this
way
we
will
avoid
the
confusion
with
two
things
called
position.
D
It's
a
synonym
of
in
index
from
the
world
dictionary,
and
I
I
think
index
is
to
is
not
so
so
clear
for
people.
A
A
So
can
I
say
that
the
bottom
line
is
that
lauren
you
publish
a
revision
with
the
poll
from
dominique
and
and
the
items
in
my
review
addressed
you
ping
me
and
I
pushed
the
publication
button.
I
already
sent
out
at
least
to
the
authors,
the
draft
of
the
shepard
review,
it's
also
available
in
the
data
tracker.
So
if
you
want
to
give
a
look
before
I
press
the
button,
please
do
so.
A
And
with
this
we'll
talk
about
the
compound
hack,
so
as
you
know,
I
was
the
original
shepherd
for
this,
so
I
wrote
the
the
original
chapel
write-up
using
the
ancient,
the
previous
format
as
it
goes,
the
esg
has
just
published
a
new
sharepoint
default,
write-up
format,
template
and
so,
as
alexander
takes
over
the
shepherding.
E
Should
have,
I
should
have
pressed
the
button
earlier
just
before
they
send
the
new
templates.
A
If
you
feel
really
really
quick,
they
won't
mind
too
much.
You
would
say:
oh,
it
was
ready.
You
know.
E
E
To
yeah
I'll
do
the
effort
to
put
it
in
the
new
format,
and
this
way
I'm
going
to
appropriate,
I
mean
it's
a
a
video
review,
yeah.
Yes,
so
I
I
already
do
it
in
based
on
on
you.
So
I
read
the
document
and
of
course
everything
is
so
there
are
no
issues
right.
So
I'm
going
to
to
finish
that
so
I'll
try
to
finish
that
by
the
end
by
so
outfit
I'll
do
that
by
next
by
our
next
interim,
okay.
A
So
it's
right
between
you-
and
I
I
mean
the
winner-
is
the
one
who
presses
the
button
the
fastest.
Oh,
let
me
win
that
one!
No,
I
need
you
yes
and
you
kind
of
know
how
to
win
the
race
and-
and
now
I
included
your
slides
in
in
the
generic
slide,
where
so
I'll
be
advancing
them.
So
before
we
go
into
a
nice
discussion,
I
would
like
to
see
if
the
authors
of
the
shakeover
sick,
fox
document
feel
ready
for
worker
plus
call
so
juan
carlos
is
not
there
but
search
you.
F
A
F
Have
modified
the
well,
we
are
doing
reordering
the
the
way
the
modes
are
classifying.
The
draft
are
described
and
we
change
the
headers
accordingly
to
what
we
saw
in
the
last
itf
meetings,
adding
the
crc.
Well,
not
the
rcrc,
but
the
resemblance
check
sequence
to
the
to
to
the
headers.
F
A
F
Yeah,
I
think
we
we
need
to
update.
We
need
to
release
a
new
version
of
the
updated
document,
yeah.
A
Well,
the
so
next
time
you
publish,
please
tell
us
on
the
mailing
list
that
you
feel
ready
this
time:
okay
and
then.
C
A
Yeah
I
mean
we'll
we'll
kick
after
the
work
of
glasgow,
and
so
now
the
the
floor
will
be
yours,
but
basically
sergio
your
shepherd
will
be
anna,
and
thank
you
enough
for
doing
this.
That's
very
kind.
A
G
So
for
for
this
version,
I
see
that
you
are
using
the
downlink
transmission
and
I
think
it's
correspond
to
the
appendix
f
of
the
rfc878724.
G
Isn't
it
so
perhaps
it
would
be
very
important
for
you
to
explain
very
well
how
this
downlink
transmission
works,
because
it's
not
in
the
is
in
the
appendix
of
the
rfc.
It's
not
really
the
the
the
text
and
in
the
second
thing
I
have
is,
you
are
using
the
tag
equals
zero.
So
you
are
not
using
the
tag.
G
F
Yeah
regarding
the
the
first
one
I
will,
I
will
check
out
the
pen
defects
if
we
have
to
take
any
consideration
required
in
the
the
quasi
direction
and
links
and
regarding
the
dtac
yeah.
We
have
some
internal
discussions
around
that
subject
and
we
are
going
to,
I
think
in
the
new
version
to
to
determine
how
the
packet
interleaving
is
going
to
be
performed
without
the
detect.
G
Okay,
so
perhaps
it
could
be
interesting
to
to
say
it,
because
if
you
need
to
inter-living
packets,
the
rfc
8724
use
dtag.
So
you
need
to
explain
how
you
are
going
to
use
to
do.
Interleaving,
packets.
F
G
G
F
The
the
comments
you
can
send
it
they
will
be
great
to
to
have
them
and
we
can
add
them
to
to
next
version.
Okay-
and
I
totally
agree
with
you
that
the
the
thing
about
packet
interleaving
is
not
clear
in
current
draft.
Actually,
it's
something
that
I've
been
also
reviewing
and
asking
how
this
is
going
to
be
performed
without
a
detect.
So
the
option
we
have
now
is
a
road
ids.
F
A
A
So
if
if,
if
you're
gonna
really
have
interleaved
packets,
then
a
d
tag
using
at
least
one
bit
of
detach
to
alternate
packets
in
the
air?
It's
very
useful.
D
Yes,
the
tag
is
just
a
convenience.
It
means
that
it's
just
after
the
id.
So
if
you
put
one
bit
it's
like
generating
two
rules,
what
is
in,
if
you
want
to
have
a,
for
example,
only
three
rules,
then
it's
better
to
generate
three
already.
If
you
want
four,
it's
better
to
use
the
tag,
it
depends
how
many
things
you
can
store
in
your
context,
but
it's
called
the
convenience
to
to
limit
the
number
of
words,
but
it's
not
mandatory.
A
A
F
A
Would
pass
the
comments?
Also,
there
is
one
big
thing:
it
allows
to
make
sure
that
you
you
stopped.
Even
if
you
don't
interleave,
you
know
that
the
pier
has
gone
to
the
next
packet
and
there's
about
on
this
packet.
So
when
for
you,
it's
it's
finished.
Well,
I'm
sorry!
If,
for
you,
it's
not
finished,
but
you
see
the
other
party
has
charged
the
d
tag.
A
A
Then
clearly
the
d
tag
of
one
bit
will
not
protect
you,
but
at
least
if
there
is
a
flip-flop
of
frabs
and
you
can
capture
packets
of
the
second
frame,
because
the
d-tag
has
changed.
You
know
you
know
that
the
previous
frame
was
finished
and
too
bad.
Whatever
you
received
is
lost
so
so
that
that's
kind
of
useful
but
as
as
you
said,
and
as
lauren
said,
you
could
also
use
different
ranges
of
varieties
due
to
the
same
effect.
B
A
A
Yes,
it
it
might
move
too
close
to
your
mouth
or
something
and
know
what
it's
doing,
but
it
scared
the
saturates.
B
B
B
But
in
this
year
it's
really
something
different.
Look
at
the
references,
it's
a
specific
mark.
Some
some
kind
of
you
know
clue
distinctive
distinctive
sign
of
something
not
nothing
related
to
a
position
within
the
collection.
B
B
G
I
want
just
to
ask
for
mbiot
for
the
draft
and
the
iot.
We
have
also
finished.
I
think
the
last
call.
So
I
will
I
want
to
to
to
send
the
last
version
today,
but
I
have
some
problems
with
my
geek
beat
up,
so
I
think
I
will
publish
it
as
soon
as
possible.
The
new
version
of
the
last
version
of
the
nbiot
draft.
A
Okay
and
then
so
I
was
I'm
wondering:
did
we
effectively
have
shepherd
for
for
nbiot
I'm
confused?
Now
I
don't
remember
that
we
have
a
championship
out
for
you.
Let
me
just
check
if
we
entered
one.
G
We
didn't,
we
didn't
attach
a
shipper
to
this
document.
We
only
do
the
last
call
after
the
iutf
and.
A
G
G
G
Right
until
my
next
publication-
because
there
are
some
changes.
A
Okay,
so
please
publish
will
remind
that
there
was
a
last
call
pending
and
we
didn't
get
much
review,
but
as
soon
as
you,
you
publish
okay,
so
the
book
and
effectively
I
don't
see-
oh
yes,
I
am.
I
will
be
your
shepherd.
I
actually
wrote
it
down
in
document,
so
it's
going
to
be
miana.
F
I
just
I
just
want
to
say
something
about
the
compound
draft
that
we
are
also
releasing
a
new
version.
This
version
I
already
sent
it
to
to
dominique
for
comments
on
the
because
I
added
a
follow-up
on
some
of
the
reviews
from
dominique
we
added
some
new
figures
and
we
changed
the
format.
So
well,
not
the
format,
but
we
changed
some
things,
so
the
padding
bits
are
undefined.
F
So
we
are
waiting
for
the
we're
not
waiting,
but
once
dominique
send
us
the
the
comments
we
will
be
publishing
a
new
version
also.
E
E
That's
a
good
news,
sergio
that
that
you're
working
on
it
I
mean
it's
good
to
improve
the
document
as
much
yeah.
F
A
E
Before
before
we
close
the
meeting-
and
there
was
something
that
got
published
by
the
by
the
laura
alliance
today
yesterday
about
the
about
the
ipv6
over
laura
one
release
of
the
technical
documents,
so
the
dolor
alliance
they
did.
I
need
to
look
at
the
link,
so
laura
alliance
did
work,
internal
work
and
basically
so
it
adopted
the
rsc
in
1911
in
the
internal
form
of
the
lore
alliance,
and
many
members
of
the
working
group
here
have
participated.
E
So
it's
a
very
clean
and
neat
document
and
you
may
you
might
want
to
see
the
press
release,
so
I
just
want
to
copy
copy
paste
the
link
and
I'm
looking
it
right
now.
It
is
actually
a
very,
very
interestingly
put
press
release
because
it
says
that
basically
I
mean
it
opens
new
verticals
and
you
know
all
the
nice
things
we
know
about
ip
well,
they
they
put
it
you
know.
E
For,
for
how
can
I
say
that,
in
a
way
that
is
appropriate
for
people
that
actually
don't
know
the
benefits
of
ip?
So
it's
actually
a
great
a
great
achievement
in
the
working
group.
E
C
E
Was
has
contributed
a
lot
to
that
document.
Thank
you
as
well.
A
That
dot
in
it-
and
that's
probably
why
so,
why
did
I
get
that
yeah
when
it
when
it
arrives
it?
Oh,
okay,
you
can't
copy
paste,
but
you
can
click
okay,
okay,
so
I
have
it
neat.
Thank
you
so
much
I
mean
that's,
that's
a
huge
achievement,
like
you
say,
a
coronation
for
for
the
work
we've
done
over
the
years.
A
E
Oh
yes,
thank
you.
Thank
you.
Yes,
so
there
is
going
to
be
a
webinar
organized
by
the
lore
alliance.
The
webinar
is
on
june
14th,
thanks
for
the
reminder,
actually
dominic
and
let
me
put
the
link.
E
Let
me
put
the
link
here
and
take
so
switching
between
navigators,
so
it's
a
little
bit
time-consuming.
Oh
thank
you!
Thank
you
dominic.
So
I
opened
the
link,
so
you
can
you
can
subscribe
to
the
webinar
and
and-
and
you
can
see
it
so
really-
a
very
significant
work
getting
done
and
the
work
that
that
was
that
was
done
in
the
in
this
working
group.
Right
now
becomes
a
major
technology
of
the
laura
one
world,
but
also
in
others,
so
so
yeah.
E
C
A
A
Well,
we
we
had
time
so
that's
cool
okay.
So
so
again,
congratulations
guys
and
thanks
so
much
for
for
all
the
work
which
is
now
you
know
being
integrated
in
the
standards.
So
that's
kind
of
very
cool,
okay
and
then
with
this
I
guess
we
will
conclude
this
this
meeting
for
today.
A
So
we
we
we
meet
next
in
two
weeks
and
so
the
to-do's
we
have.
If
I
can
summarize
that
laura
you
have
to
publish
a
new
version,
everybody
has
to
publish
a
new
version,
basically
for
the
young
model,
so
that
you
incorporate
the
the
pull
request
from
dominique
and
my
review
sergio
you,
you
said
you're
working
on
the
the
shake
of
a
sick,
fox
document,
so
you
need
to
push
a
new
version,
so
I
can
start
doing
the
last
call
for
the
mb
iot.
The
last
call
is
finished.
A
We
didn't
get
much
review
so
and
I
you
will
publish
a
new
version
and
they
will
ask
additional
reviews
on
it
to
kind
of
get
something
from
the
workers
call
and
the
compound
hack.
I
understand
sergio
you're
doing
a
few
changes
and
when
the
changes
are
done,
then
basically
the
we
can
quickly
review
them.
And
you
know
the
chairs
are
the
shepherd
alex
and
then
alex
will
press
the
button
for
submission.
A
That's
pretty
much
it.
So
all
those
documents
are
getting
shut,
so
the
next
will
be
probably
the
architecture
and
then,
if
we
need
another
document
to
express
how
we
can
provision
the
network
with
the
data
model,
so
so
the
flows
for
security
and
configuration
provisioning
and
data
model.
So
it's
going
to
be
something
and
there
is
always
spending
the
work
which
started
on
icmp
and
and
ping
and
all
those
things-
oh,
I
am
so
so
dominique
next
time.
A
A
I
missed
one
to
do
for
again
dominique,
which
was
a
review
that
the
changes
that
are
made-
I
guess
in
the
combo
deck,
oh
okay,
right
before
you
ship
the
company
sergio
and
this
time
it's
probably
it.
Okay.
So,
thank
you
all
and
talk
to
you
see
you
on
the
mailing
list
and
in
this
intro
in
two
weeks,
take
care,
see.