►
From YouTube: IETF106-6TISCH-20191122-1000
Description
6TISCH meeting session at IETF106
2019/11/22 1000
https://datatracker.ietf.org/meeting/106/proceedings/
A
To
give
a
chance
for
people
to
arrive
with
a
bit
of
latency,
and
with
this
we'll
be
looking
for
minutes
takers,
is
there
a
candidate
to
help
us
take
some
minutes?
The
etherpad
is
indicated
on
this
page.
It's
also
indicating
the
next
page.
It
would
be
really
cool
that
you
guys
can
attend
the
etherpad
and
takes
some
notes
in
particular
you're
going
to
talk
to
the
mic.
Then
please
make
sure
that
the
minutes
capture
correctly,
what
you're
going
to
say
and
with
this
so
do
we
have
a
summary,
a
ping
on
the
minutes,
nice.
A
Okay,
you
can
okay,
thank
you
meeting,
so
this
is
60s.
This
is
a
usual
ITF
meeting,
so
we
go
by
all
the
rules
of
the
ATF
and
there's
a
list
of
best
practices,
but
processes
or
response
at
cetera
and
in
particular,
we've
got
the
best
practices
applying
to
IPR.
So
if
you're
aware
of
IPR
on
anything,
that's
going
to
be
discussed
today,
please
tell
the
room
or
let
the
shares
know
by
at
the
end
of
the
meeting
with
this.
Remember
you
are
being
filmed.
The
movie
will
be
played
on
YouTube.
A
Thousands
and
hundreds
of
people
will
vision,
visualize
it
so
be
very
smart
and
so
will
publish
the
minutes
and
the
recording
you
need
to
sign
the
blue
sheets.
We
are
kind
of
going
to
pass
them
along
in
a
few
minutes.
Please
go
online
on
the
ether
pad
and
contribute
two
minutes
taking
and
validating
the
minutes
and
you'll
find
the
usual
links
about
the
60s
activities
on
this
page,
and
with
this
we
can
start
the
agenda
bashing.
A
So,
basically
we're
going
to
tell
you
where
we
are
with
this
working
group:
our
progress,
the
status
of
documents
and
related
documents,
not
a
working
row,
so
it
goes
to
the
milestones
and
then
we'll
decide
what
we
do
next
well,
we'll
have
ten
minutes
on
the
status
of
the
architecture
then
go
to
minimal
security,
that's
gonna,
be
Manisha,
then
we'll
have
Michael
if
he
joins,
he
sent
us.
Some
slides
is
not
there
yet
and
we'll
have
Jay
go
remotely
so
Jay
go.
A
A
Could
you
set
up
your
mythical
correctly
thanks,
very
well,
I
hope
it
works,
and
maybe
the
most
unusual
and
final
discussion
we'll
have
today
is
what's
next
because,
as
we
will
see,
we
have
pretty
much
concluded
all
the
work
items
for
which
we
were
chartered,
which
kind
of
means
that
either
we
reach
out
or
we
just
sell
out.
So
that's
going
to
be
the
discussion.
A
So
if
you
have
items
that
you
want
to
discuss
in
this
working
group
in
the
future,
then
it's
it's
a
it's
a
good
time
to
speak
now,
okay,
and
with
this
a
little
bit
of
status
for
the
twelve
on
the
other
working
groups.
So,
as
you
know,
six
teachers
spent
more
time
actually
pushing
work.
Another
working
group
that
doing
individual
work
inside
this
working
group.
We
we
mostly
worked
around
an
architecture
and
open
source
implementations
and
interrupt
tests
and
to
make
all
those
things
happen.
A
Then
we
had
a
number
of
dependencies
and
a
number
of
working
groups.
So
there
is
a
list
of
documents
that
we
actually
promoted
in
other
working
groups.
So
in
6lo
we've
got
all
the
fragmentation
work.
So
there
are
two
documents:
they
are
both
passed
well
groups
last
call
and
we
started
getting
feedback
from
the
affero
on
the
minimum
fragment.
So
there
was
a
new
recipient
of
the
minimum
fragment,
so
those
two
draft
are
basically
on
in
a
good
shape
for
to
go
through
iesg
and
they
are
in
suresh
carrying
health.
A
We
also
promote
you
to
work
on
6lowpan
neighbor
discovery.
So
that's
one
of
them
is
now
RFC,
eight
five,
four
five
and
then
we've
got
the
address
protection
in
the
backbone
router,
which
is
an
T
proxy.
So
also
all
this
work
is
in
good
shape
and
and
continuing
through
the
is
G.
We
have
this
new
draft
that
was
proposed
to
completely
get
rid
of
multicast
enable
discovery
using
software
registrar's.
This
is
new
job
submission.
It's
it's
related
to
the
other
three.
A
So
that's
why
it's
listed
here,
but
I
don't
know,
what's
going
to
happen
with
it,
it's
not
needed
in
the
sixty's
architecture,
it's
beyond
six
dish,
it's
on
the
backbone
and
then
we've
got
the
dead
lifetime,
which
is
also
going
through
ESG
at
four
we've
been
promoting
the
Sdn
model,
and
that's
where
the
Dow
projection
goes.
There's
also
this
work
on
a
leaf
which
would
not
be
aware
of
ripple
so
using
only
eight
five,
four
five
to
signal
the
presence
of
the
leaf
to
the
repo
rotting
and
mapping
that
into
repo.
So
that's
going
well.
A
Both
documents
have
progressed.
I
would
say
immensely
about
the
since
the
last
IETF.
There
were
several
rounds
of
each
of
those
documents
and
roll
is
planning
to
push
quite
heavily
on
the
unaware
leaf,
because
many
documents
are
actually
dependent
on
that
one,
including
the
sixties
architecture,
by
the
way
and
enrollment
priority
I
think
the
the
chairs
have
been
asking
for
reviews
on
that,
one
to
get
through
the
age
what
to
do
through
the
work
of
Wisconsin
and
then
we've
get
an
email
with
the
constraint,
voucher
and
I
hope
that
Michael
could
tell
us
about
it.
A
A
So
here
is
the
milestones
that
we
had
I
mean
two
of
them
Suresh
still
waiting
for
accept,
but
I
mean
considering
that
it's
just
about
submitting
the
draft
to
ASG
and
that
the
draft
have
actually
been
submitted
to
Zermatt.
A
So
the
the
dt
securities
you're
attached
is
mostly
empty
of
its
contents.
We
kept
it
just
to
keep
the
working
group
informed
about
the
progress
and
virtual
work,
so
I
guess
it
will
never
be
submitted
to
ASG
or
it
will
be
moved
to
animal
or
place
that
will
take
care
of
it,
and
so
there
we
go.
The
only
item
that's
left
is
evaluate
the
work
of
progress
and
talk
about
which
are
trying
or
completing
the
work.
A
The
60s
architecture
may
be
viewed
as
maybe
the
most
encompassing
work
at
six
dish,
because
that's
really
what
shows
what
we
wanted
to
achieve
in
this
working
group
and
why
we
pushed
so
many
work
items
in
different
working
groups
and
and
the
overall
picture
that
we
wanted
to
build.
So
there
are
a
number
of
components
that
we
are
done
at
home,
unlike
the
minimal
security,
MSF,
etc,
a
six
six
P
protocol
and
then
a
good
number
of
work
items
which
actually
done
outside
the
working
group.
A
A
So
there
are
really
two
pieces
in
one
document
and
we
could
have
done
to
document
some
people
said
we
should,
but
we
have
a
very
high-level
architecture
that
goes
into
the
domain
components
and
articulates
them,
and
then
we've
got
a
deeper
dive
into
some
components
which
are
most
specific,
six
and
Tom
being
explained.
For
instance,
there
is
this
concept
of
a
track.
That
is
now
inherited
by
the
work
at
Rome,
so
we
actually.
A
Well,
we
went
through
the
ASU
review
so
that
we
had
for
review
since
the
last
IETF
that
Eric
maja
Roman
and
in
German
with
Benjamin.
Obviously
we
we
went
through
a
deep
review
of
the
security
section,
but
not
only
Benjamin
has
been
going
very
slowly
through
the
whole
document
and
thanks
to
Manish
Shah,
and
you
actually
thank
you.
We
actually
did
a
lot
of
work
on
the
security
section
and
expanded
quite
a
bit.
A
So
that's
that's
one
of
the
big
changes.
We
also
extended
the
discussion
of
the
various
forms
of
diversity
that
are
viable
ratios,
which
are
not
necessarily
the
pre
off
that
of
and
in
that
net.
So
we
would
clarify
that
we
expanded
on
the
joint
process
as
well,
and
the
minimal
thing
and
I
insisted
that
we
are
well
and
not
annamaria,
but
that
we
are
too
many
references
and
they
were
not
really
well
classified
into
normative
and
non
normative,
with
the
understanding
that
this
is
not
a
Star
Trek
document.
A
This
is
a
an
informational
track,
but
there
is
still
a
meaning
for
normative
and
that
meaning
is.
If
you
need
to
understand
the
reference
to
understand
this
document,
then
the
reference
is
normative
right.
So
it's
not
normalization.
If
you
like
it,
it's
really
like.
You
can't
understand
this
document
without
understanding
the
other.
If
that's
the
case,
then
the
reference
is
normative.
Otherwise,
just
you
can
avoid
reading
it.
It's
informative
just
for
your
own
reading,
if
you
like,
so
we
actually
had
to
move
and
shuffle
around
the
references
to
clarify.
What's
normative.
A
The
problem
with
that
is
by
putting
more
documents
into
normative.
We
also
created
a
good
number
of
miss
Roth,
miss
Roffe
being
those
documents
which
are
normative
reference
to
euros,
but
are
not
yet
roc.
So
you
cannot
publish
the
architecture
before
all
of
the
mistress
are
also
published
as
RFC,
and
there
are
now
like
six
or
seven
miss
rafts,
including
the
raw
document
that
I
showed,
including
minimum
security,
including
MSF,
etc.
A
The
architecture
is
really
the
document
that
will
close
the
door
of
all
the
work
that
has
been
done
at
6:00
when
this
is
either
IFC,
then
all
the
work
we've
done
is
there
as
well,
and
that
also
makes
sure
that
the
architecture
is
completely
compatible
and
represents
well
the
work
which
is
done
in
the
other
document,
even
if
they
change
will
have
to
change
the
architecture
to
represent
that
correctly
and
that's
pretty
much
I
had
to
say
about
the
architecture.
So
we
have
any
question
okay.
A
So
so,
if
you're
interested
in
the
work
that
we
try
to
start
with
row
by
W,
reliable
variable
Wireless,
you
can
see
a
row
as
being
the
next
phase,
four
six
dish,
but
focusing
on
the
writing
aspect
of
it.
That's
why
it's
it's
done
in
the
writing
area
and
most
besides
precisely
on
the
operation
of
a
sixties
track
now.
Row
is
not
specific
to
sixty
shot
or
it
seemed
to
be
abstract
to
be
completely
lay
off
free
operation.
A
But
it's
near
it's
certainly
from
the
the
designs
that
we've
done
for
tracks
and
all
this
capability
to
have
this
direct
acyclic
graph
to
a
certain
destination
and
make
intelligent
forwarding
decision
there
on
this
track.
So
row
we'll
be
focusing
on
that.
If
the
group
forms
and
that's
pretty
much
it
for
now-
Thank
You
Posca.
B
Okay,
so
I
will
talk
about
the
minimum
security
document,
which
was
renamed
to
be
the
constraint
drawing
protocol
for
six
days
after
one
is
three
review,
the
current
statuses.
The
document
is
in
third,
it's
in
thirteenth
version
that
the
IRG
reviews
have
been
received,
as
well
as
up
there
and
sec,
their
reviews
and
general.
B
So
after
the
last
meeting,
we
discussed
some
text
prohibiting
the
mixing
of
different
levels
of
what
entity
authentication
tags.
This
was
incorporated
in
twelve.
We
added
a
new
subsection
is
an
ASN
replay
attack
after
the
last
meeting
discussions,
and
we
also
added
some
mandatory
support
for
extended
tokens
at
JRC,
as
was
received
a
comment
offline
from
Klaus
Harkin
from
core.
B
We
received
obviously
two
to
two
reviews,
be
a
pre
I,
usually
use
one
of
their
and
one
secretary
review.
There
were
mostly
needs
one
concerned,
one
added,
a
new
requirement
that
concerns
three
commissioning
of
a
device
where
we
added
the
requirement
that
the
pressure
key
must
be
changed
in
case
the
device
is
recommission
and
during
the
Secretary
will.
These
were
most
units
that
we
received
I,
usually
reviews.
We
got
four
discusses
and
three
no
objections,
and
one
year
setting
from
Suresh.
B
B
So
that
there
was
a
couple
of
parallel
discussions
on
it,
essentially
we
in
during
the
drawing
phase
we
it
was
deemed
okay,
that
we
can
use
/j
as
long
as
we
use
sixty
sharp
ax
as
the
well-known
host
name.
The
problem
was
the
after
the
join
exchanged.
The
parameter
updates
part
of
the
protocol,
where
sixty
sharper
was
missing,
well-known,
Yura
host.
So
what
we
did,
what
we
proposed
is
the
resolution
is
to
use
sixty
sharper,
also
in
that
case
as
well,
so
that
it
is
isolated
from
all
other
uses
of
slash
stray
yeah.
C
B
B
C
B
One
comment
that
would
receive
from
media
in
regards
the
parameter,
update
response
message
which,
which
is
as
it
now
stands,
empty
the
parameter.
This
is
response
to
the
parameter,
update
message,
which
is
a
cop
confirmable
message,
and
essentially
we
define
for
well
for
generality,
essentially
another
message
that
is
a
response,
but
that
is
essentially
empty.
That
is
piggybacked
over
over
coop
over
coops
confirmable
response,
and
the
question
raised
was
is
completely
legitimate
was
like.
B
Why
define
our
own
our
own
protocol
message
when
the
response
is
confirmable
and
that
when
so
that
we
have
an
indication
of
reception
at
the
sender
of
the
parameter,
update
message?
So
I
mean
why
keep
it
essentially
and
the
proposes
resolution
is
essentially
to
remove
the
parameter,
update
response
message
from
the
protocol.
Yeah.
B
The
on
the
list-
yes
yeah,
yeah,
it's
it
has
been
discussed
on
the
list
and
it
was
incorporated
in
the
document,
so
I
believe
we
are
good
to
go
there
here.
I.
Essentially,
the
issue
that
I'm
going
through
are
the
issues
that
haven't
have
had
their
resolutions
in
the
in
the
working
copy
of
the
document.
C
B
B
He
says
I
think
we
may
need
to
say
more
about
how
JP
knows
the
SEC
isn't
is
in
effect,
since
that
affects
the
critical
piece
of
the
security
posture
of
the
network
and
this
true
that
in
the
current
text
we
don't
have
the
clear
link
between
house
Exantus
set
other
than
it
was
considered
out
of
scope.
So
what
I
propose
here
we
had.
We
have
this
join
rate,
parameter
that
can
be
used
by
the
Dre
RC
to
enable
any
particular
node
in
the
network
to
act
as
a
joint
proxy,
so
I
I.
D
There
are
given
in
psychic
seemed:
is
it's
not
per
join
proxy?
It's
in
in
join
proxy
for
each
joining
node?
So
so
you
can't
set
it
until
you
know
who
is
going
to
be
joining
okay,
but
you
need
a
policy
to
yet.
You
read
the
policy
that
you
are
allowing
you
know
join,
but
the
first
complexities,
of
course,
coming
out
of
the
blue.
You
don't
know
who
do.
D
So
a
first
message
to
get
in
will
always
get
error
because
you
are
not
allowing
it,
but
then
it
up
layer.
You
can
see
that
okay
check
out
the
policy
if
they
you
know
join
generate
generate,
is
non
zero.
Then
we
allowed
configuration
except
for
that
know
that
I'm
just
sending
a
message,
but
but
it's
it's-
it's
not
like
for
forty
snow.
This
actually
for
remote
draw
that
remote
node
is
allowed
to
exempted
security
management
in
Pakistan,
okay,
okay,.
D
And
I
think
the
one
the
one
of
the
problem
is
that
when
we,
of
course
after
he
had
finished
joining,
we
want
to
drop
the
SEC
X,
and
so
he
we.
He
of
course
doesn't
send
any
anymore
pockets
in
clear.
But
some
you
know
attacker
mic'd.
You
know
send
packets
in
clear
claiming
to
be
him
because
and
actually
do
some
nasty
said,
that's
why
you
want
to
remove
the
psychics
and
so
it
that
it
take
it
dropped
automatically
by
the
Mac.
F
D
We
have
yes,
that's.
Actually.
One
of
the
comments
I
had
earlier
was
that
it
actually
would
be
better
to
for
JP
to
know
when
the
process
is
finished.
I
like
currently-
and
it's
also
you
know-
because
one
of
the
problem
there
is
that
when
we
sent
the
final
response
back,
that's
actually
sent
in
clear
because
we
don't
have
that
one
has
a
hundred
keys.
Yet
the
problem
is
that
the
AK
is
coming
back
with
using
the
same
security
level.
That
was,
you
know,
be
sending
so
and
we
have
to
respond
to.
D
We
have
to
be
able
to
receive
that
ACK,
which
means
they
still
have
to
have
an
SEC
accept
own.
At
that
point,
it's
not
transmitting
that,
and
of
course
he
if
he
doesn't
get
it,
we
have
to
retransmit,
and
so
so
there's
it
did
either
be
the
timeout
or
some
you
know
actually,
first
time
we
see
him
sending
us
any
encrypted
message.
We
know
that
he
has,
you
know
received,
he
has
the
keys
and
we
can
drop
the
SEC
extent.
That's
actually
it's
one
optional
or
or
I,
don't
know
what
this
could
could.
B
D
B
Then
there
was
an
issue
also
raised
by
Bank
a
duck
on
the
use
of
quadric
max
enjoin
attempts
parameter.
That
appears
only
twice
in
the
text,
and
this
is
essentially
a
remnant
from
the
time
join
request
was
a
non
confirmable
message.
So,
as
now,
we
rely
on
quarter
de
clave
declared
failure
to
the
application.
We
simply
can
remove
this
and
remove
it
from
the
text.
This.
This
is
the
most
simple
resolution.
B
Another
common
from
Bank
ADA
considers
the
parameter
additional
info
that
was
deemed
under
specified
because
in
one
there
is
one
example
given
in
the
text
that,
during
the
processing
of
the
error
message
in
this
parameter,
we
can
include
a
subset
of
keys
that
were
not
we
that
were
not
being
that
could
not
be
processed.
So
the
idea
was
the
Tenakee
compliant
with
link
layer.
Keystroke
that
is
defined
in
the
document
can
be
included.
But
apparently
there
is
more
text
needed
here
to
to
make
this
to
not
make
it
under
specified.
B
One
other
issue
that
was
raised
by
bank
attack
is
the
the
missing
of
the
label
of
the
elyda
t
for
each
message.
He
says
that
he
feels
that
it's
little
unusual
to
have
a
consolidate
registry
for
Quadra
parametres
that
are
used
as
map
labels
across
different
messages,
without
some
indication
of
which
map
labels
are
valid,
in
which
messages.
So
essentially,
the
resolution
to
this
should
be
to
add
a
paragraph
rate
arising
and
summarizing
c
d,
dl
c
fragments
which
already
indicate
a
bit
what
what
parameters
are
valid
in
which
quadruped
objects.
A
A
A
G
F
F
Go
okay,
then
I
would
I
think
moaning
over
there.
Okay
she's
the
mid-may
over
here.
This
is
Diego
harmony,
I'm,
going
to
present
the
the
information
element,
the
calculation
for
six
dish
and
join
our
enrollment
information,
which
is
rational,
zero,
zero.
Six
and
next,
please
be
brief,
I
hope.
So,
what's
the
status
of
the
of
this
draft,
we
have
already
passed
individual
DLC
in
case
we
have
the
last
call.
F
You
have
a
one
revision
from
Carlos
Gomez
waiter
really
thank
and
we
are
going
for
the
version,
zero,
seven
from
shore
revision
for
judges,
and
we
noticed
some
few
typos
and
I
covered
on
the
interpretation
of
the
a
of
the
link
Oh
case.
But
this
checklist
it's
a
simple.
It
doesn't
change
the
core
of
their
of
the
draft
and
they
do
I.
Also
thank
fear
of
about
the
the
enlightenment
or
how
to
manage
date.
The
name
of
the
hundred
fifteen
foreigner
naming
now
and
remember
three
sixteenths.
He
mentioned
this
and
for
the
future
next
slide.
F
Whatever
we
have
changed
here,
it's
a
few
filtration
titles,
okay
and
subsection,
1
1
pitch
1.3,
it's
the
usage
and
meaning
of
a
rotary
advertisement.
Association,
we
should
be
cleared
so
we
clear
it
up.
Okay
and
we
close
the
the
expressions
at
the
closing
of
section
1.
We
have
improved
them
and
some
feel
description
or
the
change
also
when
on
stage
section,
2
de
to
join,
how
to
arrange
them
in
the
order
word.
They
appear
on
the
hospital.
F
Then
the
cigar
decoration
and
we
can
lead
locally
glare,
can
be
using
description
of
the
pivot,
which
was
in
Austria,
each
other
should
have
been,
and
fire
finally
has
a
small
citation
explicitly
for
the
area
of
fc8
81-57
if
the
subtype
next,
please
just
very
few
questions.
Finally,
we
are
going
to
move
to
0-7,
okay.
After
the
all
the
comments
we
have
received
and
then
I
expect,
we
hope
this
draft
will
go
through
the
ilg
at
the
next
step
and
that's
it
for
for
this
draft
happy.
F
F
A
A
E
E
A
A
A
E
So,
since
the
last
re
camp
meeting,
we
have
three
main
issues
raised
up.
The
first
rate
is
the
rules
for
salaries
which
is
pointing
out
to
Pascal
and
because
of
the
curve,
its
version,
the
MSF
doesn't
specify
where
the
Souza's
pick
it
up,
though,
when
generating
the
6pu
request,
and
so
we
should
have
to
know
that
maintaining
a
candidate
is
helpful
for
this
at
least
and
know
the
world
keep
listening
for
the
clearance
of
the
cells
and
to
check
whether
is
a
collision
sound
ourselves.
E
And
then,
when
generate
the
6p
request
package,
it
will
pick
up
the
cells
from
this
cell
phone.
So
the
another
issues
is
the
schedule.
Consistency
pointed
out
by
Christine,
and
the
issue
is
when
the
mode
is
despaired
from
the
network,
such
as
when
the
power
of
the
south
is
reserved,
his
parents
will
live
in
the
parent
schedule
of
ever
without
knowing
the
neighbor
is
gone
that
we
used
to
have
the
keepalive
mechanism
for
each
node.
It
will
stand
some
Q
line
message
to
to
the
neighbor.
E
It
has
a
sales
and
half
before
and
if
this
icky
blob
is
not
act
by
his
neighbor
and
the
neighbor
will
be
Claire
declared
as
unreachable
and
and
then
it
will
stand
a
bit
clearer
to
the
neighbor,
it
may
be
filled,
but
it
will
remove
the
cell
cell
to
the
neighbor
from
its
schedule,
so
we
used
to
have
it
in
the
previous
version,
like
0-2
version
of
the
master
and
we
removed
out
from
those
trees,
because
at
the
time
we
don't
have
the
dynamic
scheduling
for
the
autonomous
house.
At
that
time.
E
The
Ottoman
cell
wall
installed
apparently
in
the
schedule,
and
in
that
sense
they
may
because
they
for
install
one
cell
for
our
neighbors.
But
now
the
MSF
has
dynamic
scheduling,
so
those
cells
could
be
removed.
The
only
installer
wouldn't
necessary.
So
so
we
bring
it
back
in
the
latest
MSF
job
and
the
third
one
is
we
added
the
downstream
traffic
adaptation
feature
for
the
latest
MSF,
and
this
is
the
main
thing
I'm
going
to
explain
in
details
and
next
one.
Please.
E
So,
firstly,
is
the
strategy
of
a
downstream
traffic
interpretation
is
pretty
similar
to
the
upstream
one
and
has
two
counters
and
the
number
cell
use
themselves
elapsed
and
them
cell
use
the
world
blood
plus.
When
the
frame
is
received
from
its
parents
on
a
our
Excel
and
themselves,
the
elapsed
number
will
increase
me
will
increments
to
win
the
current
Salazar
Excel
to
his
parents.
The
different
is
the
upstream
for
upstream
traffic.
It
has
a
negotiated
txl
at
the
initial,
but
initially
there's
no
negotiating
or
excel
for
each
node.
E
So
we
decided
to
counting
the
auto
or
excels
at
the
beginning
to
start
to
this
downstream
traffic
adaptation,
so
to
use
them
that
that
there
is
a
tool.
There's
one
problem
is
during
network
a
forming
face.
Those
cells
are
used
actually
by
its
children
and
these
parents
at
the
same
time,
and
this
CUDA
caused
a
collision
on
these
cells
and
which,
when
the
massacre
is
detecting,
the
usage
of
the
downstream
traffic,
is
actually
will
detect
a
lower
value
for
this
cell
usage.
E
But
after
the
networks
formed
and
the
traffic
from
his
parents
will
transfer
to
the
negotiate
eh
cells.
So
the
man-made
traffic
is
coming
from
his
parents,
which
is
a
downstream
traffic
on
this
auto
or
Excel,
which
is
ideal
for
a
downstream
traffic
adaptation,
so
for
the
for
the
reason
that
mostly
down
to
traffic
are
generated
after
networker
is
formed,
so
we
can
was
a
SAP
to
this
this.
A
Think
they
just
one
meter
pointer,
is
basically
when
the
network
is
being
formed,
I
mean
what
you
see.
There
is
mostly
control
in
that
data
right,
because
network
is
the
diet,
and
so
those
collisions
are
a
lot
related
to
these
and
Diu
messages,
and
we
are
working
right
now,
after
all,
to
minimize
the
chances
of
collisions
like
this
by
intruding
introducing
some
some
trickle
mechanisms
in
the
das,
so
the
prime
should
actually
also
be
addressed
at
the
operator
at
the
repo
layer
so
as
to
to
avoid
most
of
those
collisions.
A
E
For
the
information,
so
we
also
implement
the
downstream
traffic
we
implemented.
Japan
I
was
opened
up
with
an
invitation
and
will
create
a
pure
as
especially
for
this
feature
and
currently
had
marched
and
we
tested
it
through
a
test
pad
from
our
office
building
and
we
deploy
our
knows
ni
rooms
from
the
building
and
each
room.
E
First,
we
are
waiting
for
all
knows:
synchronized
and
building
networks,
and
the
figure
on
the
right
is
showing
the
structure
of
the
networks
after
the
routing
is
established
and
by
default.
There's
a
background
upstream
traffic
around
one
packet
per
30
seconds
upstream,
transmitting
and
to
generate
a
downstream
traffic,
and
we
just
using
the
ping
command
provides
by
the
OS
and
we're
paying
to
a
one
home.
E
Node
be
5f1
there
and
with
a
spade
of
one
pain
for
two
seconds,
is
roughly
once
loved
one
packet,
of
course,
offering
duration
and
the
reason
why
we
only
just
paying
this
first
hop.
Now
it
is
the
experiment,
that's
trying
to
show
the
crowning
correctness
of
the
MSF
downstream
epic
adaptation,
and
so
we
are
recording
the
data
such
as
the
traffic
changes,
which
is
when
the
pain
pinkman
stars.
E
E
E
A
E
E
E
E
Removed
from
its
schedule
and
as
well
on
the
target
mo
side-
and
you
can
see
it
installed
in
an
additional
T
Excel
because
as
beginning
has
a
1
T
Excel
there
so
so
is
a
result
with
two
at
the
exit
there
there's
a
there's
a
time
for
target
notes.
You
can
just
see
it's
showing
there's
a
cerise
out
there,
which
is
as
the
right
beginning
its
dropped
two
to
one
so
I
smelled
ourselves
faster
in
there
Ryan.
I
I
Maybe
it
would
be
nice
if
you
just
put
other
per
cell,
maybe
maybe
many
lines
cell
analysis
and
just
a
comment.
E
Yeah
I
show
you
here
is
trying
to
using
this
value
to
matching
the
speed
of
a
the
pinging.
If
I
summarize
them
together,
you
can
see
the
lines
is
actually
increasing.
One
transmission
for
two
seconds
I
was
trying
to
trying
to
connect
to
this
of
this
lines
to
the
actual
speed
of
a
the
downstream
traffic
yeah.
That's
the
idea
to
show
you
some.
A
E
A
Yes,
it's
just
basically,
it
will
occur
if
the
traffic
change
is
too
fast
for
the
6p
protocol
to
allocate
its
lanes
cells
right.
So
if
it's
a
ping-
and
there
is
a
packet
every
two
seconds,
then
maybe
that
gives
you
enough
time
to
allocate
this.
The
cells
I
think
it's
pretty
much
the
answer,
so
it
can
only
adapt.
You
know
at
a
certain
speed,
so
you
changes
of
the
demand
must
be
slow
out
and
the
changes
at
which
you
can
adapt
to
it.
Yeah.
That's
true!
E
Yeah,
so
we
are
running
so
we
believe
the
drop
to
the
set.
The
0-7
version
is
ready.
Then
we,
we
start
a
two
weeks
walking
group
let's
go
around
one
months
ago
and
we
didn't
receive
comments
either
were
responding.
Hopefully
people
review
it
in
just
no
comments
and
so
right
after
the
meeting
this
time
we
will
publish
another
version.
A
Okay,
yes,
I
mean
what
compressed
called
completed
without
any
comments,
so
I
left
it
kind
of
opening
that
maybe
that
would
be
late
commands
and
well
I
wanted
to
discuss
now
what
we
do
with
this
document.
So
we
have
not
getting
commands,
we
have
experimental
code,
it's
it
does
what
it's
supposed
to
do?
It's
just
one
possible
scheduling
function.
It
doesn't
tell
you
that
everybody
has
to
schedule
this
way.
A
C
As
an
extra
and
so
personally,
if
there's
not
interest
in
the
looking
to
progress,
it
I
don't
want
it.
But
if
it's
like
bunch
of
implementations
and
doing
it
I
think
it's
like
a
mitigating
factor.
So,
like
you
know,
just
try
a
couple
more
times
to
prod
a
few
people
personally,
the
draft
as
possible
for.
A
A
C
A
Has
been
a
long-standing
work
in
this
working
group
program,
we
started
with
other
documents
which
migrated
into
this.
One.
I
mean
it.
So
it's
been
a
long-standing
and
there
are
multiple
people
working
on
it,
I'm
pretty
confident,
it's
quite
good,
and
then
it's
like
a
no
repo
you,
somebody
could
devise
another
ones.
That's
like
we
impose
that
everybody
worked
this
way.
It's
just
just
totalizing
at
its
one
possible
way
of
doing
this
job.
That's.
C
Fine
and
as
long
as
you
document
like
in
the
Shepherd
right
up
the
implementations
as
well,
let's
could
be
very
useful.
Like
you
know,
the
question
comes
up:
why
it's?
It
needed?
Okay,
so
thankful.
If
you
can
send
out
the
implementation
info
to
like
or
past
color,
Thomas
L
know
who's
gonna
ship
at
it
or
somebody.
B
A
G
C
A
And
so
I
don't
see
Michael
coming
so
like
we
so
before
it
might
be
in
rats.
So
anyway,
I
could
I
could
just
push
these
slides
and
we
can
just
look
at
them.
Basically,
the
work
is
no
more
happening
at
60.
It's
it's
all
about
saying,
eh,
it's
happening
elsewhere.
We
keep
60,
show
where,
as
long
as
60
she's
interested
which
it
is-
and
then
you
know
that
does
not
change
the
the
destiny
of
six
in
any
fashion.
A
A
So
if
yes,
the
basically
the
rest
of
this
session,
is
to
talk
about
the
future
of
this
work,
and
the
only
document
that
could
have
still
be
opened
in
the
working
group
would
be
this
so
can
office
T.
If
the
discussion
leads
to
you
know,
closing
the
working
group
then
we'll
see
if
we
need
to
reopen
this
document
in
like
or
just
you
know,
consider
that
it's
whatever
happens
in
the
other
working
groups
is
enough
to
cover
zero
touch.
C
A
Well,
the
question
to
the
room
now
is:
do
people
see
that
there
is
work
that
should
happen
in
this
working
group
for
which
we
were
not
chartered?
Is
there
an
interesting
discussion
that
we
could,
for
which
we
would
need
to
create
new
documents?
My
reading
of,
what's
going
on
is
there
are
two
main
aspects
which
I've
been
discussed
in
these
three
of
six
station,
from
which
we
did
not
complete
work.
A
One
was
scheduling
time
slots
along
a
track
and
the
other
one
was
shuffling
the
the
cells
in
order
to
defeat
some
some
tracking
attacks,
that
was
Marcos
work,
makuu
Tanaka's
work.
In
both
cases,
it
seems
to
me
that
the
work
is
not
really
ITF
work,
but
mostly
I
Triple
E
work
if
it
happens,
and
so
I
hardly
see
that
we
need
our
canvas
which
other
6-4
under
viewing
this
work,
which
tells
me
that
all
the
work
that
I
was
aware
of
that
was
supposed
to
happen
at
6:00
is
now
past.
A
Well,
group
Glasgow,
apart
from
the
zero
touch
thing,
which
also
tells
me
that
unless
somebody
just
now
to
say,
hey,
we've
got
new
work.
This
is
probably
the
last
time
we
meet
so
I
talked
to
Eric
and
Suresh
and
well.
The
advice
was
pretty
much
this
right.
We
complete
the
document
that
we
have
running.
So
we
keep
the
mailing
list
active.
We
keep
the
workgroup
quoted
code
following
see.
All
those
documents
goes
through
the
iesg.
A
C
Yeah
suresh
krisshnan,
so
that
kind
of
makes
sense
right.
So
one
thing
I
kind
of
like
I
wanna
add
there
is
like
let's
say
in
this
period:
we
keep
it
open.
If
something
comes
up,
we
will
see
how
things
go,
but
yeah
generally
I
see
this
heading
for
closure
rather
than
a
recharger
right.
Well,
it's
like
some
amazing.
What
comes
up?
Why
not.
A
Now
it's
not
like
the
work
on
all
this
is
finished.
Like
I
said,
we
are
trying
to
create
row
at
the
routing
area,
which
will
focus
on
the
concept
of
track
and
generalize
it
over
multiple
macfay,
so
people
in
the
room
invited
to
follow
the
work
which
happens
in
reliable
and
viable
wireless,
which
is
really
the
logical
continuation
of
this
particular
aspect
of
the
six
dish.
Architecture.
J
Marko
confirm
rise
and
just
comment
to
clarify
my
personal
opinion
on
the
belonging
of
the
draft
on
circuit
of
jamming
and
schedule,
shuffling
to
ITF
or
typically
I,
think
it
belongs
to
ITF.
In
fact,
just
reading
the
introduction
from
the
Charter
it
mentions,
15-4
does
not
define
how
the
network
communication
scale
was
built.
So
this
is
a
kind
of
a
gray
area,
because
the
attempt
here
was
to
complete
the
building
of
the
schedule.
J
In
fact,
second,
in
the
latest
discussions
we
had
in
the
list
in
the
draft,
mostly
with
militia
and
Michael
Richards,
and
we
were
starting
to
think
about
how
to
handle
the
king
as
to
the
keys
use
for
that
permutation,
possibly
to
Inca
GP,
and
that
working
is
something
is
definitely
not
going
to
happen.
Ethically.
B
C
A
A
Well,
okay,
so
jamming
one
is
freaking
one,
that's
to
read
different
aspects
right,
so
do
you
want
to
focus
the
discussion
on
jamming
because
we
discussed
a
lot
on
jamming
last
time,
I
talked
to
him
about
the
details,
but
there
was
a
big
reason
why
it
could
not
be
done
at
the
apparel.
A
yeah
I
mean
I
need
to
go
through
the
minutes
of
lifestyle,
because
now
it
escapes
my
mind,
but
but
you.
J
A
Do
you
remember
something
about
it?
What
we
discussed
last
time
Tarot
last
time
we
discussed
the
you
know,
Marcos
proposal
on
shuffling
that
the
time
slot
to
defeat
tracking
attacks,
we
ended
up
with
a
problem
which
was
related
to
security,
which
made
it
so
that
you
had
to
do
something
at
I,
Triple
E,
to
make
it
work
and
I.
A
C
So
suresh
krisshnan
again
so
marco,
can
you
start
a
thread
on
the
mailing
list
trying
to
see
if
there's
an
interest
to
work
on
this
because
it
may
or
may
not
be
relevant
is
one
thing?
Is
their
energies
and
other
things
I
want
to
see
both
of
them?
Okay,
so
start
a
thread
on
the
mailing
list.
We're
not
closing
the
group
right
now
right,
so
I
would
think
like
probably
a
year
away
or
something
right
kind
of
the
timeline.
A
Yeah
but
then,
like
I,
said
last
time,
we
discussed
it
extensively
at
this
meeting
and
my
memories
that
we
could
not
do
it
here
because
it
had
dependencies
on
I,
Triple
E
that
we
could
not
solve
it
in
this
room.
That's
my
memory
of
it
right
and
maybe
wrong,
but
I
mean
you're
welcome
to
to
to
like
start
a
thread
on
the
mailing
list
and
see
where
we
are
and
make
sure
see.
If
my
memory
is
correct
on
this
for
the
scheduling
of
cells,
we
want
it
to
do
here
to
do
that
work
here.
A
Initially,
it
could
have
been
everywhere
because
it's
mostly
defining
a
data
model
that
comes
from
you
know,
magical
guard
up
there
and
it's
being
pushed
to
the
devices.
It's
a
data
model
thing
an
information
model.
So
do
we
define
this
data
model
information
model
at
the
HF
hie
I
mean
it's
this
gray
area.
It
could
have
happened
here,
but
the
point
is
we
lost
the
energy.
Nobody
really
did
anything
there
that
one
document
started
and
they
died.
So
keeping
shout
out
for
something
and
nope
no
people
doing
the
job
doesn't
help.
A
C
A
C
I
understand
right:
we
don't
want
to
discuss
actual
technology
I'm
just
talking
about
the
mechanics,
how
we
can
go
about
this
right.
So
let's,
let's
do
this
on
the
list
like
this
like
few
people
here,
that's
like
I,
don't
want
to
make
the
decision
here,
but
just
go
on
the
list,
see
this
interests
and
then
we'll
figure
out
what
to
do
right.
As
I
said,
we
have
time
time
to
figure.
A
And
then
there's
this
question
of
freaking,
so
my
memory,
my
status
in
my
own
mind
of
freaking,
was
like
most
of
the
time
it's
not
useful,
because
ASN
will
not
show
up
or
something
like
that.
Everything
is
useful
when
you
want
to
evacuate
some
devices,
because
that's
pretty
much
the
only
way
you
can
get
away,
you
can
remove
somebody,
yes,
so
so
what
else
exactly
needs
to
be
done
is
what
I'm
confused.
J
Moment,
good
membership
change
you're
supposed
to
recognize
anyway,
and
we
found
a
way
to
distribute
new
piece
for
this
permutation
together
with
the
process
described
in
code
JP,
and
that
has
some
complications
for
the
moment
where
you
have
overlapping
between
the
old
king
material
and
a
new
one,
and
we
were
starting
to
think
of
a
way
to
handle
that
issue.
There
were
similar
discussion
with
militia
in
my
own
on
the
list
is.
B
A
J
J
It's
not
our
intention
to
change
anything
in
Kahn's
opinion,
the
way
it
works.
This
is
just
an
extension,
more
keys
to
be
delivered
in
the
wreaking
messages
and
how
you
can
make
this
very
wrecking
work.
It
has
no
impact
on
code
rip
itself.
Does
it
mean
that
everything
could
be
done
in
one
document?
That's
ideal
to
me.
E
B
A
Like
we
said,
we
have
a
mailing
list,
we
don't
close
the
working
group,
we
just
don't
have
a
plan
to
meet
next
time.
That's
pretty
much
where
we
are
right
now,
I
mean
there
is
not
enough
content,
that's
very
active
to
justify
asking
for
a
meeting
next
time
unless
you
know
something
explodes
on
the
mailing
list
and
then
well
so
well
on
this
side,
which.
C
Might
sounds
good
so
like
a
suresh
krisshnan,
so
the
working
group
is
kind
of
gonna
go
dormant,
like
waiting
for
the
documents
to
finish
and
if
what
comes
up
it's
happened
before
like
same
thing
happened
for
dat
right,
so
we
didn't
meet
like
twice
and
then
they
did
like
so
documents
came
up
and
they
found
work.
So
it's
really
about
like
work
showing
up
and
interests
in
it
to
keep
it
going,
but
the
default
is
it's
gonna
close.
C
A
I
believe
that
60-ish
was
a
fantastic
experience.
It
was
very
interesting
in
the
fact
that
it
was
orchestrating
a
lot
of
work
in
multiple
documents
and
providing
the
architecture
that
ties
them
all
together,
and
so
that's
one
aspect
of
it:
the
second
spectres
driving
open-source
there
were
multiple
open-source
implementation,
and
the
working
group
also
acted
as
a
hub
outside
of
the
IETF
to
to
integrate
those
those
open-source
implementations
and,
finally,
the
interaction
with
Etsy
for
all
this
black
test.
A
So
those
three
aspects
put
together
and
orchestrate
it
from
within
an
IETF
walking,
Rock
I,
think
that
was
a
really
really
interesting
experience
and
that
leads
to
effective
code
running
and
interpreting
open-source
implementation
and
RFC's
here
and
in
multiple
groups.
I
think
this
group
did
very
well
and
I
wish
to
thank
you
know
us,
probably
the
best.
Last
time
we
meet
I
wish
to
thank
the
IDS
and
and
all
the
participants
of
other,
like
like
six
or
seven
years,
for
which
we
did
six
for
the
intensive
work
that
happened.