►
From YouTube: IETF98-TRAM-20170328-1450
Description
TRAM meeting session at IETF98
2017/03/28 1450
B
D
E
C
C
So
the
agenda
is
we'll,
do
our
update
and
then
we
have
a
presentation
by
mark
about
Sundays.
He
will
try
to
get
all
the
open
issues
resolved
and
get
this
working
last
call,
and
then
we
I
would
like
to
have
a
working.
A
group
discussion
about
turn
this
to
make
sure
that
we
are
aligned
with
the
latest
changes
in
Sunday's
and
then
the
meeting
ends.
So
there's
nothing
extracurricular
on
the
agenda
next,
so
think
of
the
tram
since
Berlin
things
have
been
moving
fast
aboard
the
tram
and
it
doesn't.
C
C
Stun
path,
MTU
discovery
has
been
revised,
there's
been
a
working
group
last
call
and
the
draft
will
be
will
be
sent
to
a
ESG
shortly.
We
just
need
to
do
the
work
and
then
stunned.
This
has
been
revised
three
times.
There
are
quite
a
few
change,
I
wouldn't
say
major
changes,
but
you
know
medium
changes,
medium
impact
changes,
and
we
want
to
close
the
discussion
today.
We
want
to
finish
this
and
send
is
the
working
class
call
and
turn
this
has
been
refreshed.
This
was
a
minor
refresh
and
works.
C
F
F
Hello
again
next
slide,
please
all
right.
So
last
time
we
presented
stunned.
This
was
back
in
Delhi
I
thought
there
was
a
lot
of
command.
Well,
not
a
lot
of
comment,
but
some
some
comments
were
made
and
we
with
my
cursor,
gonzalo
with
just
started
walking,
and
then
we
found
yet
another
egg
and
then
another
burger
and
then
adult
work
and
then
I
said
okay.
We
need
to
stop
to
stop.
There
was
something
wrong
and
I
need
to
take
a
bit
more
time.
To
do.
F
F
So,
finally,
I
did
an
ADIZ.
It
was
not
an
LED
son's
dundas,
it
wasn't
LEDs
on
start
on
ever
see.
50
389
and
I
found
a
bunch
of
public
much
I,
some
minor
of
some
less.
So
we
talked
her
talk
with
food
return,
which
we
want
to
dunk
again
for
his
help
on
all
the
surveying.
Some
and
then
we
send
a
proposal
for
ELISA
was
somewhere
I
can
forcibly
put
the
modification
in
size
as
early
story
police.
Also
so
modification
so
I
will
go
through
quickly.
F
F
We
made
something
explicit
I
mean
you
could
deduce
how
it
works
and
for
the
people
implemented
the
white
where,
but
it
was
not
explicitly
stated
how
to
to
walk
about
400
without
authentication.
My
experience
is
that
usually
need
to
explain
all
the
step
carefully,
so
we
did
that
we
found
some
of
the
should.
We
did
not
make
sense
at
all
in
the
document
in
exam
most
and
some
place.
F
F
We
so
small
modification,
which
is
I,
think
some
case,
an
improvement
in
security,
which
is
when
you
have
some
dings
announced
or
we're
changing
the
password.
If
you
still
out
the
nonce
from
the
previous
time,
we
are
still
the
bus
throughout
from
the
previous
time,
then
you
can
use
it
to
sign
the
response
instead
of
sending
naked
or
respond.
So
it's
nice.
So
you
do
it
or
you
don't
do
it,
but
if
you
do
it,
obviously
there
is
a
little
bit
more
security
than
you
can
add.
F
You
cannot
before
okay,
so
one
of
the
thing
was
that
and
only
that
it
was
401
unauthorized
which
is
kind
of
incorrect.
This
is
not
an
authorized.
This
is
an
authenticated
so
which
one
user
text
turn
is
doing
user,
separate
response
Cody
for
an
authorized,
different
beginner
and
there
were
some
tags.
I
said
you
can
add
an
in
my
message.
Integrity
and
some
rest
months
were
clearly.
There
was
no
no
possibility
at
all
to
do
this,
so
we
remove
the
possibility.
So
this
was
the
first
aid
with
the
bishop
that
was
probably
dash,
0
9.
F
So
we
remove
some
silly
things
like
the
other
preserve
the
0
in
the
security
orchestrated
make
sense,
because
this
is
a
bit
sets
obviously
the
horses
of
meeting
in
a
bit
set,
except
that
no
be
following
the
recommendation
of
the
working
group.
We
said
that
the
minimum
potential
size
for
ni
256
is
16
bytes.
F
F
F
We,
this
was
silly.
We
remove
the
truncation
from
the
message
integrity.
This
is
an
old
stuff.
We
don't
need
to
care
about
this.
This
is
going
out
all
over
the
world
missing
recommendation
from
well
Jonathan.
We
added
a
link
to
the
secret
out
of
white
subs
are
cut
out
two
tubes.
Also,
this
is
a
COC
guess
thanks.
F
So
yes,
I
think
it's
an
important
because
it's
not
easy
to
find,
and
finally,
this
was
a
response
from
an
email
from
early
during
their
lean
one
of
the
beginning.
Is
that
only
sent
and
we
replace
the
restaurant
storage
here,
2018
360
125-
and
we
are
dated
the
text
accordingly,
to
explain
how
to
verify
the
domain
name
from
the
certificate
on
and
to
verify
hit
against
the
the
Monday
next
slide.
Please
and
then
we
did
yet
another
update,
which
is
dash
11,
which
is
a
last
one.
F
So
I
went
back
for
all
these
things
and
I
am
I
fixed
a
little
bit
more
stuff.
So
first
one
is
kind
of
abuse
is
to
say
that
we
should
use
same
type
of
ash
marking
the
subsequent
collection
that
are
used.
But
again
we
really
need
to
spell
it
out
these
things
clearly
and
the
last
one
dress.
One
is
a
little
bit
more,
it's
a
little
bit
as
it
was
from
a
suggestion
from
someone
I,
don't
remember
if
you
think
it
was
in
darling,
so
add
to
this
version.
F
We
used
what
we
call
the
nonce
cookie
only
when
we
needed
it.
So
the
nonce
cookies
are
small
cookie
that
we
had
in
front
of
the
nonce
to
say
we
have
some
additional
security
properties
that
you
want.
We
want
to
use,
and
there
was
one
at
the
beginning,
which
was
a
possibility
to
change
the
password
algorithm,
and
then
we
added
a
subtle
one,
which
was
to
be
able
to
ash
username
alright.
So
the
problem
now
is
that
you
have
an
exponential
number
and
increase
of
the
number
of
cookie
with
the
itu-r
for
possibility.
F
You
need
for
cookies
well
exactly
because
you
don't
use
this
year
so
instead
of
doing
this
on,
what
we
did
we
did
is
that
we
added
for
which
four
characters
to
the
coquito
between
the
cookie
and
the
non
sir.
We
are
the
out
there,
which
is
24
bits
when
it's
connected
from
base64,
which
gives
us
24
bits
of
additional
security
properties
that
you
can
choose
in
in
the
future.
F
We
created
the
registry
for
that
and
we
allocated
the
first
to
date,
so
those
for
and
if
you
want
to
use
password,
algorithm
and
also
long
dick
is,
if
you
want
to
use
user
hush,
you
want
to
wash
your
username,
so
the
next
step
was
well.
Why
don't?
We
send
always
the
nonce
cookie,
even
before
zero
value
in
the
bit
set
to
signal
that
we
support
and
mi
to
is
that
we
support
the
new
ashmac
based
on
the
night
sure
250
six.
F
Understand
this
client
knows
immediately
that
the
server
is
also
austin
disk
Ryan
and
do
not
have
to
send
both
the
message:
integrity
and
the
message
integrity
256
to
the
server.
So
we
the
first
the
first
step-
is
a
little
bit
simpler
for
both
the
client
and
the
server.
But
now,
if
you
don't
use
any
of
the
security
properties,
then
you
just
send
a
bit
set
that
contain
all
the
other
side.
This
is
why
we
remove
it
also
konser
register,
so
that
was
always
done
and
this
is
occur
on
the
current
set.
F
A
Time
so
get
yourself
mark
before
you
go
on.
We
would
like
really
to
close
these
issues
right.
So,
let's
focus
on
closely
not
just
going
through
and,
as
you
know,
it
happens
in
many
working
groups.
So
so,
please,
you
know
each
of
the
issues.
Let's
give
it
a
thought
and
we
will
only
postpone
it
if
there's
really
need
to
happen,
something
else
like
more
information.
Otherwise,
we
would
like
to
really
make
the
decisions
today.
So.
F
Which
kind
of
make
me
not
coping
I
as
I
skipping
that
we
are
doing
the
right
things?
I
am
or
we
had
a
long
dist
description,
the
computer
or
neither,
and
we
think
that
the
ing,
maybe
not
completely
at
ease
with
the
things
we
are
doing
so
and
I
was
thinking
about
this
because
I
know
if
some
of
you
went
to
the
security
tutorial
on
Sunday,
which
by
the
way
was
very
interesting,
and
that
was
one
slide.
That
said
to
do
not
do
exactly
what
we
are
doing
so
well.
So.
F
So
they
said,
use
agility,
do
not
use
a
disc,
changer
sagittis,
so
I
do
not
use
agility
for
algorithm.
Do
not
use
a
beef
shank,
the
algorithm,
which
is
using
a
protocol
which
made
exams.
So
the
thing
is
that
I
went
back
to
the
mailing
list
and
everybody
more
or
less
a
great
such
as
was
the
right
thing
to
do.
Keep
it
simple
for
what
we
are
doing
all
right
now
it
will
go
off
in
front
of
the
AV,
the
aeg.
We
said.
Okay,
we
said
not.
Okay,
I
don't
know.
F
Question
which
is,
do
we
still
want
to
do
this
or
not,
and
I
would
like
to
add
one
small
thing,
so
this
thing
was
difficult
to
design
and
to
do,
and
the
reason
is
that
I
could
not
find
a
lot
of
guidance,
so
those
be
done
attack
prevention,
saturated
by
at
the
IETF,
a
nice
Arabs.
Also
the
literature,
Windsor
I
mean
I.
I
went
to
a
jcm
library
and
Google's
crust
and
there
is
not
people
talk
about
this,
but
that
will
explain
kind
of
farm
work,
how
together
Madonna
attack.
F
So
please
do
to
me
that's
not
an
area
which
is
really
well
explained
on
my
lands.
In
any
way
we
like
to
put
this
questioning
in
front
of
you.
Would
we
continue
this
this
way?
Do
we
want
to
charge
another
thing
that
make
mending
things
about
this?
So
just
after
the
shower
attacking
the
project,
which
is
part,
is
the
biggest
user
of
shower
ashes
and
the
people
engaged
after
to
say
well
thought
we
should
do
something
I
and
they
are
jumping
directly
from
sherwin
to
shark
week.
They
will
be
doing
shaft
32
62-56.
F
We
are
doing
shots,
ooh
256.
So
are
we
doing
the
right
thing?
Where
do
you
want
me
to
go
on
up
to
the
end
when
we
we
talk
about
this
or
we
be
open?
So,
let's
see
fixing
one
inch
at
a
time.
I
think
so.
A
C
F
G
Gonzalo
so
the
reason
that
I
think
that
this
is
an
issue
is
that
we've
made
this
decision
before
right.
I
mean
we
decided
that
we
would
stick
with,
shall
one
so
no
church
I
hope
not
not.
We
stick.
You
shall
on
your
students
which
site
was
shut
it,
but
the
idea
here,
though,
is
do
we
need
to
revisit
that
decision,
simply
to
be
essentially
to
prevent
with
what
we
think
we're
anticipating
from
the
iesg.
H
D
H
But
if
the
argument
has
it
changed,
if
it's
really
just
you
know,
that
seems
okay
and
I
think
the
argument
at
the
time
was
well,
it's
just
as
much
work
to
add
another
option
type
to
the
stun
registry,
as
it
is
to
add
another
hash
type
to
some
other
registry
that
you
would
set
up
and
let's
wear
a
be
able
to
piggyback
easily
on
a
registry
that
already
exists.
So
if
fundamentally,
it
might
just
be
a
question
of
you
know:
registry
management,
overhead
at
least
I-
think
that's
what
the
opinion
was
at
the
time.
H
I
Strong
clinics
I
mean
so
the
first
thing
I
mean
for
what
we're
using
its
board
for
HVAC
the
end
of
Shah
one,
that's
not
on
us,
not
one.
It's
not
broken
for
a
snack,
so
I'd
also
for
shot
to
I.
Don't
think
anybody
is
saying,
and
nobody
thanks
lucky
that
the
care
of
this
one
desktop
read
is
a
very
different
design,
so
it
gives
more.
You
know
it's
you're,
less
likely
to
have
a
break
that
his
belt.
In
the
same
time,
I
feel
like
I,
think
actually
are
bits.
I
F
I
F
I
And
I
feel
like
this
I,
don't
see
what
we
would
gain
by
saying
we're
going
to
put
those
additional
bit
of
sake.
What
Hatcher
algorithm
it
is
in
the
message:
integrity
into
a
mess-
integrity
active.
So
just
you
know,
get
another
registry
and
get
another
n
bits
field.
That's
take
those
space
and
messages.
Okay,
I
think
you
want.
I
G
I
How
is
that
different
from
saying
we're
going
to
define
a
patch
ability
by
putting
it
in
putting
big,
you're,
still
good
happy
to
find
hey
use
this
algorithm
whatever
this
is?
This
is
the
registry,
for
that
mean
I
mean
maybe
I
mean
if
you
want
to
say:
okay,
here's
a
lighter-weight
template
for
defining.
You
know
message
integrity
and
it
message
integrity,
your
hash
algorithm
here
you
know,
as
you
know,
a
you
know,
a
cut-and-paste
register,
aveeno
Sun
attribute
registration,
and
this
has
the
special
property
that
the
message
integrity
towards
the
end.
I
You
know
Dorothy
Africa
de
yada
yada,
maybe
having
a
somewhat
more
general
description
of
how
I
message
integrity.
Whatever
it's
worth,
that's
fine,
but
it
doesn't.
I
think
that
everyone
changed
how
they
look
on
the
wire
this
team,
so
it's
documented
and
makes
the
work
for
defining
a
new
message.
Integrity,
shot
three
or
whatever
slightly
less
harm,
and
that's
not
unreasonable.
I.
Think
that
would
have
had
that
more
emphasized
in
the
document
is
their
version
of
'deck.
Exactly
and
and.
H
I
I
got
to
say
exactly
what
Jonathan
got
to
it
in
the
end
that
if
either
this
is
our
approach
for
hash
agility,
that
the
thing
that
name
is
missing
from
the
document
is
a
clear
statement
of
this
in
our
approach
and
when
you
want
to
add
another
one.
This
is
what
it
looks
like
to
get
there.
That
may
be
what's
necessary
to
convince
those
the
quantity,
a
hash,
agility
approach
that
we
actually
have
that
covered
ya.
Gonna.
F
F
J
My
mind
will
be
quick,
I'm
Spencer
darkens,
the
responsible
yd
somebody
invoke
the
name
of
the
is
she
doing
doing
with
you
perhaps
sector
like
any
director.
That
I
think
will
accept
requests
for
early
review.
If
you
all
want
to
go
to
short-circuit
long
conversation
at
IHG
evaluation
time.
That
may
be
a
good
thing
to
do.
Report.
Okay,.
A
I
A
C
A
C
I
C
I
C
H
I,
I
think
this
this
one
get
given
the
mechanism
as
its
described
here.
I
I,
think
protection
against
pitt
downs
does
require
that
you
have
an
orderly
process
of
I've
got
you
now
I
deprecated
one
then
I
add
another
one
and
you
can
move
forward
that
way.
But
if
you
have
a
multiplicity
that
are
allowed,
it
just
gets
more
and
more
complicated
to
make
sure
that
you've
got
everything
covered.
I
Yes,
Jonathan
Ive
tonight,
I
think
and
but
there's
no
I'm,
II
nice
big
down
as
I
understand.
It
is
the
case
where
you
know
both
sides
want
to
use
the
new
hash
algorithm,
but
some
attacker
in
the
middle
is
forcing
the
good
undo
it
all
right,
so
it
by
the
size
doesn't
want
to
use
the
new
one.
It's
not
a
bit
down
shift
back,
peplum
sew
as
I
think
that
we
for
mi
56
/,
say
we
have
a
pretty
good
story
for
long-term
production
of
positive
of
your
confidential.
I
I
H
Brandon
Williams
again
isn't.
H
Am
I
remembering
correctly
that
with
the
Shah
12
sha-256,
the
thing
that
protects
against
bit
down
is
the
the.
What
are
we
calling
it?
The
nonce
cookie
and
once
we've
gone
beyond
one
and
everything
is
delivering
the
nonce
cookie.
What
will
protect
against
bit
down
for
256
to
sha
sha
22
shot
three
or
whatever?
If
everything
is
it's
a
really
good
fun,
probably
a
new
cookie,
so
probs.
F
I
D
I
I
I
D
D
I
F
A
A
We
will
wait
for
you
to
rewrite
the
dress
and
before
asking
for
any
early
review,
okay,
so
everyone
opposed
to
these
way
forward.
Please
hum!
Ok!
So
then
yeah
you
can
move
forward
with
that
by
the
way
when,
when
mark
doesn't
remember
what
we
have
discussed
and
he
goes
to
the
recordings-
everything
that
you
guys
are
saying
outside
the
mic,
you
know
it
would
not
be
captured
so
I
think
that
would
be
useful
to
to
try
to
use
the
mic
okay,
so
on.
Let
me
go
to
the
next
slide.
D
F
D
F
We
were,
we
didn't
know
exactly
what
to
do,
but
I
see
did
an
analysis
of
all
the
case.
I
know
what
we
do,
I
adopt
so
lunch
with
that
we
gave
us
some
rice
and
we
talked
with
her
Jonathan
yesterday.
I
think
about
this.
So
the
problem
is
that
the
analytics
thing
is
not
the
stunting.
Is
the
stern
usage
all
right?
F
It's
not
really
stunned
by
itself,
not
the
mechanism,
it's
how
you
use
them
that
start
to
have
issues
when,
when
you
reserve
the
man
named,
you
receive
one
or
more
ipv4
advice
and
one
or
more
ipv6
advice.
Smart,
so
I
found
three
cases,
so
the
first
one
is
more
or
less
everything
which
is
related
to
it.
D
F
F
This
is
what
among
other
things
ice
is
about
you
prioritize
ipv6,
before
I
ipv4,
you
check
the
connectivity
automatically
give
give
you
what
is
a
current
best
path
so
using
this,
and
you
don't
really
have
inside
eyes
when
you
are
well,
you
can
get
done
and
I
will
explain,
but
it's
kind
of
difficult
to
do
two
of
two
different
algorithm,
all
right.
If
we
do
at
IPL
eyeballs
and
we
have
the
first
selection,
which
is
done
by
eyes,
and
then
you
have
a
second
algorithm
which
shows
what
is
the
connection
to
use
underneath.
F
F
K
K
J
L
Schwartz,
google
so
similar
to
an
HTTP
proxy
in
the
in
the
context
of
return,
there's
also
where,
where
you're
using
turn
services
proxy,
there's
also
a
situation
where
it's,
it
can
be
logical
to
use
a
domain
name
as
your
endpoint
identifier
and
I
I
tried
and
failed
to
get
a
draft
through.
That
would
create
a
an
explicit
mode
for
turn
to
allocate
to
allocate
a
remote
endpoint,
based
on
a
domain
name
and
I'd
like
to
be
able
to
try
that
again
so
I.
Hopefully
it
won't
get
cut
out
entirely.
I
Shot
life
and
I
work
say
is
how
I
said
I
mean
how
ice
behaviors,
not
something
that
is
not
this
working.
This
problem,
so
I
mean
and
I
think
this
and
I
think
you
can
come
up
with.
Even
if
you
want
to
leave.
The
main
name
is
a
you
know.
You
know
your
local
candidate,
cetera
type,
so
parative
be
for
a
dress.
You
looked
up
for
of
your
be
four
candidates
or
the
v6
letters.
He
looks
ugly,
racist
and
eggs
or
whatever
it's
really
more
complicated
mess.
I
F
So
the
other
case
is
the
seat:
keeper
I,
which
is
much
inside
ice,
but
I
think
this
is
not
an
issue
because
I
already
know
what
kind
of
motor
car?
What
cannot
transport
the
youth.
So
it
is
not
an
issue,
so
your
proposal
would
be
led.
The
ice
people
deal
with
it
which
by
the
waist
Semin,
so
we
left
for
town,
but
we
will
get
to
this.
G
Thing's
also
here,
if
we
do
that,
I
mean
I,
figured
they're
released,
need
to
be
some
text
here
around.
G
F
K
K
I
F
F
G
F
F
I
I
A
F
Something
like
yes
I
will
topics
with
okay
all
right,
so
the
second
line
is
what
we
have
to
do
anyway,
because
this
is
the
stern
usage.
Well,
one
of
these
is
described
inside
this
up.
There
I
see
and
the
other
one
exactly
the
same,
so
it's
relatively
simple
there,
with
I
audio
from
the
place
where
to
charge
it.
F
The
right
thing
with
not
discovery
that
we
use
binding
binding,
is
an
miss
it.
What
it
impotent
required
transaction,
which
means
we
can
send
as
many
as
we
want.
We
are
not
breaking
anything,
and
we
have
some
use
case
at
Jonathan.
Explain
to
us,
it
gets
yesterday,
so
it
will
be
0
actively.
A
simple
walk
to
add
to
towards
the
text
to
the
soul
is
certainly
do
the
booze
Arctic.
Unless
there
is
question
or.
I
I
But
yeah
I
think
it's
certainly
framing
the
ice
cases.
The
terminal
allocated
your
candidates
for
all
your
interfaces
and
then
let
some
before
in
some
be
six,
so
you
send
to
the
v4
stunts
address
from
before
it.
It'll
be
nice
to
them
address
in
36
and
that's
easy
and
you
even
want
you
need
to
set
it
from
all
your
interfaces
anyway.
I
F
I
will
not
take
a
way
to
deal
with
you
talk
with
you
so,
but
so
it
remember
to
do,
and
so
the
last
one
the
difficult
one
is
the
point,
because
the
problem
is
that
another
case
in
turn
is
not
a
dedenne
potent
collections,
HW
send
navigation
requests
and
you
can
receive
a
pot
and,
and
it
could
be
or
not
an
issue.
I
am
I
I,
don't
know
so
it's
a
nice
progression
to
the
next
discussion
and
we
think
that
this
is
a
term
of
them
and
they
should
think
I'll
to
fix
this
thing.
F
F
F
F
A
G
G
G
C
Participant
yeah
I
think
it
totally
makes
sense
for
this
to
be
dependent
on
the
actual
usage,
because
done
can
be
used
at
the
IP
layer
to
discover
IP
layer
properties
or
it
can
be
used
as
a
communication
protocol.
For
example,
when
talking
to
turn
server,
you
don't
really
care
about
the
family
of
IP
that
you're
using.
So
it
depends
on
the
under
usage.
A
F
A
C
Right,
okay,
I
was
looking
at
so
Simon
as
a
participant.
Again.
I
was
looking
at
section
6.2
in
chambers,
which
explains
what
the
server
does
when
it
receives
an
alligator
request
and
how
it
detects
retransmissions,
and
so
it
uses
transaction
ID
and
the
five
tuple.
So
one
option
would
be
to
remove
the
five
people
from
the
check.
If
you
receive
the
same
transaction
ID
on
a
different
five
tuple,
possibly
from
somebody
else,
it's
the
same
transaction.
So
keep
in
mind
there
96
bits
of
randomness,
they're
lost
authentication.
K
E
I
I
Kimmy
elaborate,
basically,
you
send
an
allocate
request
with
no
authentication
information
to
both
the
forest,
except
versus
you'll,
get
a
401
response
from
one
or
the
other
or
both
at
which
point
you
send
the
one
with
your
dedication
fish
awesome
unless
somebody
figured
a
print
server
to
not
movies
with
medication
in
Todd,
restless
expect
with.
Sometimes
people
do
it's.
I
Actually,
no
we're
changing
that
for
that.
That's
because
we're
jaitley
it
was
mad
at
her
in
turn,
but
I
think
we're
changing
it
for
the
biggest
topologically
equivalent
outs
on
proxy.
Basically,
if
you
can
enforce
this
important
to
educate
offers,
eight
and
through
you
know,
network
topology
or
whatever
they
don't
have
to
have.
Ok,.
C
I
I
I
C
B
I
A
Okay,
yeah
more
things
about
term
base
mark.
Do
you
have
anything
else.
F
So
I
still
have
on
my
list
to
do
a
review
for
us
and
to
do
all
of
you
for
a
term
of
the
point
of
view
worst
undies,
so
I
suffered
with
Eterna
wanna,
and
this
is
big.
Spec
and
I
I
stopped
avoiding
to
develop
ER
times
to
be
starts.
Oh
yeah,
I
finish
what
I
will
get
back
to
take
from
yeah.
I
will
restart
my
my
review
and
I
will
send
or.
A
F
So
I
found
one
thing
which
I
didn't
like
is
a
new
term
base
and
the
problem
is
that's.
Why
turn
on
a
few
months
somewhere?
Oh
man,
I,
didn't
take
it
with
me.
So
and
I
don't
remember,
because
this
was
few
months
ago.
I
didn't
find
anything
very
important
studies
part.
So,
but
it's
normal
stuff.
This
is
this
is
aligning
the
way
they
are
doing
things
to
the
way
we
are
we'll
fix,
a
unique,
and
now
you
need
to
use
charcoal
develops.
Not
small
yeah
I
mean
I'm.
Sorry
I
should
have
taken
this
with
me.
A
Okay,
how
do
you
sort
of
vision
and
ebook
to
take
electronic
notes
just
kidding
thanks
a
lot
for
everyone,
and
you
know
I
mean
we
are
looking
forward
to
receiving
it.
Alan
and
Peter
couldn't
be
here.
Unfortunately,
that's
why
I
mean
it's
a
bit
difficult
to
drag
the
discussions,
but
anyway
I
think
this
input.
This
is
what
they
need.
Actually,
okay,
so
anything
anything
else
well,
I
mean
just
to
be
clear,
like
you
were
mentioning
before,
like
you
know,
getting
more
drugs
through
and
stuff
I
mean
to
be
clear.
A
We
are
trying
to
close
this
working
group.
So
in
that
sense,
if
somebody
wants,
you
know
to
publish
more
drugs
or
things
that
we
should
be
doing
more
work,
I
mean
now
it
would
be,
but
now
I
mean
drop
us
an
email
or
on
the
mailing
list
so
that
we
analyze
that,
because
I
mean
to
be
honest
at
the
level
of
energy
is
not
the
best.
So
so
we
want
to
avoid
just
you
know,
adding
staff
and
generating
drugs
that
they
take.
A
I
Yeah
something
then
another
word:
a
number
of
suggestions
are
the
working
group
of
potential
extensions
and
that
time
is
now,
let's
get
the
visions
of
the
first
sex
up
first,
but
I
think
there
was
sort
of
some
thought
that
you
might
go
back
to
those
things
afterwards.
There's
usually
I
feel
like
there's
often
more
energy,
for
you
know
a
queuing
new
things
and
getting
old
things
working
well
commercially.
If
there's
somebody
okay,
yeah.
A
It
could
be
not.
My
concern
is
exactly
I
mean
energy,
so
so
what
I
would
like
to
avoid
these?
Like
the
chairs?
Now
we
go
and
try
to
generate
some
energy
I
mean
this
would
be
actually
people
that
are
ready
to
spend
cycles,
and
if
they
are
I
mean
we
are
more
than
happy
and
I
mean
we
would
need
to
talk
to
Spencer,
of
course,
but
yeah.
A
L
Where
it's
google,
we
have,
we
have
working
groups,
for
we
have
our
to
see.
Well
still,
it
seems
like
we
have
size.
So
my
as
long
as
as
long
as
it's
made
clear
that
the
work
can
continue
that
basically
improvements
to
turn
can
happen
and
that
there
is
a
place
for
them.
It
doesn't
it's
not
important
to
me
which
working
group
they
happen
in,
but
I
have
at
least
felt
maybe
I'm
wrong
that
sometimes
a
working
group
closes
and
and
people
just
say:
well,
that's
it!
A
Process
so
to
be
clear,
I
don't
care
either
I
mean,
let's
focus
on
doing
the
right
thing.
If
it's
you
know,
ad
sponsor
is
a
different
working
group
is
to
keep
this
alive.
It
doesn't
matter
so
just
concentrate
on
my
point
on
is
their
energy
and
once
there's
energy,
we
will
do
the
magic
admin
support
so
that
it
happens.
So
exactly
these
more
about
whether
this
something
that
needs
to
be
done
and
we
will
decide,
you
know
how
and
what
and
when
for.
H
K
It's
having
real
implementations
going
to
add
energy.
You
think,
because
we
have
a
few
things
we
do
with
stunning.
We
would
like
to
document
and
send
to
the
working
good
you
haven't.
We
have
done
that
in
in
the
Sun
trace
regard
where
we
use
ICMP
messages
and
get
Kiba,
and
we
have
working
implementation
and
getting
feedback
from
that.
That
we
would
like
to
share
with
you
good
problem.
Is
that
getting
that
into
browsers.
A
Telescopes
really
I
mean
right
and
that's
why
we're
in
the
idea,
because
this
is
kind
of
networking
paradise
in
that
sense,
you
can
talk
to
people
from
browsers.
You
can
talk,
so
it's
not
really
talking
to
us.
We
can
help,
obviously,
but
the
idea
is
to
form
a
community
that
you
know
decides
that
this
is
important,
so
so
yeah,
that's
why
we
are
here.
We
as
everybody.