►
From YouTube: IETF105-NETRQMTS-20190724-1330
Description
NETRQMTS meeting session at IETF105
2019/07/24 1330
https://datatracker.ietf.org/meeting/105/proceedings/
A
C
E
C
D
D
C
E
C
How
about
now
excellent?
Okay,
the
other
thing
two
things
before
I
get
started,
putting
my
chair
hat
back
on
I'd
like
to
remind
remote
participants
that
there
is
a
virtual
queue
and
they
can
get
in
it
and
to
put
questions
in
the
jabberin
appreciated
by
the
word
mic.
That
would
be
helpful
and
then
the
second
thing,
personal
bit
of
administrivia,
while
I
am
just
so
I,
am
employed
by
the
Internet
Society.
C
D
C
C
These
days,
I
think
I'm
mostly
anyway,
so
basically,
the
objective
of
this
bop
and
we'll
talk
a
little
bit
more
about
how
we
got
to
this
stage,
but
was
to
gather
community
input
on
the
network,
because
you
know,
we've
been
building
this
network
and
it's
evolved
a
lot
over
time.
The
last
time
we
really
had
a
serious
requirements.
C
Discussion
was
about
ten
years
ago,
and
we
developed
the
beginnings
of
a
requirements
document
that's
up
on
the
website
and
we
haven't
really
revisited
it
since
then,
and
in
Prague
we
had
sort
of
a
a
knock
retreat
where
we
got
together-
and
we
said
well
they're
all
of
these
trade
offs
that
we're
trying
to
make
as
a
team
and
it's
difficult
to
make
those
those
trade
offs
without
understanding
what
the
community
really
wants.
Because
of
course
everything
comes.
C
Everything
is
a
balancing
act
between
you
know,
complexity
and
functionality,
simplicity,
cost
resourcing,
and
so
we
wanted
to
get
to
gather
some
feedback
from
you
all
I'm,
what
your
expectations
were
and
how
much
you
are
willing
to
put
up
with.
So
this
is
that's
the
whole
point
of
this
Boff.
The
one
thing
I
did
want
to
talk
briefly
about
was
basically
the
how
we
got
here.
C
C
But
then
what
you
get
into
and
that
sort
of
a
model
is
it's
a
one-shot
deal
each
time
for
the
host,
and
so
every
host
is
coming
in.
It's
a
big
risk
item
and
it
is
a
there's,
not
a
ton
of
benefit
to
that,
and
people
that
you
know
build
and
deploy
venue,
types
of
networks
or
event
types
of
networks.
Those
networks
are
a
little
bit
different
than
your
average
enterprise
network.
C
It
has
a
different
sort
of
concet
of
constraints,
a
different
way
that
you
build
it,
and
so
over
the
years,
some
volunteers
basically
started
advising
the
host
team,
and
so
you
saw
some
of
the
same.
You
know
the
network
team
being
the
host
and
the
host
equipment
supplemented
by
some
number
of
volunteers
and
then
over
time.
C
As
there
was
a
period
of
time
there
were
the
IHF
struggled
to
get
hosts
even
and
the
hosts
that
they
could
get
were
not
really
interested
in
the
risk
and
the
cost
associated
with
deploying
the
network,
and
so
it
became
a
volunteer.
Only
team
was
scrounged
gear
and
what
this
actually
means
is
we
would
you
know
before
any
event
before
any
I
upcoming
IETF.
C
We
would
have
to
go
and
see
if
we
could
beg,
borrow
or
steal
gear
get
shipped
in
to
wherever
we
were
going
to
be,
and
we
would
start
building
the
network
on-site,
and
this
had
burying
levels
of
success
over
time,
and
so
roughly
I
don't
have
ten
twelve
years
ago.
I,
don't
remember
exactly
how
long
ago
we
had
two
major
things
that
change.
First
of
all,
we
got
some
major
equipment
donations
and
the
second
thing
was:
is
we
started
contracting
out
a
lot
of
the
the
deployment
of
the
network
itself?
C
C
You
know,
you
start
developing
some
tools
for
how
you
manage
that
going
forward
and
then
a
few
years
ago
you
know
we
had
a
meeting
that
kind
of
went
sideways
and
we
we
did
a
little
bit
of
introspection
and
we
decided
that
you
know
that
this
network
has
gotten
big
enough
and
complex
enough
and
operational
enough
that
we
want
to
look
at
staging
and
more
standardized
operational
practices.
So
we
moved
to
a
model
of
the
volunteer
contractor
team
with
advanced
staging.
So
now
this
set
of
gear
that
we
have
gets
shipped
back
to
Portland.
C
You
know
our
racket
sets
up
and
maintenance
is
done
between
time
between
events
and
testing
is
done.
Some
development
is
done
and
also
we
built
a
whole
set
of
management
tools
and
a
whole
set
of
tools
for
configuring
devices,
and
we,
you
know,
operationalized
a
lot
of
that
and
use
a
lot
of
best
practices.
C
So
the
model
has
changed
over
time
and
and
the
resourcing
required
to
do
that,
and
because
the
other
side
of
this
being
that
the
team
is
not
just
what's
what's
contracted
it's
it's
a
real
blended
team
of
people
from
the
various
numbers
of
organizations
and
trying
to
make
sure
everybody
gets
the
support
they
need.
So
that's,
basically
the
administrative
side
of
how
the
network
is
built
and
then
the
next
thing
I
want
to
talk
about
briefly
was
documenting
requirements.
C
C
This
is
a
danger
with
notifications,
they
come
and
you
know
don't
really
month,
I
mean
it's
have
a
hard
time,
ignoring
them.
So
there's.
Basically,
the
the
network
requirements
document
has
three
sets
of
target
audiences.
One
is
the
external
provider.
So
it's
you
know
what
speed
of
the
links,
whether
the
quality
of
the
links
that
we
need
to
be
in.
How
many
do
we
need
to
bring
in
for
an
event?
How
many?
How
many
do
we
bring
in
what
kind
of
diversity
they
have?
C
So
that's
all
about
what
networks
we
bring
in
the
second
set
of
requirements
are
really
about
venue
selection.
You
know
what
do
we
need
in
a
venue?
What
are
the
characteristics?
What
types
of
access
do
we
need
to
the
venue
in
order
to
be
able
to
build
the
network
that
we
want
to
build
and
what
you
know?
What
kind
of
relationship
do
we
have
with
the
venue
itself
and
then
the
third
set
of
requirements
are
basically
targeted
at
the
network
designers,
so
the
NOC.
C
So
what
kind
of
network
do
we
plan
to
deploy
and
what
are
the
characteristics
of
that
network
so
draft
zero?
Zero
was
basically
the
current
set
of
the
currently
documented
set
of
requirements
which
goes
back
nine
or
ten
years.
Now,
it's
not
what
we
do
actually
draft
zero.
One
is
an
update
that
represents
some
of
what
we
do
actually,
but
not
all
of
it,
and
so
we're
we're
going
to
go
from
here
today
is
worn.
It's
going
to
come
up
and
tell
you
all
about.
D
F
So
started
so
yeah,
as
Russ
said,
you
know,
we
want
some
feedback
on
how
the
network
works
for
you,
the
mode,
if
you
provide
that
it's
probably
helpful
for
you
to
know
at
what
we
actually
do
currently
and
how
we
build
it.
And
what
will
we
provide
pointed
to
your
face?
Better
Wow
finally
have
to
lift
the
microphone.
Listen,
so
we're
going
on
an
adventure
if
folk
would
like
to
follow
along,
and
there
are
some
stats
dash
board
meeting
dot.
Ietf
dog
has
some
Network
stats,
you're
free
to
look
at
those.
F
This
is
a
screen
capture
from
I
think
last
night.
So
when
we
first
arrived
actually,
even
before
we
first
arrived,
we
ship
out
what
we
call
a
scout.
It's
a
little
ubiquity
router
as
a
key,
fairly
small
and
we
ship
it
to
the
site
and
either
we
install
it.
When
we
come
out
from
the
site
visit
or
we
get
the
hotel
to
install
it
and
what
this
device
does
is.
It
starts
announcing
our
address
space.
It
allows
us
to
test
the
circuits,
make
sure
the
EGP
stuff
works.
F
F
It
also
allows
us
to
start
turning
up
the
IETF
Hotel
SSID,
which
we
use
to
sort
of
pivot
and
build
the
rest
of
a
network.
Well,
why
did
that
go
down
one
there?
We
go
that
bounced
Bromley,
so
I'm
checking
the
time,
because
I've
got
a
lot
to
get
through
for
circuits.
We
get
at
least
two
circuits,
sometimes
up
to
five,
and
these
are
almost
always
donated
by
local
providers.
So
we
have
a
bunch
of
local
providers
who
we
deal
with
and
they
provide
us
with
the
circus.
F
We
try
really
hard.
What
are
you
laughing
at
which
we
try
really
hard
to
get
redundant
connectivity,
so
two
providers,
different
providers,
two
sets
of
fibers
two
sets
of
entrances,
etc,
and
up
until
now,
we've
very
much
always
been
doing
one
gig
circus.
A
number
of
providers
have
been
offering
us
ten,
but
we
found
that
one
is
up
until
now,
sort
of
largely
sufficient
and
seems
to
be
stifling
it
slightly
more
stable
but
I
think
in
the
future,
we're
probably
going
to
be
getting
ten
gig
circuits.
F
The
robbers
we
have
can
do
up
to
ten
sorry.
After
a
hundred,
we
also
require
that
the
provider
gives
us
dual
stack.
You
know
with
IP,
v4
and
v6
and
that
they
will
speak
BGP
with
us.
So
we
can
announce
our
own
address
space.
That
way
we
can
use
the
same
address
space
where
we
go
four
core
routers.
We
currently
have
two
juniper
MX
2,
o
fours.
This
is
actually
the
first
meeting
where
we've
been
using
the
MX
to
a
horse.
F
Up
until
now,
we've
been
using
juniper,
MX
ATS,
or
at
least
for
the
last
many
years.
We
need
to
be
using
her
ex
ATS
the
convergence
time
on
those
was
quite
long
as
part
of
building
the
network.
We
build
everything
and
then,
before
the
attendees
arrive,
we
do
failover
testing.
It
was
taking
about
25
minutes
or
so
for
b
gb
to
converge
with
the
MX
80s.
Thanks
to
the
new
donations
from
juniper.
It
now
converges
in
like
a
minute
or
a
minute
and
a
half.
F
The
corridors
are
obviously
our
core
routers.
They
do
BGP
for
the
last
number
of
meetings.
We've
also
do
been
doing
our
PK
validation
and
throwing
away
invalid
routes.
They
also
support.
You
know
obvious
services
like
OSPF
53.
There
are
the
DHCP
relays.
Do
our
RA
advertisements,
we
also
try
and
general
sort
of
follow
best
practices.
So
we
do
things
like
pcp
38.
F
F
So
for
switches
we
have
a
pair
of
Cisco
Catalyst,
4500
X's,
which
we
run
in
a
stack
configuration
we
also
deploy
around
10
or
so
IDF
switches
and
40
or
so
12-foot
switches.
The
12-foot
switches
are
the
little
switches
that
we
plug
access
points
into.
If
you've
ever
sat
in
the
front
row,
you've
probably
seen
a
bunch
of
them.
We
also
use
them
for
things
like
video,
streaming,
etc.
Basically,
general
switches.
Whenever
we
need
one
we
deploy
those
all
of
these.
We
now
manage
with
something
which
we
call
Joe's
magic.
F
F
We
don't
just
sort
of
deploy
the
network
once
rooms
often
get
converted
into
things
like
penury
rooms,
at
which
time
we
have
to
move
switches
around,
etc.
If
it
wasn't
for
the
CTP
type
thing,
we'd
have
to
do
that
manually
and
it
turns
out
that
IETF
participants
often
plug
in
laptops
with
DHCP
servers
or
something
like
that
on
so
having
a
system
which
automatically
configures
itself
learns
where
up
links
are
etc,
is
really
useful.
I'm
just
gonna
go
forward
or
down
it's
gonna
go
down.
We
pick
the
right
button,
I
tested
it
before
using
that.
F
So
for
wireless,
we
have
to
Cisco
wireless
LAN
controllers,
which
we
run
an
H
a
pair.
We
also
travel
with
they're
much
smaller
little
wireless
LAN
controller,
which
we
use
both
for
testing.
We
also
often
end
up
supporting
that
an
eye
socket
visor
II
committee
meeting
after
the
IETF
or
sometimes
similar
type
meetings.
This
is
a
small
little
wireless
LAN
controller
that
we
can
leave
behind
when
we
pack
the
rest
of
the
network,
we
in
general
deploy
somewhere
between
50
and
70
access
points.
This
has
a
to
do
which
I
forgot
to
do.
F
We
have
I
think
68
access
points
deployed
for
this
particular
meeting.
Yeah
68,
plus
testing
1
sir,
and
our
access
points
do
both
5
gig
and
2.4.
We
serve
steer
people
towards
using
5,
because
obviously
it
works
a
lot
better.
The
Cisco
wireless
LAN
controllers
have
actually
been
really
good
for
us.
They've
solved
a
lot
of
the
arc
problem.
If
people
remember
the
hard
meeting
of
the
previous
one,
but
the
one
before
that.
F
The
one
before
that,
while
we
keep
going
back
to
Prague
93,
we
were
having
a
lot
of
issues
with
arc,
because
we
announced
a
huge
amount
of
public
address
space
to
the
internet
and
everybody
scanning
would
cause
ARP
requests.
This
would
cause
the
wireless
infrastructure
to
send
out
Arps,
which
would
wake
up
everybody's
phones
and
so
people
getting
served
half
a
day
or
three-quarters
of
a
day
battery
life.
F
If
your
phone
now
lasts
longer
than
3/4
of
a
day,
you
should
thank
panda
who's
over.
There
say
hi
panda.
We
also
have
a
whole
bunch
of
encrypted
SSIDs
IETF
IETF
2.4.
Only
the
ITF
nat64
v6
only
nat64
unencrypted
jerome
I
suck
it's
a
whole
long
list,
something
that
would
be
really
useful
further
on
and
this.
This
buff
is
understanding
what
people
actually
want,
which
of
these
CSS
IDs
people
need
which
of
them
are
providing
useful
service.
F
If
there
any
here,
we
don't
need
any
more
or
if
there
are
additional
services
we
should
be
providing
moving
on
to
the
guest
room.
I
have
a
bad
habit
of
calling
this.
The
hotel
networks
I've
actually
end
up
calling
it.
You
know
ITF
hotel,
I,
meaning
the
guestroom
Network,
so
it
turns
out
that
in
case
you
have
noticed
IETF
what
has
been
so
weird
like
really
weird.
F
When
hotels
build
their
guestroom
networks,
they
build
them
for
sort
of
more
normal
people.
This
means
that
they
deploy
things
I
kept
for
which
do
intersecting
and
rewriting
of
DNS.
A
lot
of
them
also
do
interception
of
HTTP
and
do
serve
managing
to
make
that
we
direct
direct
people
to
capture
portals
hotel
networks
also,
generally
do
not,
and
of
course
they
need
to
do
things
like
filter,
dangerous
sports,
like
SSH
they're,
also
built
with
a
set
of
assumptions,
which
is
that
people
will
have
a
limited
number
of
devices
use.
F
F
F
So
what
we've
been
doing
has
been
taking
over
the
guest
room
network
and
replacing
it
with
a
pair
of
ubiquity
routers,
which
you
call
the
hotel,
routers
and
the
spy
passes
the
hotel's
sort
of
Hotelling
solution,
they're
captive
portal.
All
of
that.
In
many
cases
we
actually
just
get
a
VLAN
before
there
there
Hotelling
solution
and
we
turn
up
the
IETF
Hotel
SSID
and
we
generally
provide
internet
access
for
everyone
in
the
hotel.
We
believe
free
internet
for
all,
something
that's
interesting
and
people
probably
haven't
run
into
this
is
many
hotels,
have
somewhat
inventive
architectures.
F
F
F
Unfortunately,
if
your
hotel
room
is
anywhere
near
the
elevators,
the
access
point
arrives
and
you're
like
there
is
a
good
signal
and
then
you're
like
go
also:
hotels
have
billion
bent
of
physical
plants,
so
we've
seen
Ethernet
over
coax
and
I,
don't
mean
DOCSIS
I
mean
like
a
special
little
box,
which
just
runs
Ethernet
over
coax
and
something
proprietary
in
the
middle.
Of
course.
You've
also
seen
Doc's
SDSL,
you
know
cat
3
cable.
F
F
We
also
run
a
large
number
of
services
in
order
to
support
the
network.
So
we
have
three
physical
servers
and
three
other
servers
which
we
use
to
build
and
run
the
network
we've
been
using
ganodi
and
docker
to
allow
us
to
sort
of
Bob's.
Looking
concerned:
ok,
you're,
looking
you're
looking
pensive,
so
we
provide
a
bunch
of
services.
Dns
DNS
SEC,
we've
also
been
doing
recently
DNS
over
TLS.
F
F
We
have
been
trying
relatively
hard
over
the
last
couple
of
meetings
to
move
to
more
of
sort
of
a
DevOps
model,
so
we
can
design
and
build
things,
commit
it
in
something.
I
could
get
repo
and
then
be
able
to
rebuild
and
restage
things
instead
of
having
sort
of
long-running
servers
which
we
have
to
keep
maintaining
monitoring.
We
also
have
a
whole
slew
of
monitoring
things.
We
seem
to
often
end
up
experimenting
with
a
new
monitoring
thing
and
then
migrating
to
another
one,
and
we
haven't
always
managed
to
to
deprecated
the
older
ones.
F
I
put
the
slide
in
because
I
knew
people
have
questions
before
we
leave.
We
back
up
our
servers,
but
before
we
do
that,
we
may
sure
that
we
scrub
all
PII
from
the
servers
before
yeah
before
we
pack
them
up
and
also
that,
while
way
that
way,
while
they're
being
shipped
around
Oh
personally
identifiable
information,
so
you
know
logs,
DHCP,
logs,
etc.
Thank
you
one
of
the
main
purposes
of
the
network.
It
used
to
be
just
throw
four
participants
these
days.
It's
also
to
support
remote
participation.
F
We
actually
have
a
separate
VLAN
for
the
meet
echo
and
streaming
streaming
traffic.
We
also
provide
a
number
of
VMs
for
the
meet
echo
system
when
there's
sort
of
common
or
like
interesting
sessions,
it's
currently
around
60,
Meg
or
so
of
traffic
is
being
streamed
from
the
VMS
up
to
the
internet.
It's
around
60
mega
or
so
for
a
much
participation
traffic.
F
As
I
said
the
network
we
build
sort
of
makes
that
remote
participation
possible
meet
echo
and
Cascadian.
The
people
who
do
the
video
streaming
have
actually
done.
Events
on
sort
of
regular
venue
networks
before
but
they've
only
managed
to
do
streaming,
not
remote
participation
from
talking
to
meet
echo.
They
think
that
the
limited
bandwidth
map
firewalls,
like
a
v6,
would
prevent
them
from
doing
good,
remote
participation
and
Cascadian
folks
said.
Basically,
the
hotel
network
just
won't
work
for
us,
so
I've
gone
through
a
bunch
of
stuff
really
quickly.
F
You
deserve
a
kitten
for
having
sat
through
all
of
this.
Here's,
your
kitten
while
that
was
well
done.
Another
purpose
of
the
ITF
network
is
sort
of
experiments
or
eating
our
own
dog
food.
We
do
these
experiments
carefully,
obviously,
and
you
actually
need
to
get
them
approved
by
the
IETF
chair.
So
these
aren't
just
things
which
we
we
do
easily.
As
an
example,
we
ran
one
of
the
very
early
DMS
ro
TLS
services.
F
F
Another
experiment
needed
fairly
recently
was
v6
own
v6
only
and
by
that
we
mean
real
v6.
Only
we
put
the
v6
ups
and
six-man
meetings
in
a
separately
served
separate
area
of
the
hotel,
and
then
we
turned
off
I
PV
for
on
all
of
the
radios
near
them,
so
that
the
only
access
that
was
in
those
rooms
was
v6
much
hilarity
ensued.
It
turns
out
that
a
lot
of
devices
and
a
lot
of
things
become
sad
if
there's
absolutely
no
v4
network
available
to
them
at
all.
F
F
This
was
something
where
at
IETF
100
bunch
of
people
in
Toronto
in
the
hackathon
and
did
an
inventory
of
a
lot
of
different
applications
and
different
operating
systems
to
figure
out
how
things
have
I,
with
net
6
for
a
number
of
bugs
were
found.
A
number
of
things
actually
got
fixed
I
believe
that
Spotify,
for
one
at
least
fix
their.
Let's
explore.
Processing
bunch
of
things
got
got
made
better
from
this
final
one,
just
because
we're
running
here
on
time,
Mac
randomization.
F
This
was
something
that
we
did
in
sort
of
cooperation
with
the
I
Triple
E.
In
order
to
improve
privacy.
There
was
a
I
Triple,
E
sort
of
effort
to
see
what
would
happen
if
devices
randomized
their
MAC
addresses,
potentially
at
a
fairly
short
timescale,
and
we
did
this
on
our
network
to
understand
what
would
what
the
implications
in
this
some
switch,
architectures,
DHCP
servers,
router,
nd,
caches,
etc
would
be
Carlos.
F
Who
was
that
Carlos
Carlos
probably
has
more
info
and
not
Joe
keys
in
the
room
and
I.
Think
I
have
like
two
minutes
for
clarifying
questions,
but
there
is
going
to
be
a
whole
session
to
the
Sun.
You
know
which
of
these
services
are
important.
What
do
we
need
to
do
that?
We
don't
etcetera.
You
can
ask
now,
if
you
want
to,
but
I
did
it
faster
than
expected.
F
Yeah
I
mean
a
number
of
these.
Things
are
just
for
our
own
internal
infrastructure,
like
you
know,
get
repos
and
syslog.
Those
are
just
for
us,
but
one
of
the
things
I
should
have
mentioned
on
the
slide.
Is
there
are
a
lot
of
services
here?
You
know.
If
there
are
some
of
them,
we
didn't
need
that
would
be
really
good
to
know
and
that
we
can
turn
on
and
off
and
make
our
lives
easier.
F
I
can
say
so
what
yeah,
so
what
Clemens
said?
Is
we
once
that,
we've
in
recent
memory
we
had
once
buy
local
loops?
The
bandwidth
was
still
still
donated,
but
pretty
much
every
other
time
the
provider
brings
the
circuit
or
the
way
to
us
drops
it
off,
gives
us
bandwidth,
Suites
BGP,
you
know
we
put.
Their
name
are
coming
slides
this
last
time
it
will.
This
current
time
was
tell
us
and
micro
optic,
and
they
we,
you
know
kindly
gate
open,
gave
our
circuits.
So
look
us
to
us.
Okay,.
G
So
to
answer
the
question
on
circuits,
the
truth
is
that
we've
basically
gone
to
friendly
in
friendly
folks
to
get
circuits
throughout
the
years
and-
and
frankly,
it's
been
done
through
friendships
through
folks
who
are
in
the
area
that
we
have
relationships
with,
but
there
there's
always
been
money
set
aside
in
the
budget
to
be
able
to
handle
the
contingency
in
case.
Something
went
wrong.
That
said,
we
that
we
have
had
scenarios
actually
in
Korea
we
had
to
buy
circuits
as
well
as
local
loops
and
the
last
time
we're
in
San
Francisco.
G
F
F
H
There's
like
stable
electricity
and
water
in
your
room
right,
and
so
those
networks
tend
to
have
gotten
quite
a
bit
better
and
some
of
the
anecdotes
may
be.
You
know
a
bit
dated
in
sort
of
holding
us
in
your
mental
state
here
so
I
suspect,
like
when
I
travel,
most
places
I,
you
know,
can
get
great
access
and,
if
I'm
concerned
about
getting
services,
I
pop
up
a
VPN
I
suspect
that
that
might
be
a
viable
solution
to
me.
Yeah.
F
The
thing
is
women
I
mean
that
Tiffany
a
point.
We
should
discuss
that
further.
One
of
the
things
to
keep
in
mind,
possibly
is
when
you
travel
you're,
generally,
not
traveling
with
1200
other
ietf
to
take
people
who
have
a
different
use
case.
I
mean
sometimes
obviously,
sometimes
that's
not
true.
There
are
other
places
where
there
are
lots
of
usage,
fair.
I
I
can
comment
on
that
from
having
run
some
of
these
yeah
there's
a
number
of
funky
issues
with
the
architecture
of
those
networks,
as
opposed
to
just
say
their
raw
capacity
or
the
size
of
their
NAT
box
or
their
residential
gateways.
So
things
like
configuration
issues,
this
network
has
a
limit
on
the
number
for
the
hotel
network.
Here
has
a
little
bit
on
the
number
of
broadcast
packets
it's
prepared
to
see,
and
so,
when
there's
800
devices,
which
is
significantly
more
than
say,
they
normally
see
a
normal
hotel
night.
D
I
I
And
so
what
I
would
say
is
that
I
think
the
assertion
that
these
networks
are
getting
better
is
actually
pretty
accurate,
they're,
not
necessarily
as
good
as
the
expectation
that
is
currently
placed
on
the
hotel
network.
I
expect
that
over
time
personally
that
they
will
get
there.
But
certainly
that
is
not
the
case
today.
I.
J
:
JS
I'm
not
coming
I,
do
go
to
other
conferences,
full
of
technical
people
or
the
hotel
networks
generally
work,
and
when
they
don't,
everyone's
cell
phone
continues
to
work.
Is
it's
not
as
big
a
deal
as
it
was
10
years
ago?
I,
look
I,
understand
that
we
have
an
expectation
for
v6
and
that's
probably
often
not
delivered,
but
to
get
our
job
done,
which
mostly
involves
sending
email
and
getting
to
a
webpage.
You
don't
actually
need
that.
I
want
to
ask
a
question
about
the
bandwidth.
J
K
J
J
F
J
Yeah,
actually
you
there's
somewhat
different.
You
guys
keep
saying:
we've
been
bandwidth
on
strains,
I
feel
very
differently.
I
feel
that
the
bandwidth
requirements
are
causing
us
to
not
be
able
to
go
to
many
venues
we
could
otherwise
go
to
so
I
feel
we
have
been
bandwidth
constraint.
So
that's
just
my
way
so.
C
L
So
one
one
clarification
I'd
like
to
point
out
here
is:
we
do
have
really
the
two
networks.
As
you
said,
we've
got
the
guestroom
network
and
we've
got
the
network
at
large
and,
as
we
have
this
conversation,
I
think
it'd
be
really
really
good
for
us
to
keep
very
focused
on
when
we
say
well,
I
really
need
this
massive
stuff
visibility
and
we
have
1,200
people.
That's
true.
We
have
people
in
the
get
or
the
meeting
space.
We
do
not
have
1,200
people
in
the
guest
room.
L
It's
and
a
lot
of
people
are
off-site
in
fact,
I
rough
numbers.
We
do
let
about
600
people,
typically
in
a
hotel,
I
guess,
but
it
varies.
That
means
well.
Half
our
attendees
are
actually
not
staying.
Ti2
hotel.
There
are
someplace
else
that
not
these
guests
food
network.
So
keep
that
in
mind.
As
we
have
this
conversation
boy,
it
really
really
need
and
what
you
baby
don't
need.
J
Yeah
charity
again
real
quick,
something
I
said
before
about
about
the
ipv6
on
the
guest
or
I'm.
Sorry
on
the
guestroom
network
is
when
we
tried
to
do
that
once
I
think
it
was
that
nod.
We
ended
up
with
all
of
the
broadcast
traffic
on
v6
actually
going
and
taking
down
the
hotel,
Wireless
and
the
regular.
K
J
So
the
other
comment
about
bandwidth,
I
know:
I
saw
Jim
go
in
and
out
of
the
queue
you
know,
since
most
of
the
band
list
is
donated
or
functionally
free
to
the
IETF
I,
don't
see
a
reason
to
limit
that,
and
also
even
if
we
had
to
commercially
get
it
the
cost
per
bit
is
so
low.
You
know
commercially
available,
it's
you
know.
Let's
say
you
know
less
than
a
buck.
A
megabit
I
think
is
a
reasonable
number
and
if
you
actually
know
what
the
prices
are.
Oh
it's
money.
N
Were
done,
yeah
Clemence
sets
me
clément
shop,
I'm,
also
part
of
the
not
cede
my
responsibilities
are
getting
bits
into
the
building
and
also
stuffing
those
bits
back
into
the
booty
Network.
So
most
of
the
topics
we've
discussed
so
far
touch
mic.
My
my
responsibility
again,
I
have
to
concur
with
Jared
that
getting
bits
or
getting
lots
of
them.
It's
not
a
problem.
In
fact,
we've
been
turning
down
people
offering
us
10g.
N
So
this
is
this
does
not
limit
our
venue
selection,
with
one
exception
that
I've
recently
heard
of
that
I
don't
know
if
I
can
discuss
a
couple,
it's
it
sometimes
it's.
It
depends
on
a
country
and
it's
not
a
matter
of
how
many
bentwood
but
getting
been
put
into
the
thing
in
general,
like,
for
instance,
in
China
or
in
that
air
in
that
region.
But
if
we
get
stuff
or
if
we
get
bits
into
the
building,
it
really
doesn't
matter
if
it's
one
gear
ten
gear
whatever.
N
N
N
C
N
O
F
It's
not
a
huge
amount.
I
can
answer
that
other
than
people
have
been
upgrading.
So,
for
example,
last
time
we
were
in
the
Sun
hurt
in
this
hotel,
there
was
a
very
different
wireless
infrastructure
to
now,
so
hotels
are
actually
upgrading
their
infrastructure.
It's
becoming
a
lot
more
reasonable,
sometimes
actually
kind
of
driven
by
us.
You
know
we
suggest
really
strongly
that
we're
more
likely
to
come
back
if
the
network
was
made.
F
C
So
I
had
a
couple
quick
notes
before
we
start
down
this
presentation.
The
first
one
was:
we
didn't
actually
do
a
really
good
job
early
on
is
saying.
You
know
we
do
site
qualifications
Clemens
got
into
that.
He
does
a
lot
of
the
work
for
sy
qualifications.
Jim
had
done
a
lot
of
that
work,
so
we
do
as
part
of
the
venue
selection
process
go
and
check
out
the
hotels.
So
the
second
thing
is
we.
We
have
tried
to
level
up
to
a
set
of
user
facing
requirements.
C
Some
of
the
detailed
stuff
we
expect
to.
We
could
have
further
conversations,
possibly
on
a
mailing
list
and
we'll
get
to
that
at
the
end,
but
I
do
think
we
are
trying
to
get
here
as
a
team.
We
would
like
to
simplify
where
we
can
so
we're
trying
to
make
our
own
jobs
easier,
so
we're
not
so
I'd
like
to
put
that
out
there.
First.
As
you
know,
we
are
looking
to
simplify
and
then
the
second
thing
I
want
to
put
out
there
is.
C
C
So
a
couple
basic
assumptions:
we've
already
talked
about
this:
we're
going
to
focus
on
user
requirements,
we're
not
here
to
design
the
network.
Second
thing
is:
we
do
have
two
sets
of
requirements
we're
going
to
go
through.
There
is
a
difference
between
the
venue
network,
which
is
in
all
the
meeting
rooms
and
those
kinds
of
things
and
the
guest
room
network,
and
we've
already
alluded
to
that
in
some
ways,
and
so
we
want
to
focus
on
that.
So
question
number
one:
we
need
a
network
right
eye
health.
E
P
C
C
So
do
you
think
the
network
provided
by
most
venues
with
without
any
enhancements
is
sufficient,
and
that
probably
means
no
PFI
pv6
and
it
probably
has
gnats
and
captive
portals
and
what
we'll
do
is
we'll
have
a
little
bit
of
discussion
and
then
we'll
try
and
do
a
home
at
the
end
of
each
of
these
questions,
I
wanted
to
use
a
slider
app,
but
they
wouldn't.
Let
me.
J
Turn
Jared
much
previous
professional
traveler
now
currently
semi-professional
traveler
in
in
most
cases,
including
at
large
conferences
over
the
years
I
found
that
the
venue's
have
generally
upgraded
to
bandwidth.
You
know
it's
dependent
upon
the
hotel
chain
and
whatnot,
but
for
the
most
part,
if
it
works
and
even
even
in
the
technical,
you
know
the
more
technical
places
where,
like
you
know,
an
an
OGG
or
an
errand
or
write,
or
something
like
that.
Q
My
name
is
Tim,
Shepherd
I've
been
do
an
awful
lot
of
IETF
meetings
and
my
travel
plans
often
have
me
staying
over
Friday
night
and
then
flying
out
either
Saturday
morning
or
some
day
Saturday
afternoon,
or
maybe
go
on
Sunday
and
usually
at
some
point
on
Friday
the
hotel
Network
reverts
to
the
hotel's
network
and
in
my
experience
at
least
half
the
time.
It's
pretty
sad
and
it's
not
though
I
don't
believe,
is
that
the
hotel,
lock
network
lacks
capacity
to
and
from
the
internet.
Q
It's
that
the
captive
portal
system
isn't
prepared
to
deal
with
the
numbers
of
people
who
are
wanting
to
use
it
and
that's
even
when
probably
more
than
half
the
IKEA
people
are
very
much
and
I
usually
do
manage
to
get
on,
but
I
usually
have
to
try
several
times.
Sometimes
it
has
kind
of
just
worked,
but
at
least
half
the
time
it
doesn't
just
work
and
so
I
think
it's
the
captive
portal,
and
it
may
also
be
that
they
don't
have
enough
internal
DHCP
addresses
to
hand
out.
Q
So,
even
if
you
didn't
even
get
to
the
captive
portal
or
after
you
get
to
the
captive
portal
and
wants
to
do
another
DHCP
something
like
this
I've
tried
to
debug
it
and
it's
a
different,
weird
story
and
every
other
new
and
I
think
you
know.
Maybe
we
don't
need
to
bring
our
own
network
into
the
IHF,
but
I
think
we
have
to
get
there
captive
portal
out
of
the
way
and
make
sure
you
know.
Q
J
Eggs
I
think
for
the
most
part,
I
find
that
what
I
want
to
be
able
to
do
is
browse
the
web
and
do
email
and
that
those
work
fine
through
all
the
existing
hotel
networks,
maybe
yeah
ninety-eight
percent
of
time
at
this
point,
I
do
think
it's
worth
asking.
Are
there
specific
things
that
that
wouldn't
work
for
us
in
that,
and
then
the
check?
Oh
and
remote
participation
was
brought
up
earlier.
J
I,
don't
know
what
the
status
of
those
products
is,
but
I
know
that
you
know
all
the
web
RTC
based
conference,
major
every
vendors,
major
conferences
as
soon
as
web
RTC
based
works,
fine
through
v4
or
v6,
with
or
without
Nats.
At
this
point
in
time,
so
I
I,
don't
think
that
that
should
can
continued
to
be
a
problem
at
this
point
in
time
or
really
any
time
in
lost
for
five
years.
S
Melissa
Cooper
:
stole
what
I
was
gonna
say,
but
I
think
it
would
be
useful
if
people
can
speak
to
you
the
things
that
they
would
not
be
able
to
do
that
they
need
to
do
get
their
work
done
during
the
IDF's
week
under
the
different
scenarios.
Not
just
like,
would
it
be
better
like?
Does
it
make
you
sad
like,
of
course
the
captive
portal
makes
everybody
very
sad.
H
10:30
and
I
think
the
good
news
is
the
rest
of
the
world's
requirements.
Have
slowly
crept
up
well,
ours
have
kept
pretty
stable
over
time
and
the
result
of
that
is
we're
much
more
often
in
conditions
that
are
close
to
what
on-site
attendants
are
used
to
in
other
places.
They
remember
this
is
not
the
case.
Mike
st.
John's
may
remember
the
size
of
the
circuit,
but
I
seem
to
remember
that
when
we
went
to
Stockholm
the
first
time
they
actually
doubled
the
capacity
coming
into
the
country
in
order
to
deal
with
the
ietf
yeah
it.
H
So
it
was
like
56
kbps
additional
line
was
impressive
at
the
time
right,
so
they
doubled
the
country's
capacity.
Nobody
said
we
weren't
Johnny,
Appleseed
or
Johnny
applesauce.
You
know
the
capacity
and
had
that
make
some
more.
The
good
news
is
on
site.
I.
Think
we're
actually
pretty
close,
but
I
think
the
point
that
Cullen
made
obliquely
is
much
more
important
and
that
is
the
ability
to
have
remote
participation.
H
So
it
actually
suggests
that,
rather
than
focusing
on
local
participation
that
the
NOC
team,
as
it's
looking
at
these
focus
on
what
it
would
take
to
make
remote
participation,
work
or
double
the
numbers
we
have
now
and
make
sure
that
if
we
can
reach
that
without
hotel
network,
we're
good
to
go
but
I'm.
Guessing
that
if
we
can't
the
delta
between
what
we
need
to
do
to
make
the
remote
participation
work
and
what
we're
doing
now
for
local
participants
isn't
high
enough
to
work.
D
C
Before
you
say
something
I
just
wanted
to
from
Mike
from
the
jabber
room
to
answer
Colin's
comment-
and
this
is
from
the
Medeco
team
to
answer
Colin's
comment.
I
do
agree
that
remote
participation,
clients
can
work
easily
through
Nets.
Here
at
the
IETF
we
deployed
remote
participation
servers
at
the
venue
in
order
to
minimize
the
delay
and
make
it
effective
and
having
a
WebRTC
server
behind
in
that
is
far
from
ideal.
In
my
humble
opinion,.
K
L
Funding
of
two
points:
first
I
already
noticed
that
in
this
conversation,
we're
having
we're
muddling
muddling
together,
guest
rooms
and
the
meeting
space,
and
that's
we're
gonna,
try
to
capture
the
notes
from
this
session
and
pull
out
meaningful
like
what
people
really
need.
That's
gonna
make
it
really
hard,
so
I
think
we
should
all
be
really
really
careful
when
we
say
well,
you
know:
hotel
network
went
down
the
idea
found
what
went
down
at
the
end
of
the
video
well,
you're,
actually,
you're
medium
doing
work,
that's
the
guestroom
network.
L
We
were
talking
about,
and
I
also
want
to
echo
I,
really
like
the
suggestion
that
we're
talking
about
what
we
really
need
in
terms
of
getting
work
done.
So
what
I'm
sitting
here?
It's
really
awesome.
If
I
consider
and
watch
Netflix,
it's
not
an
easy
requirement,
it's
actually
probably
better.
If
I
can't
watch
that
foot
since
I'm
gonna
pay
attention,
but
email
is
a
critical
thing
you
need
to
do
while
you're
here
surf
web
pages
to
get
information.
Kristef
hear
the
conversation.
It's
critical.
J
There
are
problems
with
the
captive
portals
so
that
we're
gonna
have
to
probably
go
to
the
hotels
and
either
vet
those
ahead
of
time
that
make
certain
they
can
handle
the
capacity
we're
going
to
throw
at
them.
Or
we
need
to
speak
to
the
hotels
and
say:
can
we
bypass
these
for
the
moment
and
and
figure
out
some
way
that
we
get
the
capacity
that
we
need.
R
R
Has
to
say
it
will
we
will
I
think
except
maybe
I've
heard
in
Stockholm.
You
can
get
ipv6,
you
know
hotel.
We
will
never
see
this
enemy
and
you
know
I
think
it
among
others.
I
mean
I,
actually
use
it.
Many
other
people
actually
use
it.
The
message
the
ITF
would
be
sending
to
the
larger
community
that
well,
we
don't
even
think
it's
important
anymore
I
mean
might
as
well
just
roll
this
up
and
stop
so
I
think
this
is
very
important
that
we
continue
doing
this
and
building
networks
that
support
both
protocols.
R
T
Singapore
still
in
my
comment,
yeah
I
just
cannot
get
my
work
done
without
any
issues.
I'm
sitting
on
v6
on
your
network,
all
the
time
so
I'm,
not
sure
you
do
not
need
to
get
I
could
before
to
get
my
work
done.
But
if
you
want
opinion
here,
some
people
do
need
three:
six
as
a
primary
product,
so
I
don't
see.
Thank
you.
They
can
you
network
without
any
enhancements,
can
do
that.
I
I
don't
think
we
shall
be
using
yeah.
G
We
absolutely
found
repeatedly
that
the
way
that
those
those
captive
portals
are
set
up,
the
the
madding
that
they
do
and
the
state
that
they
maintain
for
that
Matt
doesn't
scale
to
the
ietf
size,
and
so
the
end
result
is
that,
while
you
may
have
perfectly
fine
the
available
bandwidth,
both
the
forwarding
capacity
on
the
portal,
as
well
as
the
state
capacity
for
nap,
are
exhausted
and
you
get
some
really
weird
failures
and
what
we
found
in
a
number
of
scenarios.
Actually
we've.
U
U
V
David
Sommers
Harris,
so
I
had
two
comments.
One
is
just
contributing
to
well,
you
know
what
do
we
need
list
it's
my
my
story
is
that
I
joined
the
ITF
kind
of
just
to
learn
and
find
out
what
it's
all
about
and
then
got
inspired
by
the
whole
ipv6
story.
That's
been
going
on
and
then
my
first
reaction
was
well.
What
about
mine
company?
Does
my
stuff
work
over
ipv6?
V
Try
you
know
test
the
corporate
VPN
test,
our
phone
apps
on
various
and
if
we
had
been
using
just
the
hotel
network,
I
wouldn't
have
been
able
to
do
that
and
I
wouldn't
have
been
able
to
start
those
conversations
with
the
engineering
teams
within
my
company
to
say:
hey.
Why
don't
we
do
this
I
think
that
is
one
of
the
objectives
that
the
ITF
is
not
just
writing
drafts,
but
also
I'm
moving
the
wider
community
to
adopt
the
implementations
of
the
Giants.
So
that
was
first
comment.
V
The
second
one
is
I
think
the
scope
of
the
the
users
requirements,
maybe
shouldn't
be
limited
to
just
you,
know,
obviously,
networks
mainly
from
getting
work
done
at
the
conference,
but
it
does.
It
has
other
impacts
too.
Like
me
right,
you
talked
about
how
Prague
hotel
actually
probably
upgraded
infrastructure
because
of
our
networking
background
requirements.
So
it
has
this
kind
of
impact
on
implementations,
but
also
its
messaging.
Like
Bob
said
like.
If
we
drop
a
BB
six
network,
then
it
kind
of
sends
a
message
that
we
don't
care
anymore.
V
H
Jason
looking
good
I
think
bearing
in
mind
this
is
about
the
meeting
room,
use
the
meeting
space.
Look
the
venue,
another
hotel
rooms,
I.
Think,
maybe
you
know
we're
kind
of
past
I,
don't
know
what
other
questions
you
have
but
I
think
kind
of
past.
You
know
I,
give
me
the
network
or
not.
You
know,
that's
a
simplistic,
binary
question
right
and
I.
Think
it's
not
really
asking
the
important
stuff
which
others
have
commented
on
a
little
bit,
which
is
what
do
you
really
need
and
how
are
those
needs
by
our
ties?
H
What
are
the
most
important
things
to
the
least
important
things
on
that
list?
Because
that's
where
I
think
all
of
the
difficulty
often
is
with
particular
venues
and
with
where
we
want
to
go
in
the
future,
for
this
network
I
think
most
of
us
would
say
like
yeah
we
need
we
have
unique
bandwidth
requirements
and
imagine
captive
what
if
we
know
cause
problems
so
I
doubt
anybody
in
the
room
here
is
gonna,
come
up
and
say
like
no,
it's
just
used
for
all
the
meeting
spaces.
U
H
C
Right
I
was
trying
to
decide
whether
it
was
worth
reminding
everybody
that
we
do
have
a
set
of
questions,
venue,
question
venue,
network
questions,
Plus,
guest
network
questions
and
I
was
hoping.
We
could
defer
that
the
guest
room
conversation
until
we
got
to
that
section
of
the
conversation
and
just
focus
on
the
meeting
space.
At
this
point
and
with
this
I
think
we're
done
with
you
to
go
I
cut
the
line
for
this
question
then
move
on
sorry.
Unless
you
have
something.
W
Like
that,
maybe
I
should
sit
down
for
the
motion:
oldest
customers
I'm,
not
specifically
what
immediately
so
I'm
willing
to
believe
that
we
have
demands
that
are
probably
beyond
what
most,
when
useful.
But
I
also
believe
that
we've,
what
we
put
in
is
way
more
than
we
actually
need,
but
I
don't
quite
know
where
we
are
in
terms
of
the
spectrum
right,
but
it
seems
like
we
try
to
like
painstakingly
create
this
perfect
network,
like
it,
with
architecture
to
the
nine.
W
Even
that
doesn't
actually
exist
in
reality
anywhere
else,
but
at
the
IHF
agrees
we
have
to
hackathon
the
quick
people-people
right,
the
guy
guy
on
jabber.
That
said,
you
know,
I
ever
I
have
a
local
server
out
vo
T
ITF
and
in
the
guard,
yellow
client
said
you
know,
I'm,
sorry,
I
can't
catch
us
yeah,
you
can
it's
it's!
W
It's
not
matter
then,
and
blew
their
mind
right,
because
that
just
doesn't
exist
anywhere
else
and
and
so
it's
nice
that
it
exists,
but
it
comes
with
a
cost
right
and
so
I'm
on
the
executive
committee
of
a
Texaco
which
runs
a
bunch
of
conferences
in
their
communications,
completely
void
of
students
and
professors.
There
are
not
nearly
as
crazy
as
we
are,
but
they're
not
normal
either
and
we
don't
pull
our
own
network,
but
we
go
to
the
venue
and
say
you
know
this
is
a
demanding
crowd.
You
might
want
to
get
some
more
bandwidth.
W
You
make
sure
you
have
enough.
Ip
addresses
to
head
out
turn
off
your
captive
portal
and
we
let
them
do
it
and
oftentimes
they
do
it.
Sometimes
they
don't
and
then
we
complain
on
the
first
day
and
they
fix
it
and
it's,
but
as
willing
to
believe
that
the
ITF
isn't
quite
there
yet,
but
but
I
don't
think
we
need
all
of
them.
What
the
hell
is
is
the
RTP.
So
everybody
know
I
don't
go
to
the
slide
here.
W
E
X
The
critical
protocol
is
a
stage.
It's
not
actually
anything
else.
I
can
tell
anything
right,
but
in
some
cases,
in
order
to
get
it
done
without
a
real
network
here,
to
have
to
be
three
levels
deep
in
SSH
tunnels
that
get
whacked
by
this,
then
you
pull
it
all
up
in
three
hours
that
kind
of
stuff.
So
it's
a
question
of
how
painful
is
it
to
get
your
job
done?
How
painful
was
it
now?
X
C
Okay,
I'm
gonna
interject
a
tiny
bit
right
here,
one
of
the
things
that
I
mean
I
made
an
offhand
remark
earlier
that
I
didn't
that
I
wanted
to
use
a
slider
tool
and
rust
wouldn't
mess.
Wouldn't
let
me
but
I
do
think
one
of
the
the
problems
we're
going
to
have
through
this
conversation.
As
most
of
these
questions
are
not
yes
or
no
questions,
they're
questions
where
you
have
a
spectrum,
and
so
one
of
the
things
to
be
thinking
in
the
back
of
your
mind
is
I.
C
Think
after
we
had
some
conversations
on
this,
some
sort
of
a
detailed
survey
or
some
way
of
collecting
better
information
from
the
community
on
you
know
how
important
is
this
to
you
and
you
guys
are
well-versed
in
these
kind
of
things.
Now
you
do
these
IETF
surveys
all
the
time.
Is
it
not
very
important
or
very
important,
so
I
mean
I,
think
you
can
think
of
this
as
the
first
pass
through
these
questions,
and
then
you
can
help
us
develop
what
that
survey
would
be
going
forward.
D
But
what
I
heard
was
trying
to
summarize
is
that
we
need
to
continue
to
offer
v6
we
need
it.
That's
aren't
the
worst
thing.
If
the
address
pool
is
big
enough
and
captive
portals
are
a
real
pain
in
the
butt.
That's
what
I
heard
I
mean
lots
of
other
important
details,
but
the
summary,
if
you
had
to
put
it
down
in
a
few
sentences,
if
anyone
disagrees
with
that,
please
you
know
scream
okay,
the
next
question
I.
K
C
This
is,
was
I,
was
sort
of
a
I'm,
curious
I
mean
you
all
know
what
the
rough
distribution
of
the
of
the
ietf
attendees
are.
One
of
the
conversations
that
we
often
have
internally
is,
you
know:
are
we
only
hearing
from
the
people
that
are
really
verbal
or
are
we?
Are
we
really
getting
a
sense
of
the
whole
community
and
that's
why
we
ended
up
doing
surveys?
So
what
I
would
like
to
do
is
ask
for
two
hums.
C
C
Z
K
V
Like
to
offer
that
there
is
no
such
thing
as
an
average
ITF
attendee,
so
maybe
yes,
part
of
the
problem
like
if
you've
done
user
requirements
or
that
kind
of
thing
product
management,
we
have
personas
different
types
of
us
customers
that
you're
looking
at.
So
probably
you
want
to
think
about
like
what
kinds
of
people
come
to
the
ITF,
like
you
have
the
ipv6
folks
that
are
gonna
have
very
different
average
compared
to
any
other
category.
J
Jarrid
much
I
think
the
fact
that
most
people
in
here
or
most
other
people
in
the
IETF
right
now
are
not
in
here
is
actually
a
sign
that
what
is
working
right
now
is
working
relatively
well.
Yes,
you
know,
I
think
that
that
is
a
thing
and
a
few
people
who
are
super
passionate
or
who
have
environments
that
are
really
amazing
for
them,
like
Jen,
who
has
all
ipv6
only
and
doesn't
need
v4,
I,
think
that's
great
for
them,
but
I
think.
J
For
the
most
part,
the
network
in
the
infrastructure
is
actually
working
well
and
that
this
isn't
something
that
is
very
high
on
the
list
of
things
of
you.
There's
a
few
of
us
who
posted
either
conferences
or
provided
connectivity
before
I'm
sure
Glen
can
comment
on
that
or
some
other
people
who've
done
this
before.
But
for
the
most
part
this
is
working
well,
it
works
relatively
smoothly.
J
Yeah
I,
don't
think
we
need
a
lot
of
the
extra
stuff,
but
it's
also
a
good
opportunity
for
doing
that.
I
think
we
just
need
to
make
sure
we
beat
back
the
requirements
that
people
who
want
to
use
the
attendees
here
as
a
captive
audience
to
perform
an
experiment
because
I
actually
have
a
day.
Job
I
need
to
get
done
and
I
will
ultimately
revert
to
tethering
off
my
phone.
J
L
L
It
was
the
the
network
infrastructure
hotel,
wasn't
after
our
standards
or
networking
to
structure
the
guest
rooms,
wasn't
after
our
standards
for
a
variety
of
reasons,
and
that
meant
a
lot
of
places
got
rejected.
That
potentially
we
could
go.
Do
that
and
that's
part
of
I
think
an
important
part
of
this
equation
as
we
look
through
it,
especially
when
we
get
into
places
that
are
not
like
cheer
once
big,
a
tier
1
cities
in
North
America
or
in
Europe.
We
have
a
lot
of
problems,
we're.
G
Yeah,
my
recollection
is
different
than
what
you
start
describing
I.
We
having
been
in
that
role
for
for
such
a
long
time
I.
We
did
in
fact
reject
some
locations,
particularly
in
Japan,
based
upon
the
base
upon
Network.
However,
almost
always
there
was
a
direct
correlation
between
the
infrastructure
not
being
sufficient
and
the
infrastructure
and
the
rest
of
the
pieces
that
were
necessary
to
host
an
IETF
meeting.
There
were
not
we're,
not
a
good
fit,
and
so
I
I
can
count,
probably
on
one
hand,
maybe
maybe
two
hands
over
the
last
10
15
years.
G
The
the
number
of
sites
where
Marcia
carrot
rather
Marcia,
Stephanie
and
Laura,
had
said
yes,
this
is
where
I
want
to
go
and
I'm
like
sorry.
We
we
can't
do
it,
because
we
don't
have
network
infrastructure
and
also
I'd
like
to
expand
on
that.
Just
a
hair
and
say
that
it's
very
common
for
us
to
say
the
site
isn't
acceptable,
as
is,
however,
if
you
do
X
and
sometimes
X
is
instance,
tall,
some
fiber
or
X
is
install
an
additional
copper
run
into
it
into
a
meeting
room.
G
Then
we
can
come
and
that
has
been
successful
repeatedly.
In
fact,
we
were
commenting
on
the
infrastructure
in
Prague
being
better.
Well,
there's
a
lot
of
fiber.
That's
that's
in
Prague
that
we
that
didn't
exist
there
before
and
it
dramatically
has
improved
our
deployment
there,
and
that
was
because
we
said
hey.
This
is
a
great
place,
except
we've
got
this
issue.
We
during
the
sales
cycle.
G
We
got
it
added
into
the
contract
that
they
would
add
fiber
and
it
all
worked
out
so
I
call
into
question
the
assertion
that
we've
been
rejecting
things
primarily
because
the
network.
L
Go
ahead,
Wendy,
131,
clarification,
I
didn't
say,
is
the
primary
reason:
I
wasn't
like
the
funky,
Jimmy
I
really
am
but
well,
and
we
look
at
sort
of
like
what
what
was
the
higher
reasons
more
often
not
the
primary
reason.
It
was
like
you,
ninety-nine
percent
of
the
time,
but
of
all
the
various
reasons.
The
network
was
very
often
one
of
the
key
or
the
key
reasonable
objective
in
just
like
clarification.
L
P
W
Think
we
could
give
the
seals
the
over-engineering
up.
We
could
probably
provide
quite
a
bit
less
and-
and
still
nobody
would
be
in
this
room
because
it
would
still
all
be
fine
and
I
think
that
is
sort
of
my
concern
right.
It's
not
that
the
network
isn't
good.
The
network
is
awesome,
right,
don't
get
me
wrong
and
it
could
be
half
as
awesome.
It
would
still
be
way
better
than
anywhere
else.
I
go
in
Chapel
and
I.
Just
do
fine,
but
it
mean
and
the
same
reason
in
right
at
the
seam
so
yup
our
bar.
W
What
we
accept
is
so
high
that
it
excludes
a
lot
of
things
that
work
places
I
meet,
you
always
have
to
same
venues
and
it's
it
costs
money
right
and
that's
the
concern
right.
Obviously,
I
want
to
be
able
to
you
know
the
IDF's,
the
one
place
where
I
can
run
search,
but
then
you
do
my
laptop
certificate
because
it's
the
only
place
where
actually
having
before
it.
That
is
another
nice
I
can
do
it
a
few
times
a
year.
Do
I
need
to
know
so.
R
Hi
I'm
Bobby
indeed
Jarvis
also
involved
in
the
meeting
selection,
and
you
know
most
agree
with
everything.
Jim
said
I
mean
the
other
way
to
think
about.
It
is
the
net
the
venues
that
were
rejected
because
of
the
network
that
the
existing
network
infrastructure,
those
would
be
if
you
weren't,
building
your
own
network.
That
would
be
the
last
place
you
want
to
go
because
that
network
is
just
going
to
suck
all
the
time.
H
Hi
Hans
Khun
na
comes
here
something
I
think
we
should
keep
in
mind
as
we
talk
about
scaling
back
the
services
that
we
offer.
Is
that
I
think
that's
a
good
idea.
I
think
simple
claims
always
good,
but
we
reach
a
critical
threshold
where,
if
we
don't
have
a
volunteer
team
and
we
don't
have
a
contract
services,
team
or
100%
reliant
on
what
the
folks
do
and
there's
there's
got
to
be
a
certain
minimum
amount
of
services
that
didn't
be
interesting
to
provide
and
at
which
point
it's
really
easy
to
provide
additional
services.
H
So
some
of
what
we're
seeing
is
the
fact
that
we
have
really
good
team
who's
here
who
finds
things
that
are
interesting
to
do,
because
somebody
comes
in
and
says:
wouldn't
it
be
great
if
we
did
x
and
we're
responsive
if
we're
not
on
site,
none
of
that
is
possible,
and
so
I
think
you
know.
This
is
not
a
binary
function,
but
it
does
reach
a
point
where
you're
not
providing
enough
additional
value,
at
which
point
you
say
we're
just
going
to
use
whatever
the
hotel
provides.
F
So
weren't
marios
enough
Jeff,
it's
somewhat
following
onion
for
my
concept.
Yes,
we
didn't
understand
what
are
they
looking
like
sure
here
we
go.
F
Singing
shortly
so
kind
of
following
up
on
what
Han
said.
Yes,
we
definitely
want
to
understand
what
people
use
what
they
need.
You
know
what
we're
over
building,
but
there
is
some
amount
of
once
we've
got
the
base
stuff
set
in
place.
Additional
services
aren't
always
that
hard
track
so
serve.
Smtp
was
an
example.
Once
the
base
network
has
worked
at
it,
you
know
it's
graded
heading
SMTP
is
not
that
hard,
so
there's
sort
of
a
large
hump
to
get
over
then
there's
some
additional
stuff
beautifully
actually
know
what
the
additional
stuff
is.
F
F
AA
Not
volunteer
to
two
different
things:
on
the
on
the
show
floor
side,
echoing
something
Charles
said:
I
come
from
an
academic
research
and
academic
networking
background
where
the
stuff
that's
is,
is
similar
to
some
of
the
things
that
we
are
trying
to
offer
and
I.
Think
the
idea
of
experiments
and
eating
your
own
dog
can
all
tie
into
that
in
the
Hans
is
talking
about
the
interesting
problems
of
salt.
AA
H
H
AB
Q
D
Q
My
name
is
Tim
Sheppard
I've
been
there
a
lot
of
IETF
meetings
and
I
thought.
Well,
we
just
had
this
problem
in
Maastricht
and
then
I
looked
at
that
Maastricht
was
nine
years
ago.
Yeah
and
I'm
fully
prepared
to
believe
that
hotels
have
upgraded
their
networks
since
then,
and
maybe
they're
poor,
captive
portals
and
their
DHCP.
Things
are
better,
but
I
didn't
have
that
myself
so
and
I
I
know
you're
wanted
to
separate
the
venue
and
the
hotel
rooms,
but
it
could
be
the
same.
Q
C
F
K
F
Sadly,
I
don't
really
think
it
is.
We
had
something
I
can't
remember
where,
but
it
wasn't
that
long
ago,
where
we
went
into
a
hotel,
we
told
them
it
was
an
overflow
hotel.
We
told
them
they're
going
to
be
a
bunch
of
ietf
people
arriving
and
she
please
upgrade
their
network,
so
it
would
be
ready
and
then
we
ITF
participants
arrived.
That
did
not
work
well
and
they
kept
complaining
that
we
were
attacking
the
network
and
that's
why
it
wasn't
working
what
a
wreck
she
was
just
that's
what
an
IETF
person
looks
like
I.
F
Don't
really
think
that
there's
a
way
for
us
to
tell
ahead
of
time
if
the
network
is
gonna
survive
without
having
a
lot
of
people
doing
a
lot
of
stuff,
the
issue
isn't
doesn't
seem
to
usually
be
bandwidth.
The
issue
seems
to
be
capture
portal
stuff
that
wasn't
fully
understood.
Weird
generalization,
unusual
internal
infrastructure,
weird
configuration
for
broadcast
settings
things
like
that,
so
yes,
it's
entirely
possible
that
in
some
locations
the
hotel
network
would
work.
V
David,
this
comment
probably
should
be
for
any
guest
network
related
question
coming
later,
but
I
just
wanna
point
out:
I
noticed
that
the
hotel
network
here,
if
you
want
to
use
it
cost
$20
per
day
for
their
cheapest
option
and
then
also,
if
you
want
to
get
LTE
SIM
card,
the
cheapest
I
could
find
was
fifty
dollars.
So
it's
I'm.
S
Listen
Cooper,
yeah,
I,
I,
think
the
contracts
are
actually
quite
relevant
here,
both
for
the
case
that
you
raised
Warren
because
I
don't
know
if
there
was
a
contract
with
the
overflow
hotel
that
included
anything
about
the
network.
So
so
we
contacted
with
them
and
they
agreed
that
they
were
gonna
upgrade
the
network
and
then
they
didn't
do
it.
Is
that
what
you're
saying?
Okay.
S
Okay,
it
just
seems
a
little
odd
to
me.
Like
we
fly
people
around
the
world
to
do
site
visits,
then
you
say
well.
We
can't
really
tell
in
advance
whether
things
are
gonna
work
or
not
like
what
was
the
point
of
the
site
visit,
so
I
think
we
can
tell
something,
because
we
obviously
take
hotels
off
of
the
list
if
they're,
if
they
don't
meet
some
criteria,
so
there
must
be
something
there
anyway,
we
don't
need
to
continue,
but
but
also
with
respect
to
the
contracts.
S
Like
I,
you
know
our
our
diligent
staff
put
a
lot
of
effort,
presumably
into
negotiating
with
that
hotel,
so
that
we
could
take
over
the
network
in
the
guest
rooms
and
I
feel
like
we
should
honor
that
by
providing
that
service
to
the
attendees,
as
opposed
to
you
know,
for
half
the
people
turn
it
off.
So
I
think
we
probably
need
to
find
other
ways
of
experimenting.
Maybe
that
don't
include
like
gently
backing
down
from
the
things
that
we've
put
work
into
contracting
for.
J
That
Orochi
good
I
wanted
to
see
quickly
to
Tim's
point
because
he
referenced
the
moss
tree
comment
that
I
made
and
I
want
to
go
ahead
and
throw
something
a
little
more
time,
local,
so
last
Thursday
and
Friday
I
was
in
hotel
in
DC.
Obviously,
admin
instrument
by
the
IETF
and
I
mean
we're
talking:
irrelevantly
nice
hotel,
the
Hilton
property,
the
exact
one
escapes
my
memory
and
network
there.
I
was
getting
around
3
Mbps
and
actively
used.
Ssl
tunnels
were
going
down
every
thirty
minutes.
I
didn't
really
have
time
to
diagnose.
X
Boston
not
volunteer
very
good
response
to
the
different
comments
and
one
maybe
I
haven't
know
changing
the
subject.
Several
people
have
pointed
out
that
if
the
network,
then
you
must
screwed
up
enough,
you
just
revert
to
tethered
to
their
phones
and
that's
sort
of
true,
but
they're
scheming
issues
might
want
to
think
about
what
it
really
boils
down
to.
It
was
I
believe
meeting
if
you
tethered,
you
have
thought
I
would
invite
you
to
conduct
a
thought
experiment
of
twelve
hundred
people
each
tethered
to
their
phone
in
the
plenary.
I.
N
Match
up
again,
I
just
want
to
get
into
the
can
we
predict
that
what
would
work
or
not,
having
done
in
several
site
visits
myself
and
again.
K
N
This
possible
thing
here,
I,
can
tell
you
that
sometimes
we
can
clearly
tell
that
a
network
will
not
withstand
even
the
quarter
of
it
ITF
and
he
or
attendees
or
will
walk
at
all.
So
the
negative
case
is
easy
to
detect.
Sometimes
we
find
networks
that
are
very
well
built
like
this
one
in
here
that
wipe
the
actual
coverage,
the
wife,
the
radio
coverage
is
quite
good.
Still
after
day
one,
we
found
a
problem
that
there
was
something
somewhere
deep
birth
in
their
network.
N
That
said,
no
there's
a
way
too
many
broadcast
packets
here
and
we're
going
to
shut
down
course
for
four
minutes
and
this
this
we
cannot.
This
became
our
predict
during
the
site
visit
site
visit.
It's
usually
a
small
team,
four
or
five
six
people,
maybe
it
most,
and
we
cannot
simulate
612
on
our
even
100
ITF
entities,
but
sometimes
we
can
see
I've
been
to
once
where
I
can
tell
okay,
the
network
soceity
doesn't
even
work
with
me
alone
now,
let
alone
I
tip.
AC
AC
C
So
cost
did
come
up
as
a
consideration
in
this,
so
we've
started
looking
at
some
of
the
numbers,
but
one
of
the
things
we
specifically
wanted
to
do
was
to
ask
feedback
from
the
community
and
what
the
requirements
were
and
then
take
a
look
at.
This
is
what
you're
asking
for
this
is
what
it
cost,
how
they
get
balanced,
and
so
we
specifically
I,
don't
have
numbers
to
give
you
at
this
point.
I've
looked
at
some
numbers,
but
nothing
in
detail,
but
I
think
the
the
challenge
here
is
the
cost
between
real
dollars.
C
We
pay
out
versus
cost
in
kind
and
then
also
resource
like
people,
resources
and
I.
Just
think
that
we're
probably
better
off
focusing
technically
today
and
then
considering
in
the
next
steps,
how
to
deal
with
the
cost
and
actually
handing
it
off
to
the
people.
I
mean
we
have.
We
have
an
LLC
that
that
that
is
established
to
to
look
at
budgets
and
to
look
at
all
of
those
kinds
of
things,
and
perhaps
we,
as
the
technical
team
shouldn't,
be
doing
that
to
it
to
a
large
extent.
AC
C
F
C
C
Okay,
so
we
had
do
need
ipv6
and
then
the
next
one
is
our
attendees
okay
with
being
fully
exposed,
and
this
is
sort
of
the
you
know,
we've
you
know:
do
you
want
full
network,
or
do
you
want
some
client
protection
next?
What
are
your
expectations
for
security
and
privacy
on
the
network?
This
is
whether
you
expect
the
network
to
provide
you
security
and
protect
you
or
whether
you
are
willing
to
do
it
all
yourself.
Next,
how
reliable
does
this
network
need
to
be
I
mean
we?
C
We
have
a
lot
of
complexity
and
a
lot
of
cost
associated
with
reliability,
and
it's
not
you
know
how
do
how
do
we
manage
that
next?
Oh,
you
did
sorry
how
reliable
does
a
remote
participation
need
to
be
in
a
sense
that
there's
a
difference
between
the
network
going
in
here
versus
that
folks
being
able
to
get
in
next?
C
C
C
The
last
Venu
slide
and
then
we
have
guest,
dress,
guest
room
network
questions,
and
this
one
was
basically
the
you
know:
the
ietf
hotel
or
a
default
well
known
as
this
ID.
How
important
is
it
to
you
that
you
arrive
on
site?
You
open
up
your
laptop
and
you
don't
have
to
do
anything
on
the
network
works.
C
This
is
the
without
any
hints
enhancements
which
we
spend
a
lot
of
time
discussing.
This
is
one
we
have
not
spent
a
lot
of
time
discussing
because
it
also
implies
some
limitations.
It's
is
it
acceptable
to
default
to
the
regional
filtering
constraints
on
the
guest
guest
room
network,
and
this
is
one
of
the
questions
of
one
of
the
impacts
of
depending
on
where
we
go,
whether
we
take
over
the
guest
room
network
or
not,
and
then
we're
not
going
to
get
to
more
in-depth
questions
next
steps.
C
J
Sharon
Moss
again
I
I,
think
I
think.
For
me
there
is
a
very
important
thing,
very
important
thing
to
keep
in
mind,
which
is
that
as
the
guest
room,
access
has
improved
at
the
different
conferences
that
I've
participated
in
I
need
to
go
to
the
terminal
room,
engage
with
terminal
room
or
a
common
space,
and
24-hour
access
of
that
space
has
changed.
J
It
probably
would
be
better
if
they
could
do
it
in
their
room,
but
that
is
I
think
a
balance
between
the
environment
and
you
know
people
need
to
balance
their
own
job
requirements
and
if
ITF
isn't
their
main
thing
that
they're
doing
which
in
my
case
happens
to
not
sometimes
they
may
have
to
do
that
other
work
and
so
just
providing
the
ability
for
them
to
get
that
work
done
or
or
informing
them.
No,
we
are
not
going
to
do
that.
You
need
to
make
your
own
arrangements
and
being
relatively
clear
about
what
we're
gonna.
C
Just
a
tiny
follow
up
on
that
that
used
to
be
our.
You
know
the
when
we
used
to
not
do
the
venue
thing,
but
not
not
do
the
guest
room
Network.
The
point
used
to
be
that
people
needed
to
be
able
to
go
somewhere
to
get
24-hour
access
and
if,
if
the
hotel
network
wasn't
augmented-
and
it
was
a
terminal
room
when
it
was
a
space
like
that,
and
that
was
a
different
consideration.
I've.
L
Got
to
do
so
on
the
terminal,
room,
I
think
the
fact
that
people
continue
to
go
there
and
use
it
there,
they're
funny,
useless
and
so
I
think
we
have
a
simple
test
without
what
you
look
at
the
room
and
see
if
people
are
going
there
and
if
they
are
it's
something
that
they
need,
they
probably
don't
need
VT
to
be
52
or
beat
you
with
under
terminals,
in
the
bill
and
on
the
the
other
questions
people
through
the
slots
very
quickly.
That
was
what
I
wanted
to
cover
top
that
they
up.
L
On
the
dog
food
I
think
the
dog
food
question
is
is
very
relevant
and
I
think
it
is
something
we
should
you
need
to
do
and
it
we've
learnt
a
lot.
We
continue
to
learn
a
lot
and
as
a
sort
of
a
thought
leader
of
how
networks
should
be
designed
for
a
public
health
perspective,
it's
very
good
for
the
ITF
to
to
get
its
feet,
wet
and
sometimes
I
hope,
photography,
doesn't
taste
very
good.
Z
Z
The
idea
been
talking
about
writing
a
document
that
basically
says
if
you're
on
a
network
considered
consider
the
internet
threat
novelist
as
the
thing
that
you're
operating
with
I
personally
don't
want
to
have
a
network
helping
I
found
that
in
every
case
where,
where
a
network
is
helping
their
idea
of
all
things
different,
maybe
right
and
I
would
rather
that
not
happen.
We'll
probably
not
spend
resources
on
doing
things
like
in
network
malware
scanning
and
vendor
likes.
So.
D
Z
There
are
tunnels,
guess
when
I
say:
okay,
I
don't
want
that
happening
now,
I'm,
not
sure
that
I'm
diehard
on
that
particular
point
in
the
sense
that
you
sorry
I
do
not
want
someone
helping
me
from
the
outside
and
I
would
prefer
to
be
turned
off,
but
I,
don't
think
I'm
gonna
hold
the
line
on
that
one,
because
that's
some
networks
do
these
things.
The
other
point
I
got
up
to
to
make
was
that
filtering
requirements
vary
by
region
and
some
of
them
are
more
onerous
than
others.
Yeah
and
I.
M
H
Jason,
looking
good
on
the
terminal
room,
question
I
think
the
era
of
needing
24/7
access
is
probably
past,
given
connectivity
in
other
public
spaces
and
in
the
rooms,
but
it
might
be
interesting
at
the
next
meeting.
Thinking
of
surveys,
like
everybody
coming
into
the
terminal
and
for
example,
why
are
you
using
the
terminal
room
for
how
long,
like
you
know,
are
there
other
spaces,
something
somebody
designed
a
circuit?
I
understand?
Why
are
they're
using
it?
What
are
their
specific
needs,
so
that
is
better
care
Thanks.
This.
D
AD
Sean
Turner
so
I
guess
the
pile
on
the
turtle.
Room,
I
can't
even
get
wired
access,
and
this
thing
so
I
don't
know.
AD
Ipads
or
whatever
also
I
I,
think
there's
ways
we
can
gracefully
shut
it
down
to
two
Clemens
point
like
it's
being
used,
but
we
should
say
like
do.
We
really
need
to
think
this
way
forward
in
terms
of
guest
network
access
in
the
hotel,
like
in
my
guest
room,
I,
get
back
so
late,
I
mean
I'm
working
I'm
checking
my
email,
but
like
I
can
deal
with
it
for
a
little
bit.
It
doesn't
work.
AD
That's
fine,
so
I
really
don't
think
I
need
access
like
that
specialized
superfast
access
in
Jasper
I
think
I
could
do
without
it
and
I'm
in
this
hotel.
I
can
walk
down
right
like
it's.
If
you're
in
the
venue
Hotel
like
walking
outside
your
room
and
going
down
several
flights
of
stairs,
it's
not
that
owners
and
at
the
end
of
the
day,
isn't
it
all
about
how
we're
talking
together
a
being
inside
conversations
in
life?
We
don't
necessarily
want
everyone
in
in
their
rooms,
holed
up
picking
away.
P
C
L
L
This
thing
is
awesome:
it
has
lots
of
public
space
to
sit
down
and
type
way
up
in
other
places
where
we
have
far
less
seating
like
we'll,
have
open
space,
but
not
as
many
seats
I
often
will,
if
I'm
in
the
hotel,
I'll
sometimes
drop
in
the
terminal
room
just
to
get
a
table
just
half
way
out.
So
that's
sort
of
we
do
that
experiment.
We
need
you
over
a
couple
different
venues
to
get
a
bigger
picture
because
there's
some
multiple
factors
there,
the
other
factor
I,
think,
is
a
lot
of
people.
L
Nate
guess
on
my
part,
why
they
may
be
going
there.
Is
that
the
since
a
half-hour
or
more
don't
stay
at
the
conference?
Hotel,
sorry,
the
event
Lavelle?
They
might
be
going
there
to
do
some
work
between
working
group
sessions.
It's
just
an
opportunity.
Gotta
get
some
work
done,
cuz.
They
can't
write
up
to
their
room
and
I'm
Sean's
point
I
fully
agree,
I
mean
even
today,
I
was
in
my
room,
doing
some
some
slide
stuff
and
I
went.
You
know
what
I
don't
need
to
be
of
our
room.
L
D
AB
AB
J
J
Whether
something
works
or
not,
it's
a
marketing
opportunity
for
technologists
people
believe
in
and
I
don't
I.
Don't
really
think
it
adds
very
much
value
either
as
a
marketing
effort
opportunity
or
as
a
discovery
of
how
well
stuff
works.
I
think
you
should
just
you
know,
figure
out
what
network
we
need
to
get
the
job
we
came
down
here
to
do,
which
is
around
skander's.
U
Cosmetics
yeah
I
mean
so
you
mentioned
that
the
printers
are
no
longer
in
the
terminal.
Room.
I
just
wanted
to
say
that
wherever
they
are,
those
are
indeed
extremely
valuable.
I
haven't
needed
them,
I
mean
there's
been
times,
you
know,
I,
don't
always
need
them,
but
when
I
need
them,
I
really
need
them.
So.
M
L
M
P
X
X
We
would
have
an
intro
of
event,
which
is
just
three
or
four
people
sitting
around
the
table
in
the
terminal
room,
seeing
whether
or
not
we'd
actually
match
people
with
the
same
thing
from
the
same
specs
that
was
incredibly
valuable,
I,
don't
care
that
it'd
be
called
the
turtle
room,
they
don't
actually
need
network,
but
there
should
be
some
place
for
doing
that.
It
should
not
just
be
the
hackathon
is
actually
sometimes
we
need
to
do
it
on
Wednesday
second
thing-
and
this
is
not
volunteer.
I
have
a
question
group.
Did
we
close?
X
Let's
go
to
the
privacy
protection
question
there
we
actually
have
even
within
the
NOC
two
lips,
what
opinions
about
some
of
this
I?
Don't
think
it's
really
so
much
split.
There's
just
there
are
a
lot
of
facets
to
it.
We
primarily
serve
our
operative
guy.
My
servers
live
out
on
the
public
Internet
deal
with
it,
but
it
has
been
correctly
pointed
out
a
number
of
times
that
most
people
I
have
talked
to
do
not
usually
live
out
on
the
public.
X
Internet
there's
a
question
whether
or
not
we're
actually
doing
the
community
a
service
by
exposing
all
their
laptops
to
all
this.
Now
the
comments
before
about
your
exposed
anyway
deal
with
it
you
just
maybe
it's
a
smaller
group.
Ok
stern,
lander
threat,
I,
eight
hundred
of
your
closest
friends
as
opposed
to
seven
billion
people.
Okay.
So
it's
a
scale
problem,
but
most
people's
laptops
have
not
been
exposed
to
this.
In
most
other
environments-
and
you
know
it's
a
legitimate
question:
should
we
be
doing
something
about
that?
I
have
personally
think
it's
one
of
your
people.
R
Hi
three
things
well
to
the
last
comment:
I
actually
think
it's
good
to
be
exposed
because
in
most
laptops
are
in
fact
time
to
be
exposed
that
people
connect
them
to
all
different
environments.
They
have
no
idea.
What's
there
so
they're
exposed
so
I,
don't
see,
we
should
be
different.
You
know
hard
exterior
soft
interior
doesn't
offer
support.
R
Yes
I
do
think
we
should
eat
our
own
dog
food
I.
Think
that's
part
of
the
ITF
culture
and
the
ability
to
do
that.
To
do
two
experiments
like
we
did.
The
IQ
piece-
it's
only
things,
you've
learned
a
bunch
of
stuff
there,
the
others
and
futures
and
the
ability
to
do
that.
I
think
is
important.
It's
you
know.
It's
like
the
hackathon,
so
I
think
it's
a
really
good
activity
to
make
the
ICAT
relevant.
Okay.
O
One
of
the
things
I
was
thinking
about
is,
as
people
were
getting
up
here,
Who
am
I
I'm.
Sorry,
Mike,
st.
John's
was
the
question
of
how
many
people
here
are
sort
of
key
men
and
women
with
respect
to
their
companies
and
how
much
does
having
a
good
network
here.
Allow
them
allow
your
company
will
let
you
come
here
and
do
the
work
as
opposed
to
being
there
so
and
that's
both
the
hotel
rooms
and
here
so
I,
don't
I'm
not
working
for
anybody
directly
anymore.
F
So
sorry,
Warren
Kumari,
so
a
number
of
people
have
mentioned
stuff
like
know
the
ability
to
do
meetings
in
their
room,
something
which
I
think
we
often
forget
is
many
of
us
also
call
our
significant
others
from
our
room.
Sure
one
mentioned
you
know
he
sits
in
his
room.
Does
email
you
have
a
new
kid
I'm,
assuming
that
you
also
you
know,
do
video
conferences
with
your
kid
on
the
cellular
network.
H
Okay,
so
Jason
Livengood
two
comments,
one
on
the
dog
food
question
and
I
may
take
the
pet
analogy
here.
I
do
think
that
we
are
good
at
eating
on
dog
food,
but
it's
almost
like
we're
dogs
kept
by
the
king
in
the
palace,
and
it's
super
special
and
it
doesn't
resemble
the
dog
food
that
all
the
rest
of
the
dogs
outside
the
palace
walls
eat.
So
it
might
be
a
question
of
like
how
premium
is
dog
are.
H
And
then
the
second
is
on
the
security.
There's
really
interesting
questions
to
think
a
lot
more
about,
and
this
space
is
evolving
pretty
quickly
and
even
user
expectations,
as
we
know
here,
are
changing
quickly
and
I.
Think
it
ties
into
the
maybe
the
question
about
measurement
data
that
maybe
somebody
else
raised
earlier
you're
really
interested
to
see
more
data
about
the
things
that
are
happening
in
the
network.
You
know,
or
is
there
a
lot
of
c2
traffic
you
know?
H
J
J
There
may
be
some
other
things
that
we
want
to
actually
look
at
offering
somebody
earlier
I
think
mentioned:
v6
PD,
you
know
so
if
I
want
to
go
and
actually
run
something
to
an
experiment,
it
may
also
make
sense
to
just
dedicate
that
off
on
a
separate
SSID,
which
I
know
gets
into
the
engineering
thing
that
you
guys
not
to
do,
but
I
think
there's
I
think
there's
other
areas
to
explore
for
this,
but
I
think
in
general.
You
know
when
we
talked
about
the
security
of
openness.
J
A
lot
of
us
are
plugging
these
devices
connecting
them
to
unknown
wireless
networks
all
the
time,
and
that
is
just
kind
of
the
natural
state
of
things
and
I
have
a
corporate
managed
device.
It
has
protections
in
place
to
to.
You
know,
do
stuff
whether
or
not
they're,
the
right
ones.
You
can
argue
about
in
the
hallway,
but
that's
the
case,
but
as
far
as
working
remotely
warned
the
line
was
cut.
J
Picking,
I'm,
sorry-
and
you
know
the
you
know
as
far
as
being
able
to
work
remotely
and
such
you
know,
the
network
here
is
excellent
and
I
and
I've
had
no
issues
remoting
in
and
like
I
said
I've
seen
people
video
conferencing
out
here
as
well,
including
co-workers.
So
as
far
as
being
a
critical
employee
being
able
to
get
stuff
done,
I'm
able
to
do
that.
F
C
Okay,
we.
C
So
I
will
say
that
I
I
had
originally
thought
that
we
would
rip
right
through
these
questions
and
be
done
in
an
hour
and
I
should
have
realized
that
that
was
not
going
to
be
the
case.
But
there
were
not
food
fun.
There
were
no
food
fights
so
but
I
don't
think.
We
have
answered
the
question
sufficiently
to
help.
C
C
So
we're
on
the
what
next
question
and
I
think
that
at
this
point,
I
feel
like
we
haven't
really
quite
answered,
we
haven't
provided
enough
feedback
to
the
NOC
to
help
guide
some
of
our
design
sessions
going
forward.
I
think
we've
also
not
provided
enough
information,
that's
going
to
help
ministration
side
of
the
house,
so
we
have
a
mailing
list
and
so
I.
C
My
we
have
like
about
five
minutes
to
discuss
next
steps.
I
would
like
to
think
about
the
concept
of
using
the
mailing
list
to
develop
a
survey
that
we
could
then
send
out
to
the
community
using
the
mailing
lists
to
further
discuss
requirements
and
then
any
other
suggestions
in
the
next
five
minutes.
C
AA
N
C
AB
J
C
H
Kuhn
I
have
a
comment
about
next
steps.
Is
there
any
idea
about
how
to
communicate
to
the
community
as
this
process
moves
forward,
will
update
to
be
posted
to
ETF
all
or
only
to
the
network
requirements
mailing
lists
or
what's
the
best
way
to
communicate
with
this
audience
as
more
information
to
this
discussion
about
budgetary
issues
and
you
know,
meeting
requirements
etc.