►
From YouTube: IETF111-GAIA-20210730-1900
Description
GAIA meeting session at IETF111
2021/07/30 1900
https://datatracker.ietf.org/meeting/111/proceedings/
A
Or
like
yeah,
we've
been
introducing
a
lot
of
people
to
althea
they're,
actually,
a
grantee
where
we
have
a
project
in
with
a
big
bank.
Truist
has
been
supporting
some
work
that
we're
doing
in
the
southeast
part
of
the
united
states.
We
had
a
grant
proposal
launch,
and
these
are
networks
like
yours,
jared
that
are
profit.
Some
are
profit,
some
are
quasi
profit,
nonprofit.
A
It's
needed
all
right,
well
we're
at
the
top
of
the
hour.
I'm
gonna
start
us
going
forward
because
some
of
the
colleagues
here
on
the
line
who
are
presenting
are
in
africa
and
the
uk,
and
we
want
to
make
sure
that
they
get
the
chance
and
it's
late
where
they
are.
So.
Thank
you.
Everyone
for
joining
my
name
is
jane
coffin.
I
work
for
the
internet
society
I'm
here
in
my
personal
capacity.
However,
I
do
spend
a
lot
of
time
on
network
infrastructure
issues
and
the
intersection
of
technology
policy
and
development.
A
A
So
we
took
over
gaia
from
arjuna
and
we're
happy
to
just
continue
the
goodness
of
bringing
different
presentations
here.
There's
an
rsc
that
came
out
under
arjuna
we've
been
trying
to
think
as
to
whether
or
not
we
want
to
continue
to
look
at
updating
that
rfc
and
that
could
be
part
of
the
conversation
toward
the
end.
You
all
have
the
agenda
and
I'm
going
to
stop
talking
soon,
so
that
we
can
turn
to
that
agenda
and
run
down
the
presentations
that
we
have
today.
A
A
C
So
good
evening,
everybody
well,
I
think
it's
different
times
different
places,
I'll
just
say
greetings,
then
my
name
is
mary,
jane
mary
jane
sully
I'm
a
lecturer
at
the
university
of
jos
in
nigeria,
but
currently
I
am
doing
a
postdoc
at
the
international
center
for
theoretical
physics
in
tracy,
italy
and
my
research,
I'm
actually
I
used
to
be
a
network
administrator
at
the
university
of
georgia
before
I
moved
to
the
academia
strictly,
but
I'm
at
the
department
of
computer
science.
So
it's
still
like
a
build
up
on
my
foundation.
C
C
Then
there's
joel
goguin,
who
is
a
deputy
director
at
the
university
of
jos
icc
directorate
at
the
university
of
so
these
are
basically
the
introduction
of
the
people
involved
in
the
project
and
I'll
go
into
my
presentation.
Now.
My
presentation
is
going
to
follow
this.
The
following
overview.
I'll
talk
about
the
importance
of
the
topic.
C
C
That's
in
the
whole
world,
the
out
of
school
children
in
the
whole
world
60
are
coming
from
nigeria
and
that's
a
big
number
and
so
and
out
of
that
60
also
they're
coming
from
one
particular
part
of
nigeria.
That's
the
northern
part
of
nigeria
and
this
out
of
this
same
number
again.
60
percent
of
this
out
of
school
children
are
girls,
the
northern
nigerian
girls.
C
It's
not
actually
surprising,
even
though
I'm
another
nigerian
girl,
but
I
have
not
had
to
face
the
religious
and
cultural
challenges
that
most
northern
nigerian
girls
from
core
not
face
they.
So
because
of
this,
not
cultural
and
religious
challenges.
A
lot
of
northern
nigerian
girls
don't
go
to
school
actually
and
for
those
who
go
to
school,
there's
a
limit.
You
don't
really
get
to
the
top
of
it.
There's
a
limit
to
how
much
education
you
can
get
because
most
times
they
believe
that
the
girl
child
should
go
and
take
care
of
the
home.
C
You
know
like
what
she's
meant
for
is
to
take
care
of
the
home,
so
nobody
most
parents
from
northern
nigeria,
don't
really
consider
the
adult
child
education.
Some
of
us
have
been
lucky,
it's
a
different
story
for
us,
but
majority
of
the
core
north
and
nigerian
golf
are
having
challenges
to
go
to
school
and
we
know
that
for
any
development,
any
national
development
in
any
country
or
in
the
world
at
large.
C
C
C
C
You
know
so
the
the
idea
is
for
these
communities
to
to
enjoy
some
some
benefits
of
having
an
institution
in
their
community,
there's
so
much
to
consider
in
with
internet
and
education.
But
my
research
is
going
to
be
limited
to
a
particular
scope
and
I'm
limiting
this
research
to
basically
just
provides
quality
online
education
to
high
school
graduates.
Just
I'm
not.
This
research
is
not
considering
online
education
for
maybe
primary
school
students
secondary
just
high
school
graduates
and
for
the
out
of
school
youth.
C
You,
if
you
I
I
said
at
the
beginning-
that
there's
the
cultural
and
religious
challenges
that
the
girl
child
faces,
even
if
we're
ready
to
push
and
provide
education
for
these
girls,
we
have
to
provide
it
in
such
a
way
that
the
community
is
not
going
to
reject
it.
You
know
we
want
it's,
it's
supposed
to
be
a
win-win
for
both
the
community
and
the
girls
and
the
institutions
too.
C
So
it's
a
customized
application
that
will
still
teach
them
about
mathematics
and
we're
thinking
of
expanding
it,
not
just
mathematics
but
other
courses,
but
also
using
these
courses
to
teach
them
home
economics
home
management
recipes.
So
then
the
communities
will
accept
it
without
really
looking
at
it
like.
This
is
not
part
of
us
and
these
communities
are
where
you
most
of
you
must
have
heard
about
the
boko
haram
and
the
right
boko
haram
means
western
education
is
a
sin
right,
so
these
communities
are
in
these
places
where
the
boko
haram
has
strongholds.
C
So
you
just
want
to
show
them
that
okay,
the
education
is
not
as
bad
as
it's
made
to
look,
it's
also
a
good
thing,
and
they
can
use
this
to
also
learn
how
to
manage
their
homes
better
and
look
up.
Looking
at
the
scope.
Now
we
want.
How
do
we
deploy
this?
How
do
we
make
it
available
to
both
the
institutions
and
the
communities?
C
The
idea
is
for
this
community,
these
institutions
to
come
as
one
as
one
academic
networks
and
network
and
share
resources,
so
they
can
have
lectures
together.
Their
students
can
have
lectures
together
taught
by
the
best
lecturer
from
any
of
the
of
the
institutions,
but
having
the
same
class
and
then
having
access
to
the
same
lecturer
to
to
give
them
quality
education
to
to
make
them
be
a
part
with
their
with
their
mates.
Anyway,
these
lecturers
do
not
have
to
come
from
only
nigerian
institutions.
C
We
are
partnering
with
other
institutions
or
we
hope
to
partner
with
other
institutions
to
get
these
resources
to
provide
these
resources
to
these
institutions.
So
the
I,
the
the
by
the
time
this
project
is
fully
deployed.
You
have
academic
institutions
coming
together,
sharing
library,
resources,
human
resources
together
and
then
their
own
students,
too,
are
gaining
from
the
best
instructors
or
best
lecturers,
and
these
these
are
the
institutions
here,
as
you
can
see,
and
they're
going
to
share
these
resources
over
the
cloud.
C
So
what
are
the
resources?
The
the
services
with
all
this
deployment?
What
are
the
services
that
will
be
will
be
deployed?
We're
certain
that
the
remedial
sciences,
the
remedial
studies
that
are
run
by
these
institutions
will
be
run,
will
be
run
online
and
to
be
run
on
this
academic
network
to
provide
access
to
these
communities.
Some
communities
are
very
very
far
like
the
community.
We're
working
with
on
this
project.
C
So
they
can
run
these
remedial
studies
online
for
a
year
to
prepare
them
to
get
admission
into
any
science
course
later
on
in
in
any
of
these
institutions
that
are
partnering
with
them
and
once
they
are
running
these
remedial
studies,
they
have
access
to
the
library
services
too
in
any
of
these
institutions,
and
they
can
get
this
live.
They
can
get
books
from
these
libraries
by
just
going
online.
C
Checking
that
there
are
the
books
are
available
and
the
idea
is
to
have
these
books
transported
to
the
local
government
offices
in
these
communities
and
then
they
can
get.
They
can
go
and
collect
the
books
or
they
can
have
access
to
it
online.
If
it's
an
online
resource,
if
it's
a
physical
book
they
want,
then
they
can
take
it
down
to
the
local
government
office
and
they
can
get
the
book
and
another
service
that
will
be
deployed
on
this
network
is
having
lectures
online.
C
Like
I
said
earlier,
you
can
have
a
lecturer
who
is
very
good
in
a
particular
course,
and
he's
now
able
to
teach
these
children
better
and
put
them
through.
So
you
have
one
institution
coming
together
to
work.
You
have
one
lecturer
attending
to
different
institutions
and
providing
services
to
these
children
at
the
same
time,
and
so
we
have
the
interactive
mathematics
for
children
and
for
girls
that
will
teach
them
cooking
and,
at
the
same
time
also
teach
teach
them
mathematics.
C
So
we
had
a
proof
of
concept.
We
did.
We
ran
a
proof
of
concept.
We
had
the
ictp,
the
ictp
was
gracious
enough
to
provide
us
the
cloud
infrastructure
so
and
then
two
part
two
universities
participate
at
the
plateau
state
university
and
the
university
have
just
participated
where
we
ran
two
different
courses.
C
We
used
module
as
the
learning
management
system,
and
we
also
had
the
customized
application
that
was
able
to
teach
the
girls
how
to
do
some
fractions
in
mathematics
using
the
normal
recipe
we
had
from
our
poc
the
findings.
We
had
some
findings
and
the
findings
showed
that
the
female
students
actually
participated
better
than
the
male
students,
and
so
we
saw
that.
Okay,
actually,
technology
improves
on
on
this.
C
This
is
a
screenshot
from
one
of
the
classes
we
had
and
one
of
the
things
we
saw
that
even
people
who
were
sick
were
able
to
still
attend
classes
like
this
girl.
Here
she
was
sick,
but
she
was
able
to
attend
classes.
So
we
saw
how
it
has
improved
attendance
for
students
too,
and
there
were
also
some
interaction.
You
see
the
lady
at
the
bottom
still
had.
She
could
manage
her
shop
and
still
attend
lectures
again
at
the
same
time,
so
she
wasn't
divided
in
her
in
her
engagements
like
that.
C
So
this
is
the
recipe
where
you
can
see.
You
have
60
minutes.
You
soak
rice
for
half
an
hour.
How
many
more
minutes
do
you
have
like
it's
an
interactive
application?
The
video
is
there.
If
anybody
later
on,
I
can
share
it
with
anybody
who
wants
to
see
it.
So
this
next
screen
shots
to
are
from
the
video
of
the
interactive
applications.
C
There
were
some
challenges,
of
course,
while
doing
this,
while
trying
to
work
on
the
proof
of
concept,
we
saw
that
one
of
the
challenges
we
might
have
in
doing
this
thing
in
fully
deploying
it
with
the
phones
ictp,
had
to
provide
all
the
funds
for
us
to
run
this
proof
of
concept
which
were
happy
they
did,
but
after
it,
how
do
we
sustain
this
thing?
Then
the
unreliable
power
supply
problems
that
we
always
have
just
this
week.
C
The
national
grid
has
gone
down
and
we
know
that
there's
always
need
for
power
supply
for
internet
facilities
and
online
facility
applications.
The
human
resources
might
not
be
there
because
in
the
call
com
in
the
remote
areas,
most
of
them
are
running
back
to
the
urban
areas.
So
how
do
we
keep
local
technical
people
remaining
in
the
villages
and
still
have
this
on,
and
there
were
some
solutions
for
the
problems
we
identified
and
in
conclusion
we
saw.
C
We
said
that
using
ictp
I
sorry,
ict
facilities
would
definitely
improve
education
and
the
social
economic
development
of
the
host
community
also
and
the
female
child
can
still
go
to
school.
Even
in
the
core
northern
nigeria,
where
there
are
challenges
of
the
female
child
going
to
school
and
there's
a
saying
that
says:
train
a
woman
train,
a
nation.
So
if
you
can
get
education
to
the
female
child,
then
you're
sure
of
a
better
nation
and
a
better
world,
and
it
will
help
nigeria
to
meet
its
sdg
goals
too.
C
D
Hi,
mary
jane,
thank
you
just
a
quick
one,
because
I
had
to
look
away
for
a
moment.
The
communication
is
it
just
basically
institution
to
the
participant
or
the
student,
or
is
there
connectivity
with
the
outside
world
as
well?
So
is
it?
Is
it
a
closed
network
or
or
is
you
know,
can
people
reach
stuff
out
in
the
world.
C
A
A
Excellent
and
if
folks
didn't
hear
the
first
part
of
this
mary
jane's
working
with
a
good
colleague
of
many
of
ours,
marco
zanaro
from
and
I
assume,
amano
pietro,
somali
too
from
the
international
center
for
theoretical
physics
in
trieste,
italy,
they're,
a
great
team
of
experts
and
we've
used
them
in
the
field
to
do
some.
Radio
frequency
training-
and
I
believe,
hermano-
has
the
world
record
on
a
point-to-point
connectivity
from
italy
to
another
island
but
anyway,
side.
Note
all
right,
mary
jane!
Thank
you!
So
much.
E
C
C
So
if
you,
if
you,
if
you
are
replicating
this
like
for
the
apps,
you
just
replicate
it
in
such
a
way
that
it
will
reflect
the
cultural,
the
cultural,
no
norms
of
that
area,
but
and
then
for
the
internet
and
for
the
network
connection,
it's
almost
certain
that
it's
the
same
thing
everywhere.
It's
universal
everywhere,
sort
of
the
communications
building
a
a
community
cloud
is
the
same.
It's
the
same
way
you
can
build
it
anywhere
in
the
world.
C
So
when
it
comes
to
the
applications
or
the
services,
you
just
build
it
around
the
cultural
norms
of
that
area.
And
basically
the
thing
is,
you
don't
tell
somebody
that
your
culture,
the
things
in
your
culture,
are
bad.
You
just
tell
them
that.
Okay,
I
can
use
your
culture
to
get
something
that
am
good
out
of
it
and
you
will
it
will.
It
can
be
replicated
anywhere
actually.
F
C
Okay,
can
I
go
move
to
this
slide,
which
I
just
quickly
rushed
over
just
to
give
just
to
give
an
answer
to
your
question.
The
thing
with
the
infrastructure
is
that
I'll,
I
will
say.
Yes,
we
have
to
consider
some
designs
of
the
infrastructure
when
we
are
considering
countries
developing
countries
where
power
is
an
issue,
you
you
build
low
power
infrastructure
that
can
serve
them
for
as
long
as
possible.
C
I
mean
it's
not
really,
there's
a
limit
to
how
much
the
devices
can
run,
but
if
you
can
have
low
power
equipment
it
will
help
a
lot
because
some
of
the
equipment
gets
pulled
because
of
the
on
and
off
like
of
power
going
on
and
off
on
and
off
in
the
park
situation
is
really
bad.
It
might
not
be
very
obvious
for
people
in
developed
countries,
but
very
bad.
I
in
my
parents,
house
back
home
there
are.
C
We
still
have
like
three
sources
of
power,
but
there's
still,
there
are
still
some
times
that
you
might
go
without
power
at
all.
So
considering
design
infrastructure
is
very
good.
Considering.
Protocols,
too,
will
be
very
good
for
us
to
like.
Okay,
look
at
the
tv
white
space
technology
now
we
have
still
not
done
anything
in
nigeria
with
it,
because
this
this
they
are
still
working
on
the
policy
I
mean
for
how
long
you
know,
and
so
yes,
there
are.
F
A
We
need
to
connect
you
and
thank
you
for
that,
colin,
because
that's
a
problem
we're
seeing
so
many
places
around
the
planet
and
mary
jane.
We
should
connect
you
with
the
apc
team,
the
association
for
progressive
communications
who
are
working
right
now
with
the
government
of
nigeria.
So
this
is
where
technology
and
policy
start
to
interface
to
try
and
get
more
infrastructure
deployed
and
get
those
rules
and
regs
out
there
for
the
spectrum-
and
we
are
at
time
here
for
this
one.
Thank
you
so
much
mary
jane.
A
A
D
D
Just
to
give
you
a
little
bit
of
background
and
where
I'm
coming
from
here,
formerly
a
university
professor
computer
science,
networking
was
my
domain
as
part
of
that
about
10
years
ago,
my
goodness,
almost
exactly
10
years
ago
now
I
worked
with
a
few
other
colleagues,
other
professors
in
the
network
engineer
and
we
co-founded
a
community
broadband
isp,
but
of
a
different
order,
and
this
emerged
just
after
gui
fee
started
to
emerge
which
will
be
familiar
to
many.
So
what
we
found
was
with
just
a
little
bit
of
training.
D
The
the
real
problem
was
that
internet
services
access
to
the
wider
internet,
if
it
could
be
found
at
all,
was
prohibitively
expensive
for
any
community.
So
what
we
did
with
this
with
this
isp
was
we
launched
it
as
a
as
a
cooperative
for
backhaul,
specifically,
so
the
subscribers
of
this
isp
were
community
networks
and,
as
a
consequence
of
being
a
cooperative,
the
community
networks
were
also
the
members,
so
they
would
help
decide
the
fate
of
of
the
isp
any
additional
revenue.
D
They
had
a
hand
in
deciding
how
it
was
spent
far
more
often
than
not,
which
was
back
into
the
service,
which
is
a
good
thing
and
and
and
as
a
result,
people
who
were
you
know,
200
miles,
300,
kilometers
away
from
the
nearest
exchange
or
more
often
no
roads.
In
around
some
of
these
places,
they
were
getting
better
and
faster
internet
than
people
in
the
middle
of
of
glasgow
and
edinburgh
at
the
time.
D
So
this
is
where
I'm
coming
from.
In
terms
of
of
background,
I've
been
with
cloudflare
research,
as
I
said
for
a
little
while
now
and
this
week
they
launched
a
product,
and
I
promised
jane
there'd-
be
no
marketing
involved
in
this
one.
But
I
think
it's
important
to
understand
the
context
here
after
launching
that
product
on
monday
and
I'm
happy
to
talk
about
it
offline
if
anybody's
interested
in
preparing
these
slides.
It
really
started
me
thinking
about
about
what
would
be
the
next
thing.
D
D
D
In
in
the
next
case,
there's
this
notion
of
backhaul
to
an
exchange
or
an
internet
connection,
point
okay
and
I'm
gonna-
I'm
gonna
touch
on
this
in
a
moment
in
the
next
slide,
but
I
want
to
distinguish
between
the
notion
of
carrying
packets,
long
distance
and
and
moving
those
packets
on
and
off
the
internet
on
behalf
of
the
community
network.
So
the
back
hall
I'm
going
to
use
in
this
context
is
just
taking
packets
long
distance.
D
So
even
the
backhaul
community
networks
have
have
a
good
hand
in
in
deciding
these
can
be
self-owned
highly
unlikely.
I
acknowledge
they
can
be
via
some
sort
of
a
university
or
educational
network.
We
saw
little
bits
of
this
in
the
previous
talk,
thank
you
or
even
the
publicly
funded
national
networks
so
there.
So
there
are
ways
forward
and
again
I'll
touch
on
on
this
again.
D
There's
this
last
one
though,
which
is
the
internet
services
themselves.
So
people
have
various
names
for
this,
but
it
is
effectively
when
the
community
networks,
packets
are
routed
to
the
internet
and
the
responses
are
routed
out
of
the
internet
back
into
the
community
network.
Okay,
I
want
to
point
out
here
this
is
most
often
purchased
with
backhaul
service.
D
Usually
it's
the
telcos,
not
the
educational
networks,
for
example.
So
you
find
a
local
cabinet.
You
pay
a
small
fortune
if
you
can
afford
it
to
to
to
carry
packets
to
the
internet
exchange
and
since
it's
also
the
same
company,
that's
provided
you
with
the
backhaul.
It's
usually
cheaper
to
buy
them
together
as
a
bundle.
Okay,
but
I
want
to
emphasize
here
it
is
not
a
requirement,
so
it
is
possible
to
buy
the
backhaul
completely
independently
of
that
service
to
and
from
the
internet.
D
D
There
are
a
few
that
tend
to
be
in
urban
environments
and
mesh
along
them.
Okay,
the
second
is
going
to
be
the
one
that
is
most
familiar
to
people,
so
you
have
a
community
network
on
the
left.
Typically,
there's
a
cabinet
or
some
connection
point
that
is
at
least
reachable
community
network
sticks,
its
box
in
there
connects
to
the
backhaul
providers
box
and
then
the
back
hole,
which
is
the
red
component
here,
will
take
those
packets
onward
and
may
or
may
not
provide
those
internet
services
at
the
exchange.
D
Okay,
there's
this
third
model,
which
I
think
is
increasing
in
popularity
and
in
the
last
couple
of
months
I've
realized.
There's
more
of
these
out
there
than
than
I
understood
previously,
which
is
grand
and
it's
this
idea
that
the
backhaul
service
and
the
internet
service
in
particular,
is
so
expensive
beyond
any
community
network's
budget.
D
So
the
nearby
community
networks,
even
they
might,
they
might
be
100
or
200
kilometers
from
each
other
jane
pointed
out
the
longest
distance
wireless
link
across
an
ocean
today
on
the
west
coast
of
scotland
right
now,
there's
run
that's
that's
25
miles
or
40
kilometers,
and
it
runs,
I
think,
at
about
100
megabits
per
second
with
just
straight
ordinary
wi-fi.
So
you
get
a
bunch
of
communities
together
that
might
connect
to
each
other
for
resilience.
But
crucially,
they
all
connect
to
some
point
at
one
end
of
a
block
called
service.
D
D
D
So
what
are
we
going
to
do
here?
Here's
what
here's?
What
I
would
like
to
ask
and
kind
of
spark
conversation
and
point
out
that
that
cdns
they
are
not
internet
service
providers,
not
by
any
stretch
of
the
imagination,
but
they
are
very
well
connected
networks
and
in
fact
some
cdns
are
more
connected
than
any
tier
one
isp
that
we
can
imagine.
D
Okay.
This
is
really
really
interesting
by
the
way
when
I
say
well
connected
here,
I
want
to
point
out.
This
is
well
connected
relative
to
to
their
own
scale.
Okay,
so,
yes,
you
have
your
mammoth
international
cdns,
your
cloud
flares,
your
akamai's
and
your
fastlys,
but
there
are
also
smaller,
regional
or
national
type
cdns
that
people
create
all
on
their
own.
D
So
relative
to
their
operating
environment,
cdns
are
very
often
well
connected,
and
so
the
question
then
that
comes
that
falls
out
of
this
observation
is
what
might
cdns
and
other
non-isp
entities
do
to
contribute
to
the
community,
network's
connectivity,
okay
and
crucially,
here,
and
this
one's
really
really
important.
This
little
bottom
caveat
is
also.
We
have
to
consider
the
reasons,
I'm
not
going
to
say
that
they
might
want
to
do
so.
I
claim
that
they
should
want
to
do
so.
D
So
let's
look
at
cdns
and
these
non-isps
and
and
the
facilities
and
the
features
that
they
offer
right
in
order
to
operate
internally.
They
route
data
around
within
the
infrastructure
and
they
probably
run
additional
services
on
top
of
that
related
to
things
like
content,
delivery
and
caching
or
security,
or
both
and
more
okay.
So
they
run
familiar
type
things
internally,
they're
well
equipped
externally.
D
Well,
they
have
to
be
able
to
interface
with
the
rest
of
the
internet,
and
so
they
have
reliable,
high
quality
connections,
okay
to
go
to
the
wider
internet,
and
they
have
addresses
that
they
can
announce
via
bgp,
just
like
everybody
else
does,
and-
and
one
thing
something
that
applies
equally
irrespective
of
size
is-
is
that
is
that
the
differences
are
associated
with
scale
alone.
Okay,
so
yes,
it's
true
that
a
regional
isp
will
have
a
smaller
pipe
to
the
outside
world
than
a
than
a
mammoth
global
cdn.
D
D
So
that's
fine,
I
mean
from
a
technical
perspective.
Cdns
have
many
of
the
elements
that
you
would
need
from
an
isp-like
entity.
Okay
and
then
one
has
to
ask
well
look:
cdns
aren't
transit
providers
by
any
stretch
they
have
some
content
and
they
send
it
out
to
the
world
when
it's
requested.
So
what
are
the
incentives
for
them
to
do
just
a
couple
of
things
differently:
okay
and
and
here's
the
interesting
thing
is.
I
actually
firmly
believe
that
the
incentives
here
to
engage
with
community
networks
offer
connectivity
or
services
to
community
networks.
D
It
aligns
much
more
tightly
with
cdns
than
it
ever
will
with
isps
okay,
personal
claim.
So
let's
consider
this
think
about
the
additional
bandwidth
and
service
costs.
So
if,
if
there's
a
cdn
that
offers
connectivity
to
a
community
network,
clearly
that's
going
to
increase
the
the
the
bytes
in
flight.
D
That's
great!
Now,
here's
the
big
incentive.
As
far
as
I
I'm
concerned,
if
there
are
more
connections,
remember
you're,
a
cdn:
you
don't
know
how
to
originate
content.
You
have
customers
that
originate
content.
You
need
to
make
them
happy
whether
that's
a
government
or
or
a
commercial
doesn't
matter.
D
If
there
are
more
connections
and
more
users
who
are
connected,
then
there's
a
larger
audience,
and
that
means
happier
customers
and
it
means
happier
customers
in
a
number
of
ways.
Right
might
actually.
Not
only
is
there
is
there
a
larger
audience
for
those
customers
to
reach,
so
imagine
you're
a
retailer.
D
So
not
only
might
you
reduce
your
own
costs
by
a
bandwidth?
Not
only
will
you
make
your
hap
your
customers
happier,
but
you
might
actually
reduce
your
customers
costs
of
delivery
service.
Okay.
What
about
charging
models?
Well
come
back
to
this
first
bullet.
If
you're
a
large
isp,
you
probably
don't
feel
it
and
if
you're
a
smaller
isp
you
could
you
engage
in
a
cost,
sharing
model,
okay,
and
so
the
trick
here
is
that
you
should
never
charge
as
a
cdn
any
more
than
cost,
because
there
are
all
these
other
benefits
alongside.
D
D
That
first
question:
should
the
itf
or
similar
decide
somehow
interfaces
or
best
practices
with
respect
to
these
community
network
to
cdn
connections
when
they
appear?
I
have
no
answer
it's
hard
to
know.
Ideally,
cdns
use
some
sort
of
open
standard
to
do
this
so
that
people
can
move
around,
not
in
addition
for
for
interoperability,
but
you
can
imagine
some
cdns
might
want
to
retain.
D
What
about
commercial
interests?
I'm
going
to
come
back
to
that
previous
slide,
happier:
customers,
okay
and
this
last
one
kind
of
came
up
while
I
was
thinking
about
this
generating
the
slides
this
week
is:
could
community
cooperative
models
that
we
know
work
well
in
the
communities
that
we
now
see
work
well
for
backhaul
and
onward
internet
services?
Could
a
community
cooperative
model
also
extend
to
this
space,
the
cdn
space
both
for
content
and
for
community
network
connectivity?
D
D
He
started
a
student
and
he's
a
young
student
pre-university
and
he
started
a
40-pop
anycast
cdn
out
of
his
bedroom
during
covid,
because
he
wanted
to
learn
more
about
networking
and
he
figured
this
was
a
good
way
to
do
it
and
it's
a
full
stack
service
right
from
ecmp
load,
balancing
at
the
bottom
end
to
https
service
and
certificate
management.
At
the
top
end,
he
has
running
on
his
cdn
foss
host,
which
is
a
non-profit
cdn
for
open
source
software.
D
A
B
One
of
the
biggest
challenges
I
had
was
finding
that
backhaul
capacity
and
I
could
get
it.
But
I
could
get
it
a
mile
easter,
a
mile
west
of
where
I
wanted
to
be
and
the
additional
cost
to
build.
That
was,
you
know,
somewhere
between
50
and
100
000
dollars
to
get
that
extra
mile,
and
that
poses
a
unique
challenge.
I
don't
I'm.
B
I
I
think
there
there's
going
to
be
a
lot
of
challenges
in
trying
to
convince
a
cdn
such
as,
for
example,
akamai
or
cloudflare
or
netflix
or
or
somebody
who
has
this
content
to
pay
for
the
last
mile
networks.
That's
what
a
lot
of
the
isps
have
wanted
have
kind
of
said.
You
know
well
you're
using
our
lines.
You
should
pay
us
that
the
economics
dynamics
at
play,
of
how
you
do
that
are
very
broad
and
a
huge
challenge,
and
I
mean
85
percent
of
the
cost.
B
But
getting
getting
to
that
connection
point
is,
I
think,
the
big
challenge,
and
how
and
and
who-
and
how
should
close
that
gap
is-
is
an
incredibly
complex
dynamic
of
public
policy
in
the
geolocation
where
that
access
is,
and
I'm
not
sure,
I'm
not
sure
where
that
intersection
is
where
you
know
with
the
standard
setting
organizations
or
the
research
organizations
are
because
a
lot
of
that
seems
to
be
outside
outside
of
that
wheelhouse.
So
that's
that's
my
commentary
on
it,
but
I
I
do
think
that
there
is.
B
There
is
some
interest
or
value
in
you
know
the
cdns,
but
we
found
the
cdn's
so
switching
hats
to
more
of
my
day,
job
cdn
hat,
like
we
build
deeper
into
these
networks
because
we
have
to
in
order
to
get
the
content
to
the
end
users
and
we
kind
of
chase
the
end
users
to
some
extent
to
those
those
regional
aggregation
points
and
so
yeah.
We
should
be
building
more
of
those
and
they
should
be
in
more
neutral
locations.
B
But
when
you
get
to
even
a
large
market
like
detroit,
you
don't
have
you,
don't
even
have
the
two
major
players
interconnected
in
the
same
locations
like
the
two
major
incumbents,
let
alone
anything
else,
and
so
just
those
dynamics
that
are
at
play
are
are
really
hard
and
things
get
very
diffuse,
very
fast.
Outside
of
the
large
regional
aggregation
points.
D
So
so
so
I
will
say
a
couple
of
things.
If
I
may
first
is
remember,
there
are
many
differences
throughout
the
world.
So
what
we
see
in
the
us
we
see
we
can
often
see
different
things
elsewhere,
but
but
you
point
out
some
some
really
really
good
ones.
You'll
find
policies
in
a
lot
of
places
are
are
bizarrely
designed,
and
I
can
I
can
personal
opinion
as
to
why
that
might
be
so.
For
example,
the
the
the
cost
of
lighting
fiber
is
based
on
distance
in
most
places.
D
I
know,
which
makes
very
little
sense
to
me,
considering
that
once
a
fiber
once
something
is
in
the
ground
and
I'll
acknowledge
that
initial
cost,
the
the
cost
to
light
a
fiber
over
a
kilometer
like
100
meters
or
a
thousand
meters,
is
the
same
and
why
they
should
be
priced
differently.
I've
never
understood,
and
that
is
a
policy
thing
as
far
as
actually
getting
the
connectivity
yeah
when
you're
laying
fiber
there
are
some
serious
complications.
If
you
rely
on
on
the
large
telcos
to
do
it,
it's
never
going
to
be
cheap.
D
In
my
experience,
what
happens
is
the
community
co-ops
will
put
up
wireless
connections
and
then
use
the
revenues
from
that
over
time
in
order
to
lay
their
own
fiber
and
typically
the
places
where
you're
laying
fiber
are
places?
You
know
land
owners
who
already
have
little
tractors
and
things,
but
you
can
rent
them
and
do
it
as
at
a
fraction
of
the
cost
of
of
what
you're
suggesting
not
saying
it's
that
easy.
I
know
there's
all
kinds
of
policy,
land,
ownership
and
other
issues,
but
but
I
think
it's
doable.
D
B
A
Hey,
I
just
want
to
make
sure
we
we
caught
up
with
anyone
else
who
wanted
to
be
in
the
queue
and
just
a
recognition
that
there
are
a
lot
of
different
embedded
regulatory
and
policy
schemes
that
entered
there's.
Also
just
a
lot
of
what
goes
on
the
municipalities
of
different
entities,
having
a
lock
on
those
municipalities
or
even
where
I
am
now
on
the
buildings,
which
is
ridiculous
right.
H
D
There's
no
reason
that
if
you
couldn't
get
the
isp
to
agree
at
least
to
route
to
those
boxes,
you
could
get
couldn't
get
similar
types
of
service.
So
by
no
means
am
I
offering
suggesting
that
all
the
answers
are
here,
but
I
think
it's
a
it's
a
space,
that's
worth
exploring
and
and
and
and
talking
about
jeffrey.
I
Kind
of
proposing
a
re-architecture
of
the
way
the
internet
works
right
like
right
now,
there's
there's
kind
of
the
backbone
providers
who
have
a
funding
model
that
I've
never
really
understood
and
you're,
saying
that
it
makes
sense
to
to
replace
them
with
cdns
whose
funding
model
is
to
charge
the
content
provider
and
use
community
networks
to
get
kind
of
the
last
mile
and
some
sort
of
data
transfer
between
the
big
cdns,
so
that
so
that
you
can
kind
of
cross
those
those
worlds.
But
it's
I
mean.
I
I
mean
they
have
to
in
order
to
do
what
you're
proposing
like.
If
they
provide
that
backbone
service,
then
the
traffic
has
to
be
able
to
escape
the
cdn,
and
so
if
this,
if
this
happens
like
you've,
you've
replaced
the
architecture-
and
I
don't
know
if
that's
good
or
bad,
but
it's
it's
an
interesting
implication.
D
On
a
on
a
personal
level,
I
might
say,
I'm
eternally
grateful
for
everything
isps
have
given
us
the
traditional
isps,
but
it's
also
been
20
years
that
we've
been
trying
to
solve
this
problem
and
and
universally
isps
are
seeing.
I
would
love
for
them
to
be
bigger
partners
than
they
are
perceived
to
be.
Let's
put
it
that
way.
A
D
A
No,
the
it's
a
part
of
the
function
of
how
you
have
it
set
up
on
your
screen
and
I'm
learning
how
to
make
the
setup
better.
Okay,
well
note
that
marwan's
presentation
is
up
in
data
tracker
and
you
know
where
to
find
him.
Thank
you
so
much.
This
is
fabulous
and
a
food
for
thought
about
the
architectural
changes.
J
This
is
a
discussion
about
how
the
state
of
vermont
has
enabled
empowered
local
jurisdictions
to
actually
do
ban
deployment
and
how
fortunate
we
are
to
have
that
public
policy
in
place,
as
well
as
the
funding.
That's
coming
our
way.
So
we
are
not
we're
not
struggling
like
many
of
the
folks
that
I've
admired
over
the
years
who
are
working
around
the
world,
installing
networks
do-it-yourself
networks
so
who
we
are
and
what
we're
doing
and
how
we're
going
to
pay
for
it.
J
J
But
they
were
not
a
municipality.
They
were
just
a
group
of
that
had
signed
local
agreements
among
themselves
with
regard
to
delivering
broadband
in
the
area,
and
so
they
were
looking
for
the
recognition
of
the
status
of
being
a
municipality.
So
they'd
go
to
the
municipal
bond
market
and
as
a
result
of
this,
they
went
to
the
legislature
and
at
that
point
in
time,
vermont
in
2015
passed.
J
This
alleged
enabled
the
creation
of
what
they
call
communication
union
districts,
which
enable
towns
to
form
the
cud,
and
these
cuds
are
a
non-profit
single
purpose
and
the
goal
is
to
deliver
symmetrical
100
megabits
per
second
to
all
the
underserved
in
vermont.
That's
the
goal
this
year
in
2021
they've
set
up
a
ban
board
which
purple's
purpose
is
to
facilitate
and
support
the
cuds
and
their
efforts,
and
they
also
set
up
this
20.
This
vermont
community
broadband
fund
into
which
is
being
poured
all
the
grant
money.
J
J
J
So
our
mission
is
to
deliver
affordable,
reliable
internet
service
to
the
underserved.
Our
first
priority
is
the
underserved
and
the
goals
of
cv
fiber
are
to
make
that
service
available
to
half
of
the
underserved
in
our
20-member
community
by
80
by
the
following
year
and
95,
but
the
year
after
that
and
notice
the
little
asterisks
on
the
bottom.
There
subject
to
fund
material
and
contractor
availability,
which
is
going
to
be
the
challenge.
J
J
J
So
this
is
this:
is
the
process
of
building
the
network,
something
I'm
sure
you're
all
familiar
with.
I
have
to
do
poll
inventory,
so
we
have.
I
have
three
contractors
standing
by
waiting
for
a
notice
to
proceed.
While
we
wait
for
grant
funds
to
come
in,
and
so
they
can
launch
and
launch
their
poll
inventories,
but
we're
going
to
be
taking
information
about
each
poll
in
our
in
our
communities
in
some
25
000
polls.
Within
this
communication
union
district
and
we're
collecting
about
50
bits
of
data,
including
the
height.
J
What
things
are,
what
the
attachments
are,
how
old
is
it
whether
it
needs
to
be
replaced,
etc,
etc?
So
that's
the
first
part
of
then
going
into
network
design,
high
level,
design
and
detail
engineering
and
then
make
ready,
of
course,
you're
probably
familiar
with,
and
that
is
just
moving
things
on
the
pole.
So
we
12-inch
space
where
the
fiber
is
going
to
go
a
very
expensive
proposition
by
the
way,
and
then
we
get
to
the
construction
part
of
actually
stringing
the
fiber
up
there
and
then
the
last
part
is
the
service.
The
connection
bit.
J
So
here's
the
breakdown
having
broken
down
the
pre-construction
construction
operations
bit
here
is
the
bus
we're
expecting
for
the
work
over
the
years
poll
data
about
eight
hundred
thousand
dollars.
You
can
see
the
make
ready,
work,
six
and
a
half
million
dollars,
construction,
etc,
connections,
bringing
us
to
20
to
46
million
dollars.
J
So
this
is
a.
This
can
be
a
succession
of
images
here
about
what
we
expect
to
do
and
when
we
expect
to
do
it,
we're
expecting
to
do
all
of
the
pre-construction
work
in
2021.
That
is
this
year.
So
we
see
coal
inventory,
high
level,
design
and
detailed
engineering
poll
inventory.
We
have
grant
for
area
a
some
200
000
plus
to
do
that.
J
J
Here's
the
construction
phase,
one
the
difference
here-
is
that
this
this
is
addressing
the
underserved.
So
you
see
the
phase.
One
miles
is
different
from
the
miles
total,
so
the
phase
one
miles
for
area
a
for
example,
is
150
miles
is
intended
to
address
the
needs
of
those
underserved
within
that
within
those
five
towns,
and
you
see
to
do
construction
and
service
in
2022
and
make
ready
for
most
of
for
b
and
c
also
in
2022,
and
you
see
area
3,
which
is
300
miles
in
2023..
J
J
We
expect
to
get
federal,
arp
or
grant
funds
about
20
million
dollars
and
you're
probably
aware
that
the
federal
government
american
rescue
plan
act
has
set
aside
broadband
funds
and
that
the
state
of
vermont
is
getting
us
about
250
million
dollars
that
we
expect
to
get
about.
20
million
of
that
towns
themselves
are
also
getting
opera
grant
funds
and
we're
expecting
to
get
about
five
million
dollars
of
contributions
from
them.
J
J
All
of
these
rural
co-ops
were
set
up.
Well,
this
this
washington,
electric
co-op
that
we're
working
with
has
the
possibility
of
get
obtaining
two
and
a
half
percent
loans
for
the
purpose
of
building
out
a
network
and
about
500
miles
of
their
possible
network
might
be
in
our
district,
and
so
we
might
have
access
to
loans.
J
Our
goal
is
to
provide
affordable
subscription
rates
and
and
you're,
probably
well
aware
that
the
cost
of
the
cost
of
subscription
rate
is
dependent
upon
the
cost
that
it
costs
to
build
this
thing,
so
we're
expecting
50
million
dollars
of
that
the
operations
has
yet
to
be
determined.
What
that's
going
to
cost
and
the
other
bit
large
cost
is
the
cost
of
money,
and
so
the
more
at
the
lower.
J
The
subscription
rates
are
going
to
be,
and
to
give
you
an
example
of
the
impact
of
that
ec
fiber,
which
I
mentioned
earlier
as
being
the
first,
the
first
one
is
offering
a
72
rate
for
25
25
service
and
debt
service
is
about
fifty
percent
of
that
rate,
and
so,
if
they
didn't
have
ones
that
had
some
50
million
dollars
in
bonds
outstanding,
they
didn't
have
those
kind
of
bonds.
Then
their
rate
could
be
considerably
lower.
J
J
Let's
see
here,
and
so
this
year
we
have
been
very
busy
and
you
can
see
that
we've
done
a
lot
of
work,
especially
in
the
last
few
months,
with
issuing
rfps
and
launching
inventories,
signing
contracts,
etc,
etc.
So
the
great
there's
a
great
deal
of
traction
under
what
it
is
we're
trying,
and
I
have
to
acknowledge
the
fact
that
my
years
with
the
ietf
and
all
the
experience
that
I
had
with
rfps
and
requests
for
bids
and
work,
orders,
etc,
etc.
K
Ray,
can
you
hear
me
I
can
this
is
this?
Is
an
awesome
project?
I'm
really
really
glad
to
see
that
that
you're
taking
this
on,
because
I
suspect
this
is
an
awful
lot
of
work
for
you.
Besides,
I
mean
what
you've
just
shown
us
here
is:
is
substantial
you're
going
to
you're
planning
to
run
most
or
all
your
fiber
above
ground
or
above
ground
between
towns,
or
how
do
you
I'm
not
sure
where.
K
K
K
J
I'm
also
looking
that
we
might
be
able
to
use
grant
funds
for
driving
down
the
cost
of
installation
so
that
installation
fees
are
non-existent
for
those
homes
that
are
within
400
to
600
feet
of
the
pole
and
for
those
that
are
longer
we're.
Looking
at
providing
customer
assistance
programs
money,
some
grant
money
that
we
might
be
able
to
obtain
so
to
drive
that
down
and
there's
also
would
be
an
opportunity
for
reducing
prices
for
those
who
can't
find
even
fifty
dollars
a
month
as
being
affordable
for
them.
J
J
L
J
So
there's
existing
cable
companies-
there
is
a
the
consolidated
communications,
is
overlapping
its
own
telephone
lines
and
they
have
about
400
million
dollars
in
investment
that
they're
that
they're
accelerating
their
their
delivery,
and
they
can
do
it
a
lot
of
than
us
because
they
don't
have
to
make
ready.
They
don't
have
the
their
expenses
are
far
less
and
their
time
needed
is
is
less.
J
A
A
M
To
see
you
likewise
and
I'm
glad
to
see
you're
staying
busy,
I
have
sort
of
two
unrelated
questions.
I
was
talking
with
george
sadowsky
actually
a
couple
of
weeks
ago
about
ec
fiber
and
I'm
wondering
how
your
experience
compares
to
them,
partly
because
they're
pioneers
and
partly
because
they're
they're
part
of
vermont
is
a
lot
richer
than
yours.
J
Yeah,
so
I've
been
talking
to
george
as
well
and
to
to
the
current
chair
or
vc
fiber
is
also
the
chair
of
something
called
v
cooter,
which
is
the
association
of
cuds,
and
their
experience
is
that
they've
had
to
go
to
the
bond
market
and
have
all
that
have
all
that
heavy
debt
to
to
pay
off.
They
have
an
operator
where
we
are
currently
rfp
out
for
an
operator,
and
it
kind
of
appears
that
the
tail
is
wagging
the
dog
a
little
bit
over.
J
M
J
J
We
have
the
right
to
acquire.
We
can
acquire
companies,
for
example
existing
companies.
The
other
thing
we
could
do
is
we
can
condemn
and
take
easements
nobody.
Nobody
was
interested
in
my
idea
of
condemning
the
telephone
lines
and
taking
the
easement,
which
would
allow
us
to
overlash
their
telephone
lines
that
wasn't
very
popular,
but
I'd
like
to
have
done.
That
would
have
saved
a
lot
of
time
and
money.
M
Basically,
you
know
the
comment
was
I
mean
yeah
vermont.
Has
these
r
licks
and
they're
all
being
rolled
up?
You
know,
like
yeah,
my
sister's
companies
got
rolled
back
mike
mike
my
my
company
just
got
rolled
up
to
the
same
one
that
rolled
up
the
shore
company
and
these
guys
just
want
money.
So
anyway,
I'm
scratching
my
head
is
there
some
way
that
we
can
make
it
more
in
their
interest
to
cooperate?
But
presumably,
if
there
were,
you
would
be
doing
it.
J
J
But
what
we
might
find
ourselves
doing
over
a
period
of
the
next
five
ten
years,
for
example,
is
that
we
could
one
instead
of
having
nine
different
operators
running
nine
different
networks,
we
could
find
that
there
there
may
be
just
one
two
or
three
operators
that
are
that
are
operating
these
the
operating
these
networks,
but
the
contract
is
going
to
be
that
we
own
the
public
infrastructure
and
they
are
under
contract
with
us
to
operate
the
network.
They
have
no
financial
investment
in
it
and
a
lot
of
companies
are
looking
for.
A
A
The
rules
aren't
necessarily
as
complicated
from
state
to
state
as
they
or
from
place
to
place
in
those
places
so
they're,
looking
at
actually
creating
new
rules
and
regs.
That
would
open
up
the
availability
of
municipal
and
community
led
networks.
So
I
think
ray
you're
up
against
a
lot
of
entrenched
legacy,
not
only
mindset
but,
of
course,
the
regulations
which
impact
the
technology,
which
is
pain
in
the
neck.
A
E
E
Yeah
well,
I
I
would
focus
on
commenting
that
there
is
an
ongoing
topic
to
this
to
focus
on,
which
is
that
we
all
know.
Technology
has
changed
quite
a
lot
in
the
last
years
and
then
we
have
an
rfc
that
reflects
the
state
of
the
world
more
than
like
five
years
ago,
and
definitely
by
the
time
we
finish
a
revision.
Things
would
have
changed
significantly
and
not
only
in
the
technology
part,
but
also
in
the
business
part
in
the
regulation
part.
E
I
was
wondering
when
I
was
hearing
rave
to
what
extent.
What
is
being
done
now
could
be
done
before
copied,
for
instance,
not
only
the
perception
of
the
importance
of
network
connectivity,
but
also
the
implications
of
that
in
in
in
policy
technology
and
everything,
and-
and
this
is
something
that
we
definitely
have
to
do
and
and
probably
the
way
to
start.
It
is
by
by
sending
an
analysis
of
the
of
the
rfc
and
calling
for
individual
contributions
to
it.
A
Sandra
is
opening
up
the
conversation
to
us,
updating
the
rfc
that
currently
exists
in
the
guys
based
aleandra.
One
thing
I
would
throw
out
and
for
everyone
here
is
the
cost
elements
that
we've
been
hearing
about
and
jared
on
what
labor
costs
are
ray
on.
You
know,
studies
from
whole
cost
to
or
poll
analyses
and
or
environmental
analyses
that
go
into
it
or
from
mary
jane,
just
the
cost
of
dealing
with
them
all
of
some
of
the
challenges
that
she's
got
in
front
of
her.
A
I
don't
know
if
we
had
a
robust
cost
analysis
in
that
rfc.
I've
got
to
take
a
look
again,
but
I
like
the
idea
of
opening
this
back
up
to
see
if
we
can
add
sort
of
more
sprints
and
agility
to
annexes,
to
the
rfc,
potentially
on
new
tech.
That's
out
there.
We
know
a
lot
of
new
equipment,
cheaper,
equipments
out
there
from
ubiquity,
microtic,
etc
and
just
different
ways
of
building
networks
and
all
over
the
planet
which
have
shown
us
that
there
are
different
models
that
can
be
used.
A
A
E
A
A
F
Yeah
I
mean
I
you
know.
Obviously,
if
you
think
things
have
changed
and
there's
new
experiences
and
new
information
to
add
to
that
rfc,
it
seems
like
it
would
be.
It's
potentially
worth
thinking
about
an
update
or
or
another
document
to
discuss.
What's
happened
since
then
yeah
absolutely
seems
seems
a
reasonable
thing
for
the
group
to
do.
A
F
F
A
Electricity
issues
getting
equipment
in
I
mean
that's
just
a
challenge
and
of
itself
sometimes
with
customs
and
the
question
you
raised
earlier
about
should
the
protocols
and
john
and
others
were
going
back
and
forth
in
the
chat.
Do
some
of
the
protocols
need
to
shift
depending
on
different
environmental
concerns
in
a
certain
country,
yeah
yeah?
Maybe
an
aggregate.
F
A
Just
about
everyone
we
work
with
has
generators,
solar,
other
other
or
two
generators.
You
know
the
hot
cut
over
those
are
from
both
internet
exchange
points
and
community
networks
and
actually
there's
some
really
interesting.
Iot
related
analyses
that
one
community
network
we
know
in
the
republic
of
georgia
is
doing
with
respect
to
the
solar
panels
that
they
can
troubleshoot
same
in
greece.
That
could
be
some
really
interesting.
A
D
A
Put
something
in
the
chat
we,
we
were
actually
about
200
kilometers
above
the
arctic
circle
about
four
years
ago
and
okay.
You
know,
we
all
know
what
permafrost
is,
but
boy
was
that
something
that
we
wanted
if
the
local,
the
local
experts
there
said,
yup,
there's
no
way
you're
digging
you're,
putting
it
above
on
these
funky
like
troughs,
which
were
interesting
along
the
road,
but
to
your
point,
people
don't
think
about
that,
and
there's
also
fiber
that
some,
some
colleagues
we
know
at
liquid
telecom
in
sub-saharan
africa
have
had
to
design.
A
D
When
you
buy
little,
when
you
buy
little
cabinets
to
put
your
little
mini
micro,
tick,
router
or
something
inside
of
it,
and
it
just
because
it
says
waterproof
doesn't
mean
that
it
is
and
when
you
make
it
waterproof,
you
know
on
a
hot
day,
so
in
many
places
that
actually
need
it.
These
things
will
melt
so.
E
Yeah
that
raises
the
question
to
what
extent
producing
a
document
that
talks
about
the
unconnected
requires
the
unconnected
to
participate
and
if
because
otherwise,
it
would
reflect
the
views
or
the
knowledge
or
the
experiences
of
those
who
already
are
so
well
connected.
And
and
if
it
is
just
a
matter
of
writing
or
is
it
requires
some
research
and
some
involvement
or
questionnaires
with
those
who
are
not
yet
connected
or
partially
connected,
or
that
do
not
speak
english
and
that
that
do
live
in
places
where
there
are
no
markets.
E
A
B
A
E
A
Excellent
all
right:
well,
we
have
our
marching
orders,
meaning
things
for
us
to
do.
Thank
you
very
much
everyone
for
being
here,
and
especially
you
land
with
you
for
your
moving
houses,
but
thanks.
Everyone
for
joining
gaia
provides
an
opportunity
to
hear
from
different
people
who
are
building
and
executing
on
networks
rather
than
and
to
test
things
in
the
wild.
A
So
we'll
take
all
of
the
suggestions
on
board
and
the
great
chat
that
we've
had
here
and
some
of
the
expertise
from
all
of
you
who
are
actually
out
there,
building
the
networks
and
testing
things
in
real
time
and
ray
who's,
developing
networks
and
designing
them
mary
jane.
Thank
you
very
much
ray.
Thank
you
very
much
nmr1
and
we'll
be
back
at
you
online.
So
thank
you.
We'll
close
a
little
early
and
give
you
back
15
minutes
of
your
day,.