►
Description
On our May 21 call, Arky (twitter.com/@playingwithsid) introduced us to offline knowledge hotspots that create low-cost, wifi-enabled and solar-powered digital libraries in resource-constrained schools.
Learn more about the monthly calls of the IPFS Local Offline Collaboration Special Interest Group: https://github.com/ipfs/local-offline-collab/issues/25
B
If
you
saw
a
prompt
asking
you
about,
recording,
please
hit
okay.
If
you
were
okay
with
this
meeting
being
recorded
hello,
my
name
is
Dietrich
Iola
and
I
work
on
the
I
confess
team
and
I'm
happy
to
welcome
you
all
here
to
the
monthly
local
offline
meetup,
where
folks,
who
are
interested
in
technologies
that
work,
offline
may
or
may
not
use
IP
fest,
but
even
discussing
the
use
cases
and
application
and
needs
in
offline
communities
comes
up
regularly
a
loosely
knit
group
of
people
with
shared
interests.
Today,
I
would
like
to
welcome
our
key.
B
A
A
B
B
B
A
So
thanks
for
being
right,
my
name
is
rocky,
and
this
talk
is
more
about
my
journey
trying
to
bring
education
to
remote
places
where
I
was
I
was
living.
It
was
my
journey
through
India,
where
I
was
born
and
raised
through
Kenya,
where
I
lived
for
a
while
and
and
all
the
way
to
Southeast
Asia,
where
I
live
and
work
right
now,
I
love,
libraries,
I
spent
most
of
my
school
days,
living
in
libraries
and
I,
have
a
strong
belief
that
if
we
provide
knowledge,
children
could
self
learn,
but
but
the
places
where
I
live.
A
It's
not
always
I
call
them
resource
constrained
countries,
because
quite
often
it's
we
don't
have
access
to
broadband
access
to
even
powers
and
I
work
for
I
was
lucky
enough
to
work.
A
lot
of
interesting
projects.
I
was
part
of
Mozilla
for
a
long
time
working
on
the
localization
team
and
after
my
work,
I
am
a
strong
believer
on
the
power
of
digital
technology
to
bring
equitable
access
to
make
things
equal
for
people
and
the
project
that
I
am
currently
talking
about
is
one
such
project.
A
I'm
always
kind
of
designing
my
solutions,
so
I
don't
have
to
import
stuff
from
anywhere.
So
it's
always
driven
by
the
resources
that
are
available
and
the
challenges
that
that
I
face
on
the
ground.
So
one
of
the
one
of
the
challenges
that
I
would
quite
often
see
living
in
in
Southeast,
Asia
and
during
my
time
in
Africa
is
education
is
quite
often
is
the
only
way
for
people
to
to
make
a
good
life
for
themselves.
That
includes
me.
A
I
had
I
was
lucky
enough
to
have
a
very
good
education
and
I
was
able
to
to
make
a
good
life
for
ourselves,
and
so
I'm
always
looking
at
that,
and
this
project
is
about
trying
to
do
try
to
bring
high
quality
education,
but
making
sure
that
that
it's
openly
licensed
so
quite
often
in
most
of
the
countries,
even
the
textbooks
are
not
open.
Licensed
though
it
is
made
by
the
government,
these
textbooks
are
owned
by
private
publications.
A
So
I
am
quite
aware
that
it's
really
important
to
have
educational
content
under
open
license
and
also
if
we
are
building
a
solution,
I
tried
my
best
to
use
open
hardware,
so
we
don't
get
into
a
lot
into
any
kind
of
windows.
Also
not
just
this
solution
is
not
just
for
students,
it's
actually
for
teachers
as
well,
because
in
most
countries
teachers
rarely
get
sort
of
refresher
courses.
Quite
often
they
never
go
to
any
refresher
courses.
A
So
this
this
kind
of
offline
libraries
is
not
only
just
for
students,
not
only
just
also
for
teachers
where
they
can
refer
to
to
the
to
the
resources
in
there,
though,
this
is
primarily
targeted
for
schools,
but
it
could
be
also
deployed
in
places,
uncommon
places
like
disaster
relief
camps,
refugee
camps
out
and
also
schools
that
community
centers
in
certain,
not
world
news
places
in
a
big
city.
So
it's
so
they're
the
inspiration,
the
origin
story
of
this
whole
thing
goes
back
to
this
project.
Perhaps
you
have
heard
about
the
hole-in-the-wall
project
in
India.
A
This
was
an
experiment
by
su
gato
Misha.
He
noticed
that
the
kids
living
in
Islam
next
to
his
office
could
never
had
access
to
any
form
of
formal
education.
So
what
he
did
is
he
literally
made
a
hole
in
the
wall
and
put
an
end
cold.
Thank
you,
computer
with
a
keyboard
and
a
mouse
that
was
access
to
thing,
and
it
was
just
an
experiment
to
see
these
kids
barely
speak.
English
I
mean
this
whole
computer
is
in
English,
but
very
quickly.
A
He
realized
that
kids
learnt
they
mastered
how
to
open
programs,
how
to
launch
a
browser,
how
to
traffic
and
then
very
quickly.
They
started
teaching
each
other.
This
kind
of
self-paced
learning
words
until
then
they
were
like
it
was
always
like
Oh
children
has
to
go
to
school,
they
have
to
stay
in
a
very
structured
environment
and
a
teacher
will
impact
education
and
the
children
will
learn,
but
hole-in-the-wall
project
has
shown
them.
Given
the
right
tools.
A
A
A
This
is
a
Linux
machine
with
multiple
screens
multiplexed
and
on
and
the
internet
was
delivered
through
reset.
This
was
before
3G
had
come
to
East,
Africa
and
there's
a
solar
panel
that
was
providing
the
battery
the
power
for
this
whole
setup.
So
those
who
could
pay
could
actually
access
internet
and
they
would
pay
the
shop
the
kiosk
owner
some
money,
but
the
children
would
use
this
resource
for
free
I
worked
on
this
project.
A
I
did
a
little
bit
of
integration
for
this
project
and
when
this
was
this
was
completely
offline,
and
this
was
like
really
really
inspiring.
Even
today
about
10
years
later,
people
are
still
asking
us
like
hey.
We
would
like
to
deploy
these
machines
in
places
like
South
Sudan
right
now.
Use
of
Mesa
is
no
longer
required
because
we
do
have
pretty
good
2g
and
3G
connections
in
some
parts
of
Africa.
So
this
was
my
first
experience
in
trying
to
bring
education
offline.
A
So
since
then
things
have
changed
a
lot.
I'm
gonna
switch
gears
here
and
talk
about
hardware.
One
of
the
first
successful
projects
that
was
done
worldwide
was
the
One
Laptop
Per
child
project.
This
was
not
really
a
hundred
dollar
laptop,
but
that
was
the
goal
to
have
a
inexpensive
but
rugged
London
that
could
be
used
for
children
for
in
schools,
and
this
was
really
really
advanced
for
its
time.
It
actually
had
two
Wi-Fi
cards
and
it
was
doing
mesh
networking,
which
was
super
advanced
like
if
they
like.
A
If
in
a
classroom,
environment,
children
could
actually
see
each
other
out
and
they
could
actually
collaborate
work
on
this
platform.
So
this
was
really
great,
but
the
challenge
is
making
hardware
is
really
expensive.
It
was
really
hard
to
sustain
this.
This
kind
of
model
trying
to
build
hardware
15
years
later
right
now.
What
we
use
is,
we
have,
we
are
using
these
raspberry
PI's,
which
are
the
single
board
computers
right
now,
I
just
check
the
prices,
it's
about
like
30
dollars,
and
it
has
the
same
performance
as
the
One
Laptop
Per
child
device.
A
One
of
the
reason
we
are
moving
away
from
laptops
to
Raspberry
Pi,
which
does
not
have
a
screen
which
does
not
have
a
keyboard,
is
because
of
the
proliferation
of
mobile
phones.
Now
this
has
changed
everything.
This
is
the
last
year's
stats
of
mobile
penetration
and
Cambodia,
where
I
live
and
I'm
so
shocked.
Actually,
Cambodia
population
is
about
13
million,
but
there
are.
There
are
more
phones
than
people
which
is
kind
of
strange,
but
it's
not
very
uncommon.
Like
people
have
two
or
three
SIM
cards,
so
I
guess
that's
that's
a
thing.
A
A
I
forget
the
name
of
the
company
that
produces
it,
but
that
particular
charge
controller
is
connected
to
a
a
car
battery
and
a
large
solar
panel
on
the
the
red
box
is
the
Raspberry
Pi,
with
a
128
Meg
gigabyte
SD
card
that
has
all
the
learning
material
and
the
white
box
is
actually
a
Wi-Fi
router,
that's
running
open,
wrt,
lassparri
pi
has
a
onboard
Wi-Fi
chip,
but
it
it
is
not
powerful.
It
is
not
reliable.
A
So
if
you
it
can
connect
to
ten
clients,
but
as
more
clients
join
the
network,
it
just
drops
and
it
consumes
a
lot
of
power
when
I
talk
about
challenges
of
building
such
libraries,
Wi-Fi
is
remains
one
of
the
biggest
challenge.
The
good
thing
is:
all
of
these
devices
could
be
run
using
a
power
bank
that
you
use
to
charge
your
mobile
phone,
because
the
chip
that
is,
that
the
board
that
is
in
your
phone,
is
actually
the
same
on
chip.
That's
running
these
devices,
so
the
power,
the
power
consumption
remains
very
very
little.
A
A
It's
it's,
it
runs
on
the
Raspberry
Pi
and
you
can
see
the
interface
you
could
once
you
join
this
Wi-Fi
network.
It
will
point
you
to
a
series
of
mock
style
courses,
so
you
can
choose
whether
you
want
to
learn
through
Konica,
Khan
Academy,
or
you
want
to
learn
about
how
to
be
teachers,
how
to
deal
with
bullying
in
schools.
So
it
is
self-paced
social
learning
and
you
can
also
download
stuff.
So
that
is,
you
have
some
storage
on
your
phone.
A
So
quite
often
you
could
go
out
when
you're
in
school
or
in
a
community
space.
You
could
access
the
access
to
content,
download
it
and
you
could
you
could
view
it
when
you're
back
home.
One
of
the
most
important
thing
about
this
kind
of
environment
is
in
its
classroom
environment.
The
teacher
you
have
to
follow
the
pace
set
by
the
teacher,
but
in
a
learning
environment
like
this,
some
kids
who
are
smart
could
go
through
the
material
really
quickly,
but
others
who
would
need
a
little
more.
They
can
learn.
C
A
Their
own
pace,
so
it
is
it's
it's
very
adaptable,
so
this
this
is
some
of
the
volunteers
in
Cambodia
teaching
the
kindergarten
teachers
how
to
access
it.
We
are
trying
to
develop
a
lot
of
local
content,
videos
and
music,
so
they
can
be
used
in
kindergartens
and
primary
school.
This
particular
device
actually
has
about
100
gigabytes
of
content
from
Khan
Academy.
C
A
A
To
sum
it
up,
this
is
applying
knowledge
hotspots
that
basically
are
a
way
to
deliver
contents
offline,
but
also
try
to
create
a
learning
environment
that
could
be
very
sub
space.
We
we
have
deployed
this
and
ran
it
for
about
a
year
now,
unfortunately,
right
now
due
to
go
with
19,
the
schools
are
closed
until
number
2020.
A
So
so
I
its
the
schools
are
closed,
but
the
devices
are
still
running
I
hope
we
have
found
out
that
that
we
had
very
little
problems
except
one
solar
panel,
which
was
faulty,
which
was
poor
quality
we
had
to
replace,
but
otherwise
it
was
running.
This
library
is
also
fixed
in
schools,
but
we
also
have
a
mobile
to
talk
or
a
small
three
wheeled
cart
that
goes
around
different
schools
and
we
have
tablets
and
in
the
in
the
tiptoe.
So
children
in
the
communities
could
actually
borrow
the
tablet.
A
Do
an
activity
or
watch
a
video
or
use
their
parents
are
quite
often
busy.
They
borrow
their
parents,
phones
and
trying
to
access
it.
The
next
step
for
us
is
to
have
a
lot
of
local
content
on
this
on
these
devices.
It
is
really
important
for
us
to
have
all
of
this
offline,
often
quite
often,
because
the
school's
themselves
cannot
pay
the
power
bills.
I
mean
they
have
some
out.
They
do
have
power
in
schools,
but
quite
often
they
don't
have
budget
to
pay.
C
A
Calibri,
the
software
that
is
running
in
the
libraries
have
some
p2p
system
so
that
they're
thinking
about
adopting
IVFs
the
implementation
is
has
still
hasn't
happened.
So
from
the
hardware
side
I
found,
there
is
no
Wi-Fi
device
on
any
of
the
SOC
birds,
whether
it's
Raspberry
Pi
or
the
boats
from
Olli
makes
or
any
of
the
banana
I
and
all
these
variants.
The
the
Wi-Fi
chip
is
quite
often
very
unreliable.
A
It
does
not
scale
it
quite
often
it
come.
It
consumes
so
much
power
that
it
becomes
unreasonable
to
use
it
on
a
solar
setup.
So,
right
now
we
have
to
use
a
open,
wrt
device
which
actually
somehow
is
much
more
power
efficient
and
it
scales
to
more
than
50
lines
per
device,
and
also
we
could
do
a
mesh
network
where
we
could
actually
extend
the
reach
of
the
digital
library
to
the
whole
village
by
using
a
mesh
network.
A
So
that
makes
it
really
so
it
is
really
when
building
this
system
I
had
to
use
a
lot
of
hacks,
because,
unfortunately,
you
you
cannot
have
a
captive
portal.
I,
don't
know
if
you
know
what
a
captive
portal
is,
but
if
you're
in
a
hotel
or
at
the
airport,
and
you
connect
to
Wi-Fi,
you
get
this
splash
screen,
which
is
called
the
half
the
bottle.
Interestingly,
captive
portal
relies
on
DNS,
so
if
you're
doing
deployments
like
this,
where
there
is
no
DNS
capital,
puffles
don't
work.
A
So
one
of
the
tricks
that
we
use
is
called
it's.
It's
called
a
DNS
black
hole
where
we
are
basically
listening
for
the
request
and
sending
them
to
our
local
local
server.
It's
a
it's
a
hacky
way,
but
this
needs
to
be
replaced.
Dns
should
work
service,
discovery
should
work.
Offline,
I
say
should
because
it's
absolutely
needed
mobile
devices.
A
Primarily,
we
are
looking
at
Android,
but
Apple
iOS
devices
as
well.
At
the
moment
they
hit
a
captive
portal
there
they
check,
for
they
do
something
like
call
back
home
where
they
check.
If
this
device
is
really
connected
to
internet,
if
it
doesn't
the
new
Android
and
iOS,
they
actually
drop
the
Wi-Fi
or
they
give
them
a
say.
Saying
like.
Oh
sorry,
if
there
is
no
internet
on
this
Wi-Fi
device,
I
will
not
connect.
Now
this
becomes
an
issue.
A
Once
you
connect
to
the
once,
you
have
connected
to
the
DNS,
you
have
a
captive
portal.
You
have
to
come
out
of
the
captive
portal
to
a
web
browser,
so
we
could
redirect
you
to
our
learning
materials.
This
captive
portal
hand-off
really
doesn't
work,
because
when
you
are
doing
captive
portal,
both
iOS
devices
and
Android
devices
give
you
an
iframe
which
is
not
really
a
true
browser
and
they
don't
allow
us
to
fire
up
a
browser.
A
So
this
whole
onboarding
experience
is
really
hockey
and
it's
really
difficult,
especially
when
you're
work
when
you're,
when
you're
trying
to
deploy
it
in
somewhere,
where
people
don't
have
experience
in
technology
before
so,
it
becomes
a
bit
of
an
issue
to
successfully
to
do
do
the
deployment
of
this
system.
Someone
needs
to
have
a
good
internet
connection.
A
So,
if
I'm
doing
this
in
AP,
I
need
to
find
someone
who
has
a
really
good
internet
connection
in
Haiti
to
actually
download
all
the
materials
first
and
then
copy
it
off
to
the
SD
the
SD
card
and
deploy
it.
So
it
is.
It
is
really
dependent
on
having
someone
on
the
ground
with
good
internet
connection
to
through
the
deployment
and
also
when
you
are
when
you
are
watching
videos,
there
is
no
mesh
network.
So
there's
a
lot
of
load
on
the
system,
especially
when
you
are
broadcasting
videos.
It
becomes
a
really
an
issue.
A
A
Syllabic
is
the
largest
portion
of
the
deployment
cost,
the
board,
the
Raspberry
Pi
and
the
access
point
they
are
very
inexpensive.
Even
the
SD
card
is
only
like
$15
now,
but
the
solar
panel
itself
is
120
dollars
and
in
some
places
it's
even
$200
and
they
don't
last
for
a
very
long
time.
So
for
these
kind
of
deployments
to
work,
actually,
solar
power
should
become
more
accessible
and
more
reliable
and
I
hope
it's
getting
cheaper,
but
I
hope
the
prices
will
drop.
So
we
could
do
more
deployments.
A
Schools
are
closed,
so
we
can't
do
any
more
deployments
anymore,
but
the
plan
was
before
Corbin
19
to
expand
this,
to
form
our
schools
and
to
have
a
lot
of
local
language
learning
material
myself,
I'm,
trying
to
design
if
you're
a
product.
If
you
have
some
experience
in
software,
engineering
or
hardware
manufacturing
I,
would
love
to
talk
to
you
because
it's
possible
to
shrink
this
package
into
one
small
device.
A
All
you
have
to
do
is
unpack
this
device
pop
in
and
SD
card
get
some
content
and
it
should
be
ready
to
go
and
that's
the
vision.
I
would
really
love
to
have,
but
right
now,
I
don't
have
the
resources
or
the
technology
skills
to
make
it
happen.
There
you
go.
That's
all
for
me.
I
would
love
to
answer
any
questions
that
you
might
have.
A
B
You
for
coming
and
speaking,
people
who
figure
I'll
use
the
clap
response
in
the
chat
that
was
fascinating
and
I.
Think
there's
a
there's,
a
number
of
questions
in
the
chat
first,
one
there
is
whether
or
not
there
are
media
playback,
speed
controls
and
some
of
these
self-guided
learning
systems
can
do
to
beat
them.
Media
playback
speed
controls,
so
people's
ability
to
determine
the
speed
at
which
the
media
is
played
back
faster,
slow,
also,
probably
the
rate
at
which
they're
going
through
learning
materials,
it
seems
like,
would
be
relevant
as
well.
A
B
A
A
C
B
Fantastic
thanks,
I
had
a
question
about
Calibri:
do
you
bundle
the
it
does?
Calibri
have
distributions
and
releases
that
have
content
bundled
in
or
do
you
on
these
devices
install
Calibri
and
then
dump
everything
onto
all
the
media
and
content
on
the
Westie
card?
How
tightly
coupled
are
those
systems
so.
A
A
It
takes
forever
to
download
content
both
during
deployment
time,
but
also
I
would
love
to
have
when,
when
you
have
people
who
have
already
downloaded
content
on
their
mobile
phone,
it'll
be
nice
to
actually
send
packets
laterally,
then
coming
from
the
server
and
also
updates.
Now
that's
another
thing.
The
way
this
system
is
updated
is
by
swapping
SD
cards,
so
I'm
working
on
a
system
where,
when
a
mobile
library
goes
to
a
school,
it
could
talk
to
their
system
and
say:
hey
I
have
some
new
files.
A
B
A
Sir
cubics
was
the
project
based
out
of
Mali,
that
did
it
and
they
used
a
format
called
sim,
which
was
a
format,
but
the
idea
is
to
actually
have
an
open
systems
that
could
and
right
now,
whoever
videos
our
videos
are
the
most
large
content.
That
needs
to
be
synchronized,
and
there
is
no
good
solution
for
that.
D
I
think
yeah
so
like
putting
Zim's
those
like
offline
snapshots
of
big
Wikimedia
websites,
such
as
Wikipedia
in
different
languages,
is
something
we've
been
experimenting
recently
and
the
recently
you
mentioned
the
upgrade
mechanism
right
now
is
based
mostly
on
swapping
SD
cards,
and
that
gives
me
an
idea
that
we
have
lucky.
If
we
start
thinking
about
introducing
ipfs
somewhere
for
things
like
that,
we
we
don't.
We
we
cannot
assume
there
will
be
actual
connectivity
in
place
to
do
the
update
of
bigger
corpus
of
data.
D
However,
in
recent
OFS
release
we
had
this
import
export
functionality
added
where
data
that
was
previously
like
properly
chunked
into
smaller
pieces
and
imported
to
ipfs
repository
that
internal
format
can
be
exported
to
a
car
file
which
is
sort
of
like
a
zoom
and
that
can
be
just
put
on
SD
cards
and
then
in
a
different
place,
just
imported
type,
EFS
node.
So
that's
sort
of
like
a
sneakernet
approach
to
distributing
bigger
data
sets
that
are
already
on
ipfs.
D
B
A
What
I'm,
trying
always
trying
to
do
is
is
there
a
project
on
the
ground
that
is
doing,
work
with
teacher
education
or
empowering
children
are
working
with
refugees
and
trying
to
sneak
this
technology
enjoy
lately
since
go
with
nineteen
people
have
actually
approaching
me
because
this
I
designed
this
for
offline
world,
but
with
the
current
epidemic
we
have
people
realizing
that
this
is
a
need.
That's
there
on
the
ground,
so
the
right
now
I'm
trying
not
to
scale
this,
but
try
to
figure
out
all
the
challenges
that
I'm
facing
trying
to
address
them.
A
D
C
Absolutely
so
with
the
content,
so
open
content
try
to
combine
two
words
though
and
horribly.
How
are
you
finding
the
challenges
of
getting
like
content
producers
to
licence
that,
in
a
way
that
protects
their
interests?
Obviously,
it's
very
difficult
to
create
engaging
and
useful
content,
but
also
enables
projects
like
this
to
succeed,
about
being
strangled
by
the
costs
of.
You
know
that
every
content
creator
wanting
to
be
paid
20
dollars
per
view.
A
We
try
to
get
these
open,
licensed
content
and
Calibri
already
does
the
curation,
so
we
are
pulling
from
those
those
repositories
right
now.
The
content
that
is
that
we
are
producing.
We
are
trying
to
force
people
to
open
licensed
it
because
in
Cambodia,
for
instance,
most
of
the
content
development
is
funded
by
EU
or
one
of
the
large
funding
organization,
and
lately
they
have
been
stipulating
that
if
you
use
public
funding
for
developing
content,
it
is
recommended
to
kind
of
release
them
under
open
license.
A
So
that
is
one
kind
one
way
of
ensuring
that
people
would
release
content
that
is
open
for
all,
but
the
challenges
it's
really
difficult.
The
trend
is
either
to
package
all
the
content
they
develop
into
an
app
and
release
it
on
App
Store,
and
that
kind
of
login
always
happens,
and
it's
really
hard
for
us
to
to
kind
of
change
that
mentality.
But
using
funding
as
a
kind
of
a
weapon
to
open
knowledge
bases
is
what
we
are
trying
to
do.
A
B
Appreciate
it
there's
a
great
one,
nice
to
see
you
all,
we'll
probably
have
another
one
of
these
in
another
month,
so
feel
free
to
join.
If
you
have
a
topic,
you'd
like
to
share
on
and
speak
about,
whether
it's
a
formal
talk
or
even
just
a
conversation
topic
or
something
you
have,
a
question
would
like
to
learn
more
about
discuss
with
the
group
feel
free
to
add
it
to
the
agenda
and
we'll
pretty
soon
here,
post
a
new
comment
in
that
issue
for
next
month's
meeting.
Thanks
a
lot
all
right,
you.