►
From YouTube: IPFS All Hands 🙌🏽📞 Nov 19, 2018
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
Hey
there,
hey
I,
rely,
hi,
hi,
hey
I,
see
Hannah,
see
Terry,
I,
see
I,
see
a
lot
of
people
now
go
through
all
the
names.
Let's
begin
hello,
my
name
is
Portia
and
welcome
to
the
ipfs
all
hands
on
call
this.
The
call
where
you
get
a
chance
to
listen
to
some
of
the
exciting
projects
happening
in
the
ipfs
community
of
today's
name.
Speaker
is
Michael
Rogers
and
the
name
of
his
presentation
is
called
the
internet
of
data.
But
before
we
get
into
the
main
presentation,
we
will
start
with
announcements.
E
From
the
beads
point
from
last
week's
call
that
everyone
who
is
participating
in
okay,
ours,
please
make
sure
you're
scoring
your
mid-quarter
okay,
ours.
We
do
it
asynchronously,
but
it
helps
us
keep
track
of
all
of
the
awesome
initiatives
that
are
making
progress
and
which
things
we
need
to
put
more
wood
behind.
So
we
can
get
them
done
before
the
end
of
the
course.
So,
thanks
to
everyone
for
spending
some
time
on
that
this
week,
thanks
Molly.
E
C
Okay,
so
just
some
quick
background,
while
I
get
this
you're
working
so
yeah,
so
this
so
often
like
way
in
advance
I
have
to
like
commit
to
when
a
topic
is
for
a
particular
talk,
and
so
this
was
like
you
know
four
months
ago,
or
so
maybe
five
when
I
had
to
commit
to
Acadia
and
so
I
decided
that
I
was
going
to
trying
to
talk
that,
but
basically
taking
our
vision
for
the
next
version
of
the
decentralized
web
I'm
trying
to
boil
it
down
to
something
that,
like
everyday
web
developers,
can
understand
and
create
sort
of
a
narrative
like
why
that's
important
and
sort
of
some
of
the
primitives,
in
our
view
of
kind
of
content,
addressing
as
a
way
to
really
open
up
something.
D
C
Didn't
block,
though?
Okay,
do
you
see
again,
you
don't
see
the
zoom
tool,
cool,
okay,
all
right
so
yeah,
so
that
that
talk
was
call
up
the
future
of
the
web
vision
for
an
Internet
I.
Think
for
this
audience,
Internet
of
data
is
probably
a
little
bit
better
so
anyway.
That's
me
all
right
so
start
with
back
in
the
90s.
Even
though
this
is
a
future
talk,
we
need
to
kind
of
start
back
at
the
beginning.
C
C
D
C
Literally,
like
you
had
to
type
all
that
in
the
browser
wasn't
fun
enough
yet
to
have
for
you
so
there's
life
like
from
a
consumer
standpoint.
One
of
these
was
significantly
worse
in
terms
of
experience
but,
as
we
know
like
the
web,
one
and
and
the
reason
that
the
web
one
was
that
we
had
URLs
and
links
essentially,
so
the
starting
point
for
putting
content
on
to
a
o/l
was
like
it
a
keyword,
so
that
was
like
you
know,
be
Oprah
like
if
you're
Oprah,
you
can
get
keyword
at
Oprah.
C
The
value
here
is
the
network,
not
any
individual
piece
of
the
network
right,
okay,
so
to
the
weather
like
built
on
the
network,
effective
lakes
like
in
the
modern
era.
When
we
talk
about
the
web,
instinctively
developers,
think
about
you,
know,
standards
and
web
frameworks
and
java
scripts
and
all
the
stuff,
but
like
you
really
have
to
go
back
to
the
beginning
there
to
realize,
like
the
foundational
building
blocks
here,
are
linked
in
URLs,
so
look
at
links
and
URLs.
So
an
earl
has
a
transport
which
is
like
it.
So
actually,
let's
go
backwards.
C
C
So
that's
that's
problematic,
because
the
notice
that
like
centralization
is
baked
in
the
network
as
a
whole,
like
the
entire
web,
is
federated
because
anybody
can
put
these
up,
but
actual
URLs
are
centralized.
You
have
to
talk
to
your
particular
location.
Comes
applies
to
more
than
just
the
HTML
for
that
content.
To
some
extent,
HTML
and
static
assets
are
really
a
big
problem
like
when
I
pull
up,
twitter.com
or
facebook.com
I
can
and
do
really
manipulate
that
page,
because
once
it's
in
the
browser
and
as
part
of
the
open
web
I
can
remove
ads.
C
I
can
like
change,
works
like
that
stress
me
out,
like
Trump's
words
that
don't
like
daisies,
like
you
literally,
do
that
in
browsers,
amazing,
but
the
web,
and
whether
the
locations
are
much
more
than
static
assets
like
modern
web
applications
pull
a
ton
of
their
content.
Information
in
dynamically
from
web
services,
and
these
web
services
are
all
like
addressed
by
these
URLs.
So
they
are
centralized,
and
so
what
you
end
up
getting
is
like
data
centralization
kind
of
baked
in,
even
though
you
can.
D
C
To
multiple
services
and
like
holding
the
contact,
many
where
anybody
can
put
up
a
website
if
I
want
to
have
relationships
between
two
pieces
of
data,
I
have
to
put
that
in
a
content
silo
behind
an
API,
and
the
thing
about
this
is
that
there's
no
good
way
right
now
to
not
build
this
way.
If
I
want
to
make
a
multi-user
application,
I
essentially
have
to
go
and
build
a
data
silo
so
like
every
web
developer.
C
It's
not
ten
years
per
execution
and
it
only
grows
over
time
and
so
because
applications
cost
money
to
build,
especially
multi-user
applications.
That's
where
we
have
to
basically
create
actors
in
business
models,
as
businesses
is
a
primary
business
on
the
web,
because
in
order
to
participate
in
those
network
effects
that
we
saw
before
you
essentially
have
to
have
the
content
open
to
the
public
and
linkable
and
so
having
content.
C
The
hot,
like
behind
some
kind
of
pay
wall
makes
is
very,
very
hard
to
participate
in
the
network
effects,
so
that
really
incentivizes
people
to
create
a
driven
business
models.
And
what
we
see
now
is
that
on
a
long
enough
timeline,
these
ad
models,
like
they
start
adding
machine
learning,
algorithm,
learn
and
what
they
learn
is
that
the
best
way
to
continue
to
attract
our
attention
is
to
push
us
towards
more
and
more
extra
content,
and
so
essentially,
like
radicalization,
is
sort
of
like
the
end
state
of
a
driven
business
on
the
web.
C
It's
what
they
inevitably
get
to,
because
it
is
just
a
most
efficient
way
to
make
us
all
watch
YouTube
all
day
and
we
have
like.
Unlike
you
know,
static
assets
that
can
loaded
in
the
browser
that
you
can
manipulate.
We
have
like
no
visibility
in
these
algorithms.
We
can't
change
the
data
that
comes
out
of
them.
C
They
just
don't
participate
in
the
open
web
in
the
way
that
you
know
the
rest
of
the
resources
that
are
standardized
do
okay,
so,
like
breakfast
Rick
is
in
a
little
bit
more,
like
you
have
web
application
to
talk
to
multiple
api's.
All
those
api's
talk
to
different
databases
right.
So
if
we
want
to
get
rid
of
just
this
API
layer
that
is
so
problematic,
we
have
to
look
at.
You
know
what
we
might
need
to
connect
you,
what
we
might
need
to
replace
once
you
start
to
drill
into
this.
C
What
these
databases
do
is
they
take
data
structures
in
some
kind
of
serialized
form,
and
then
they
actually
serialize
them
again
to
disk?
Or
often
you
know
to
a
network
of
servers
that
are
actually
writing
them
to
disk
and
they
provide
certain
consistency,
guarantees
both
on
the
read
and
write
edge,
and
it's
it's
hard
if
you
haven't
done
database
engineering
to
really
understand,
but
API
calls
that
users
make
when
they
make
just
very
simple
API
is
really
depend
on
and
take
for
granted
the
consistency,
guarantees
of
these
data
structures
and
databases
there.
C
There
are
literally
sort
of
on
disk
guarantees
at
like
the
eff
sink
layer
and
a
lot
of
databases
that
bubble
all
the
way
up
to
the
end
of
an
API
call.
And
if
you
want
to
start
replacing
the
API,
is
you
literally
have
to
go
all
the
way
down
and
start
to
think
about?
Okay?
What
do
we
do
when
we
take
a
data
structure
in
memory
and
it
has
to
leave
memory?
C
If
we,
if
we
have
like
a
global
decentralized
network,
we're
not
going
to
have
this
namespace
own
website
anymore,
and
so
we
essentially
can
hash
the
content,
and
this
is
where
we
start
to
get
into
a
larger
view
of
hashing
I.
Think
it
been
around
for
a
very
very
long
time,
but
it
seems
like
all
the
time,
we're
figuring
out
new
ways
to
work
with
hash
content,
new
primitives
that
we
can
build
on
top
of
hash
a
it's
an
incredibly
powerful
concept
and
I
promise
I'm
not
going
to
talk
about
watching.
C
That's
an
important
part
of
the
talks
that
we
don't
asleep
watching
Oh
cut.
So
this
is
content
accessibility.
Basically,
if
we
start
with
rather
than
referring
to
things
by,
you
know
URLs
or
human
readable
names
that
we
create
if
we
instead
say
okay,
the
way
that
we're
going
to
refer
to
data
is
by
the
hash
of
that
data.
Then
we
get
some
really
interesting
benefits
out
of
that
one.
C
Is
that,
like
this,
primitive
doesn't
require
us
to
centralize
the
data
data
can
come
from
anywhere
and
we
have
a
way
to
graphically
to
verify
that
we
are
talking
about
the
same
data
and
that
I
got
the
right
data.
So
this
proof
is
really
important
because
there's
a
lot
but
really
liberate
us
from
having
to
talk
to
an
authority
or
do
some
kind
of
access,
control
or,
like
you
know,
authentication
dance,
because
if
I
have
the
hash
of
data
and
I
get
it
from
literally
anybody
over
any
transport,
it
doesn't
matter
how
I
get
it.
C
I
needed
that
we
worked
and
we
were
talking
about
the
same
thing:
okay.
So
to
give
like
a
very
kind
of
high-level
view
of
what
this
looks
like
in
sort
of
a
pseudo
code,
we
at
protocol,
as
we
have
a
primitive
called
C
ID
or
a
Content
ID,
which
is
a
hash
with
like
a
little
bit
of
extra
data
in
front
of
it.
That
tells
you
how
to
interpret
it,
how
to
pull
additional
links
out
of
it.
C
So
if
I
you
know
published,
a
blog,
I
would
have
like
the
title
of
that
blog
and
some
posts,
and
then
those
would
essentially
have
C
ideas
in
them
that
are
links
to
that
content
and
allies
network.
And
so
then
I
would
say:
oh
hey,
yeah
I
go
I
need
to
go,
get
that
hashed
out
and
I
literally
get
this
hash
from
anywhere
anybody.
Any
transport
I
can
have
it
in
a
local
cache
indefinitely
offline.
C
I
could
have
it
in
you
know:
a
local
network
not
even
connected
to
the
Internet
I
could
grab
it
of
the
peer-to-peer
network.
I
could
get
out
of
a
CDN
anyway.
That
is
the
fastest
and
simplest
way
to
get
this
content
doesn't
matter
how
I
get
it
and
then,
of
course,
Sally
creates
a
list
of
her
favorite
blogs.
C
Not
all
the
web
sites
are
in
the
same
place,
but
we
can
still
link
to
each
other
and
create
a
network
of
value,
and
here
we
actually
have
data
structures
like
you
know
this
aggregator,
you
know
points
to
the
data
in
my
blog
directly,
but
somebody
else
is
hosting
it
or
maybe
I'm
hosting
all
of
this
I
pulled
it
all
into
a
local
cache
to
work
offline.
All
of
that
works.
We
don't
have
to
change
the
data
structure.
We
don't
have
to
do
any
kind
of
extra
dance.
C
This
is
you
know,
I've
done
offline
stuff
for
a
really
long
time.
This
is
sort
of
this
is
really
interesting,
where
you
can
actually
stop
thinking
about
authenticating
to
a
place
and
then
in
your
offline
cache
saying
oh
well.
This
is
like
the
authenticated
data
that
I
got
from
this
particular
location.
You
don't
have
to
keep
any
kind
of
reference
around
for
how
you
got
that
data.
It
doesn't
really
matter,
it's
all
stash.
C
C
C
You
can
start
to
imagine
how
to
build
a
b-tree
IQ
figure
if
I
had
to
do
that,
google
interview
and
then
never
use
any
of
that
data.
Again,
you
can
actually
like
Bill.
You
know
these
fancy
data
structures
that
are
very
big
and
we
can
start
to
recreate
what
databases
do
in
a
very
siloed
centralized
way
for
the
web,
and
this
is
a
quote.
C
This
is
Michael
and
I
said
this,
but
actually
after
I
said
this
is
like
this
is
just
for
this
audience
it
into
the
Cascadia
after
the
first
time,
I
set
this
out
loud.
It
was
at
once
house
with
Stephen
and
right
afterwards.
Stephen
corrected
me
instead,
no
they're,
like
pointers
in
rust,
because
they
can
only
be
for
it
was
like
that's
true,
okay,
anyway.
Okay,
all
right,
so
so,
but
the
end
state
here
is
that
we
again,
you
know
we
create
something
like
the
web
for
data
we
can
start
to.
C
The
thing
that
we're
not
doing
it,
implying
that,
like
everything,
we've
done
so
far,
is
the
right
way
to
do
things
and
that
there's
a
new
there's,
a
new
set
of
things
that
we
need
to
do
is
that
you're
solving
for,
but
if
you
just
like
recreated,
Twitter,
recreated
Facebook
or
something
like
that,
and
it
could
participate
in
the
network
effects
of
this
data.
You
we
blow.
C
You
know
Twitter
and
Facebook
out
of
the
water,
because
the
the
value
of
that
data
is
so
much
more
valuable
and
like
to
public
into
everybody,
then
Twitter
or
Facebook,
because
it
is
not
locked
up
in
this
content
silo
like
we
basically
start
to
create
this
Internet
of
data
and
that's
what
we
want
to
push
for
and
so
yeah,
so
we're
working
on
this
and
I
feel
be
we're
running.
Some
respects
were
literally
in
the
earliest
stages
possible
of
this.
C
B
B
B
A
In
terms
of
like
the
relationship
with
blockchain
I
think
it's
almost
better
to
talk
about
blockchain
and
declare.
This
is
not
explaining
blockchains.
This
is
explaining
some
technical
principles
that
blockchains
do
use,
but
you
could
use
them
too
and
like
make
that
case,
because
I
think
a
lot
of
people
are
curious.
They
want
to
know
more
about
blockchain,
so
showing
like
hey.
This
is
this
is
connecting
with
that
domain,
but
is
just
using
the
parts
that
are
useful
without
committing
to
a
whole
like
mindset
of
something
that
might
be
a
fad.
C
A
C
Mean
obviously
like
a
lot
of
people
that
are
familiar
with
content,
addressing
are
familiar
with
blockchain
like
that's
how
they're
familiar
with
it
I
think
that
the
one
the
hard
parts
of
the
hard
thing
that
breaking
that
out
is
they
think
of
content,
interesting
in,
like
a
like
a
stylet,
almost
namespace
way
right
where
it's
like.
Oh,
this
is
a
great
way
to
address
content
like
in
the
blockchain
or
in
the
Bitcoin
network
or
in
the
etherion
Network.
C
D
I
was
wondering
D
Phillips,
you
know,
or
how
important
is
it
in
these
sort
of
the
intro
talk
stage
versus
later
on,
and
they
get?
You
know
more
in-depth
to
discuss
sort
of
the
difference
between
we'll
call
it
like
paying
for
everything
and
things
being
free,
both
from
sort
of
in
both
directions.
Most
of
the
fact
that
we
get
everything
for
free
because
of
ads,
but
also
and
in
a
decentralized
way.
Maybe
it's
a
little
different
and
the
opposite.
C
That
is
actually
a
possibility
right
and
like
I
know
that
Google
has
people
like
looking
at
the
centralized
web
technology
and
sort
of
the
built
indexes
of
around
it,
and
what
they're
probably
doing
is
just
taking
all
of
the
hashes
ever
and
putting
them
into
a
giant,
centralized
store
and
then
look
and
then
examining
the
links
of
them,
and
so
I
think
that,
like
once,
we
have
a
real
flow
for
people
to
make
some
structured
data
and
then
publish
it
and
host
it.
C
We
have
a
whole
workflow
around
that,
where
you
know
here's
how
you
host
it
by
yourself
and
definitely
here's
how
you
pay
somebody
else
to
host
it.
Once
we
have
that
workflow,
then
we
can
start
to
talk
about
how
people
get
it
and
how
they
keep
it
alive
indefinitely,
and
what
the
cost
structure
is
for
that,
but
like
until
we
happen
until
we're
literally
demoing
that
flow
I
would
rather
not
check
to
get
into
those
issues
right.
G
C
So
obviously,
things
are
really
early,
so
we
want
to
make
sure
that
we
don't
try
to
standardize
too
early.
That
said,
like
there
are
some
parts
of
our
stacks
that
we
know
aren't
changing
right,
like
multi
hash
is
not
changing
and
and
if
it
is
like
that
would
be
like
a
big
breaking
change
and
effectively
like
a
new
standard.
So
we
there's
there's
some
low
level
stuff
that
we
do
want
to
start
to
get
on
standards
track,
but
probably
not
that
at
the
w3c
at
the
IETF.
C
Instead,
the
ITF,
it's
just
like
a
bit
easier
to
work
with,
for
a
whole
variety
of
reasons,
and
a
lot
of
like
web
standards
are
at
the
ITF
by
the
way.
So
it's
not
like
out
of
the
room
cost
like
yeah.
It's
only
only
a
section
of
web
standards,
mostly
all
the
Dom
stuff,
is
only
at
the
w3c
/wg
like
HTTP
to
most
cryptography
stuff
is
all
I
get
the
IETF,
so
it
kind
of
makes
sense
for
us
to
go
that
route.
F
Hey
I'm
just
curious
if
you
got
any
questions
about
private
data,
authenticated
data
just
because
someone
who
normally
builds
web
applications
when
I'm
not
working,
I,
PFS,
most
things
and
have
an
authentication
system
and
then
privatized
data.
C
Yeah
yeah
I
think
that
the
the
analogous
part
of
that
story
for
from
my
stock,
like
people
did
ask
about
that
afterwards
and
I
just
said
like
you
need
to
encrypt
the
data
like
frankly
like
the
issue
was
having
like
globally,
you.
C
C
B
Okay,
excellent.
That
concludes
our
presentation
and
before
you,
log
off
I
want
to
I
was
having
problems
with
my
Wi-Fi
connection.
So
I
am
sorry
if
you
were
not
able
if
it's
cut
out
so
I'll,
be
sure
to
address
that
problem
for
next
week.
Thank
you
very
much
and
I'll
see
you
next
week.
Bye.