►
Description
Browsers 3000 is a six-week virtual event to explore and accelerate the development of web3 in the browser through experimentation and challenge prizes of up to $20,000 worth of Filecoin (FIL). Learn more at https://events.protocol.ai/2021/browsers3000.
Like all IPFS Community events, Browsers 3000 expects all speakers and participants to follow the IPFS Community Code of Conduct (https://github.com/ipfs/community/blob/master/code-of-conduct.md).
A
Hey,
thank
you.
I'm
going
to
talk
a
little
bit
about
ion.
This
is
the
centralized
identifier
network
that
we
we
built
collaborated
with
others
to
build
actually
not
at
microsoft,
at
decentralized
any
foundation,
but
we
were
major
contributors,
let's
dig
in
so
understanding.
The
problem
first
thing
to
understand
about
decentralized
identifiers
is
like
what's
the
actual
issue
that
they're
trying
to
solve.
You
know
we
don't
just
want
technology
for
technology
sake.
Two
types
of
identity
data
will
help
sort
of
focus
on
the
issue
of
decentralized
identifiers.
A
So
there's
some
issues
with
that
right:
the
status
quo
of
identity.
Your
identifiers
can
be
snapped
out
of
existence
right.
I
can
just
be
dusted
anytime,
a
provider
wants
your
connections,
go
with
that.
So,
if
you
have,
you
know
your
your
id
severed,
email
address
or
a
username
and
an
app
you
know
like
twitter
or
something
like
that,
all
the
connections
and
followers
and
all
the
stuff.
You
built
the
relationships
you
built
over
time.
A
A
So
it's
really
it's
really
important
that
we
get
this
right
because
we're
seeing
this
more
and
more
in
the
world
today
where
the
answer
seems
to
be
because
we've
co-mingled
identity
is
that
we
sever
people
from
what's
increasingly
becoming
more
important,
which
is
your
digital,
your
digital
persona.
A
So
this
isn't
hypothetical
and
I
know
you've
all
seen
it
tons
of
people
are
getting.
You
know
severed
from
twitter,
which
you
know,
deletes
all
their
contacts
and
everything
else,
and
that's
just
really
not
acceptable.
Interestingly
enough,
in
the
time
since
I
gave
this
talk
with
this,
this
one
slide
that
happened
to
be
in
it
that
I've
presented
before
I
was
actually
suspended
from
twitter
for
12
hours.
A
So
I
got
a
little
taste
of
it,
even
though
I've
talked
about
it
for
a
long
time,
and
I've
seen
you
know,
friends
who
are
very
much
mainstream,
get
suspended,
and
I'm
sure
you
all
know
folks,
it's
it's
a
little.
It's
shitty
to
have
happened
to
you
and
you
don't
want
that
to
be
the
case
and
it's
all,
because
we've
co-mingled
identity
with
apps,
because
we
don't
have
a
digital
identity
layer
for
the
web,
so
apps
become
this
sort
of
de
facto
identity,
which
is
is
really
not
how
things
should
be.
A
So
how
do
we
fix
this?
Well,
we
have
this
really
cool
new
spec.
That's
just
about
finished!
It's
going
to
be
an
international
standard
here
in
a
matter
of
months,
called
w3c,
decentralized
identifiers,
it's
it's
codified
in
the
same
body
that
codifies
web
technologies
and
browser
standards
and
the
kind
that
dietrich
and
I
used
to
work
on
mozilla.
A
So
it's
it's.
You
know
it's
a
robust
standard.
It's
been
worked
on
for
many
many
years
over
half
a
decade
and
finally,
you
know
arriving
and
essentially
what
it
is,
is
the
ability
to
give
you
identifiers
that
are
yours
that
are
truly
owned
and
controlled
by
you.
It
defines
a
data
model
and
a
uri
scheme
which
you
can
see
over
there
on
the
right
of
my
screen.
The
uri
scheme
is
basically
did
colon.
A
You
know
like
you
would
have
http
colon,
that's
where
I
fast
cool
and
that
sort
of
thing
then
you've
got
your
method
portion.
The
important
piece
there
you
see
ion
is
that
there's
not
just
ion,
isn't
the
only
type
of
decentralized
identifier
that
could
adhere
to
the
spec.
There
could
be
other
constructions
as
well,
but
ion's,
one
of
them
and
that's
what
that
segment
means,
and
then
the
other
gobbledygook
on
the
right
side
is
the
actual
identifier
itself
and
these
identifiers
are
backed
with
cryptographic
keys
and
they
also
can
include
things
like
routing
endpoints.
A
So
it's
not
about
identity
data.
That's
the
one
big
thing
I
want
to
you
know
like
we
talked
about
the
first
slide.
There's
your
identity
data
there's
identifiers
ion
is
not
a
system
of
putting
your
identity
data
somewhere
public.
It
is
a
system
of
identifiers.
So
it's.
How
do
you
prove
you
own
this
string
and
then,
where?
What
are
the
routing
endpoints
that
I
could
go
to
find
data?
You
know
in
you,
know
private
data
or
communicate
with
you.
A
So
what
what's
the
real
core
like
technical
issue
and
why
aren't
dids
like
why
haven't
they
been
around
for
a
long
time?
Well,
there's
actually
like
this
really
really
interesting
problem.
That's
been
around
for
40
years.
You
know
really,
since
the
dawn
of
like
digital
identity
itself
and
the
ca
system,
which
is
you,
don't
really
have
the
ability
to
know
what
keys
are
related
to
an
identifier.
A
What
certificates
you
know
in
the
jurisdiction
are
related
to
an
identifier
like
a
domain
in
a
decentralized
way,
so
we've
had
to
create
these
hierarchies
and
the
reason
why
is
because,
as
we
see
here
in
the
example
of
the
identifier
theseus
there's
no
there's
never
been
a
system
of
record
that
keeps
track
of
historical
change,
linear,
historical
change,
that's
decentralized
and
immutable.
We've
never
had
that
before
right
and
that's
that's
actually
what
satoshi
solved
with
bitcoin
was
was
creating
this
thing.
A
Up
to
that
point,
we
had
ca
systems
that
were
based
on
hierarchies
and
central
authorities,
and
someone
saying
well
this
domain
name's
associated
with
this
set
of
you
know
set
of
cryptographic
material
because
I
say
so
right.
You
don't
have
to
know
the
history,
it's
just
whenever
I
say
it's
different.
It's
different,
not
a
great
thing
for
basing
like
human
identity
on,
because
that's
a
human
rights
risk,
in
my
opinion.
So
what
we
see
here
in
you
know
this
example
of
theseus
is
when
an
id
starts
out.
A
It
could
have
one
set
of
keys
and
one
set
of
endpoints
with
it
right
and
theseus.
Is
this
greek
hero
he
left
on
a
journey
with
the
ship
and
over
the
course
of
his
journey,
everything
about
the
ship's
change
so
like
every
nail
and
board
everything
about
it.
So
it
looks,
looks
very
different.
So
over
the
course
of
time
you
know
you,
roll
rotating
keys,
you're,
switching
out
end
points,
and
then
you
know
far
down
the
track.
Nothing
about
your
your
did
or
its
lineage.
A
Its
secure
lineage
is
the
same
right,
so
we
haven't
had
that
ability
to
track
that.
So
in
this
case,
in
this
example,
when
these
ship
comes
back
to
the
dock,
if
you
imagine
someone
who
saw
it
saw
it
when
it
left,
they
would
question
and
say
well.
Is
this
the
ship?
I
don't
know,
I
can't
securely
understand
or
know,
because
I
wasn't
there
when
every
board
and
plank
was
replaced.
So
that's
what
you
need
for
pki.
A
You
need
something
that
can
track
the
backbone
of
key
lineage,
and
you
know
what
routing
points
are
supposed
to
be
associated
at
any
given
time.
So,
let's
talk
a
little
bit
about
ion
how
it
might
solve
this.
It's
an
open
public
and
permissionless
layer,
2
network
that
runs
atop
bitcoin,
provides
robust
decentralization
interdiction,
resistance
and
scale,
so
we
can
do
did
scale
right
on
top
of
of
bitcoin.
A
A
We
didn't
want
any
of
these
things
when
we
built
it
because
we
wanted
it
to
be
completely
a
public
utility
and
control
controlled
by
users.
What
is
it
at
ten
thousand
feet?
And
this
is
where
we
start
getting
into
the
ipfs
portion
of
this?
So
if
you
look
at
this
sort
of
layout
here,
you've
got
a
couple
different
ion
nodes.
If
we
assume
there's
only
two
ion
nodes
in
the
world,
this
is
kind
of
a
you
know
how
it
would
look
first
ion
node.
A
Could
batch
one
or
multiple
operations
so
creates
of
ids
updates
of
keys
that
sort
of
thing
together,
and
this
could
be
anyone.
This
could
be
alice
or
house.
This
could
be
like
microsoft
running
a
node.
Whatever
the
only
point
of
aggregation
here
is
that
you
know
some
people
might
not
want
to
be
running
bitcoin
or
doing
cryptocurrency
transactions
or
paying
so
you
can
aggregate
them.
People
can
do
this
for
you
all
the
transactions.
The
operations
are
signed
before
they
leave
your
your
device.
A
So
if
I
as
alice,
wanted
to
have
microsoft
anchor
one
of
these
operations,
for
me,
it's
not
like
they
can
mutate
it.
They
can't,
like
you
know,
do
anything
insecure.
They
can't
violate.
You
know
any
part
of
the
the
operation
itself,
because
it's
cryptographically
secured
on
the
device
once
those
are
assembled
by
ion
node
one
into
a
batch
they're,
anchored
into
bitcoin
with
an
ipfs
c
id.
So
what
ion
one
did?
A
Was
it
sort
of
assembled
this
file
structure,
a
few
different
files,
and
then
it
put
it
together
in
through
some
linkages
and
anchored
that
in
in
a
cid
in
bitcoin
and
what
all
the
other
nodes
are
doing
is
they're
all
watching
bitcoin.
So
when
they
first
start
up
they're
like
going
to
the
block
number,
that's
the
starting
number
for
ion
and
they're
essentially
reading
forward,
and
then
they
continue,
observing
and
they're.
A
Looking
for
these
cid,
these
marked
transactions
in
bitcoin
and
when
they
find
them,
they
fetch
the
data
and
then
circulate
it
through
the
you
know,
ipfs
and
underlying
lib
pdp
protocols
and
get
it
and
process
it
all
the
same.
So
it's
a
it's
a
deterministic
system,
there's
no
agreement
between
the
iom
nodes,
there's
no
secondary
consensus.
It
is
purely
data
and
deterministic
determinism.
A
A
So
what
ion
delivers
massive
scale
right?
We
can
do
thousands
of
pk
operations.
Second,
you
know
tens
of
billions
operations
annually.
We
can
support
the
entire
globe
if
we
needed
to
just
on
ion
cost
efficiency.
You
know
even
people
ask
well
you
know
what
about
the
fees-
and
I
say
you
know
even
a
hundred
dollars
per
bitcoin
transaction.
A
A
So
that's
you
know
one
cent
pretty
cheap
right.
They
probably
charge
you
more
for
the
the
sim
card
inside
your
plan,
decentralization,
it's
actually
decentralized,
unlike
some
other
protocols
out
there
that
you
might,
you
know,
see,
there's
no,
like
rug,
pull
smart
contract.
You
know
master
key.
There's
none
of
that
stuff
iron
nodes
are
lightweight;
they
can
run
on
raspberry
pi's,
they're
they're,
not
bottlenecked.
On
cpu.
It
doesn't
take
you
a
long
time
to
start
up
a
node.
A
It's
mostly
storage.
The
system
is,
is
a
lot
ipfs
a
little
bit
bitcoin
and
then
some
processing
logic.
A
So
it's
an
intrinsic
utility
consumer
ion
consumes
the
most
intrinsic
utility
of
bitcoin,
which
is
just
the
actual
transaction
space.
It
doesn't
need
a
whole
lot
for
for
about
a
hundred
transactions.
We
could
support
the
globe
with
50
billion
diadeops
per
year.
A
A
The
foundation
of
verifiable
credentials
is
the
ability
to
sign
a
proof
from
someone
to
someone
else
right.
In
this
case,
we've
got
woodgrove
an
issuer
of
an
employment
credential,
for
instance,
and
they're
going
to
sign
this
credential
to
their
employees.
Their
employee
can
go
anywhere,
they
want
and
prove
that
they
are.
In
fact,
an
employee
of
this
place
ion
in
this
case
is
just
used
to
look
up
the
public
keys
behind
the
ids
and
do
the
signing
right.
A
So
it's
it's
really
there
for
that
pki
substrate
another
one
that
we're
working
on
dietrich
is
you
know
we
talk
to
a
lot
of
the
groups.
Frequently
you
know,
groups
like
textile
and
fission,
and
ceramic
is
personal
data
stores.
So
when
I
talked
earlier
about,
dids
can
have
associated
with
them
routing
endpoints
well
to
where
right.
What
are
those
routing?
Endpoints
too?
We
think
in
the
future.
A
It's
going
to
be
these
personal
data
stores
and
if
you
know
someone's
id,
you
can
look
up
the
routing
endpoint
and
you
can
send
messages
and
other
exchange
other
pieces
of
data
with
them.
A
couple
use
cases
I'll
quickly
go
over
hey
like
so
signal
and
telegram
are
awesome
right,
but
wouldn't
it
be
better
if
we
just
had
you
know
a
standard
encrypted
messaging
layer
for
the
web,
I
think
it
would
be
and
that's
exactly
what
did's
most
personal
data
stores
are.
A
If
you
know
someone's
id,
you
can
encrypt
the
message
by
looking
up
their
keys,
send
them
the
message
to
their
data
store
and
now
there's
a
substrate,
that's
common,
so
you
could
build
apps
like
signal
and
telegram
and
others
on
a
common
infrastructure.
That's
standardized
where
the
app
itself
is
mostly
just
ui
and
sort
of
some
affordances.
You
build
around
it,
but
the
underlying
layer
isn't
like
a
per
app
sort
of
infrastructure
and
is
much
harder
to
block
in
places
that
want
to
like
suppress
communication.
A
Another
one
that
we're
doing
right
now
is
you
know
the
ability
to
go
to
an
id
and
fetch
some
data.
That
proves
maybe
what
it
is.
So
in
this
case
a
school
right.
Do
you
know
if
it's
accredited?
Well,
if
you
find
the
did
the
school-
and
you
want
to
go
to
those
endpoints
and
go
to
its
data,
store
and
say,
hey,
give
me
this
accreditation
credential
right.
A
A
A
A
A
The
default
is
our
microsoft
server
because
we're
running
a
node
that
resolves
in
anchors
so
we're
covering
the
cost
of
the
bitcoin
anchoring
right
now
and
we're
doing
the
resolution
and
that's
fine.
You
could
put
your
own
endpoint
in
if
you
want
like,
if
you
run
your
own
ion
node,
but
we're
just
doing
this
as
a
help
to
the
community,
so
people
can
get
started
as
quickly
as
possible
and
just
to
show
you
it's
not
bs.
Here's
me
creating
a
bunch
of
ion
dids.
A
You
can
see
here
we're
just
running
through
them
and
I
might
like
say
you
know,
I'm
gonna
put
this
jws
and
test
it
out,
and
you
know
it's
it's
basically
generating
jws.
It's
signing
it
and
then
verifying
that.
In
fact
the
word
test
was
signed
over
by
the
did.
That's
been
created
down
here,
so
you
have
all
these
capabilities
in
a
library.
That's
very
easy
to
use
in
the
browser-
and
you
know
that's
basically
iron
in
a
nutshell:
we'd
love.
You
know
your
support
and
you
know
get
involved.
It's
done.