►
From YouTube: IPFS 201: Track introduction - Daniel Norman
Description
Dapps are all the rave, but there's no consensus about how they should be built.
There's a cambrian explosion of developer tools and many emerging design patterns using IPFS. This introduction will set the stage for some of the fantastic speakers we have during the day who'll be sharing the state of the art in app design patterns and developer tools.
A
I
just
want
to
set
the
stage
and
I'm
going
to
split
this
intro
into
three
parts.
I
want
to
talk
a
little
bit
about
the
qualities
of
protocols.
What
is
ipfs
very
briefly
and
really
the
theme
of
today,
which
is
what
you
can
do
with
ipfs
and
so
the
qualities
of
protocols
I,
think
protocols
should
mostly
stay
out
of
the
way
of
you
know:
users
and
Developers
for
a
good
experience
and
so
informally.
A
A
Is
it
using
TCP
under
the
hood
UDP,
as
is
the
case
Now
with
HTTP,
3
and
quick,
and
so
this
is
really
like
the
power
of
abstractions
when
they're
not
leaky
and
when
they
do
a
good
job,
and
so
good
abstractions
really
make
our
life
easier,
and
you
know
arguably,
as
I
mentioned
before,
most
application
developers
don't
really
want
to
think
about
protocols.
They
just
want
to
use
them,
and
so
some
qualities
to
these
great
protocols
and
standards.
A
A
But
we're
slowly
getting
there
we're
seeing
more
implementations
that
you
know
are
accounting
for
more
platforms,
more
languages
and
more
use
cases,
and
so
some
of
you
may
have
attended
the
implementations
track
yesterday
and
you
may
have
seen
ipfs
embed,
which
is
you
know,
written
in
rust,
and
it's
really
intended
for
these
low
powered
devices
and
so
I
think
another
good.
This
sort
of
kind
of
leads
me
to
SQL
I've
been
a
long
time
kind
of
like
full
stack
developer.
A
So
to
say,
I
mean
back
in
the
day
it
was
a
webmaster
and
then
it
became
a
web
developer,
and
you
know
now
it's
a
full
stack
developer,
but
you
know
almost
throughout
my
career,
I
was
using
SQL
I
first
learned
SQL
when
I
was
about
15
and
I.
Did
an
internship
at
Vodafone
in
Australia
and
I
was
working
on
their
Billing
System
and
that's
when
I?
You
know
first
got
the
chance
to
work
with
Oracle
and
you
know:
50
SQL
is
now
50
years
old.
A
A
It's
concise,
it's
interoperable
and
it's
abstractable,
and
and
for
that
reason
you
know
if
you're
working
with
relational
databases
and
SQL,
you
have
so
many
different
options.
When
it
comes
to
abstractions,
you
can
use
raw
SQL
or
you
can
use
or
Rams
and
you
can
use
Query
Builders,
and
so
you
have
this
whole
range
of
tooling
that
you
can
build
on
top
of
this
are
standard.
A
Now
it's
hard
to
not
address
the
elephant
in
the
room.
I
think
a
lot
of
us
here
at
ipfs
camp.
We
came
here
because
we're
interested
in
decentralization-
and
you
know
some
of
the
tensions
between
centralization
and
decentralization
and
I.
Think
it's
probably
worth
saying
that
you
know
it's.
It's
probably
a
false
dichotomy.
You
know
to
say
it's,
you
know
it's
either
decentralized
or
centralized.
You
know,
most
of
us
familiar
with
the
OSI
model
or
the
tcpip
model
realized
that
you
know.
A
Centralization
is
a
spectrum,
and
you
know
you
can
make
judgments
about
things
on
different
layers
of
the
stack,
and
so
it's
never
black
and
white,
and
it
really
depends
on
which
layer
but
I
think
what
we've
seen
in
the
last
sort
of
10
to
15
years
with,
like
social
media
and
the
big
clouds
sort
of
rising.
You
have
these
Giants
and
these
Giants
they're.
You
know
driven
by
quarterly
earning
reports
and
efficiency,
and
you
know
on
the
user
side
we
care
about
convenience.
A
A
We
just
you
know,
get
a
bunch
of
you
know
Netflix
Hulu
subscriptions,
and
you
know
we
like
that
lean
back
experience
and
so
I
think
that
has
been
the
forcing
function
for
a
lot
of
that
centralization
that
we've
seen,
but
obviously
there's
also
the
counter
Force
and
that's
you
know
a
lot
of
the
people
who
are
here
today
and
so
there's
many
I,
don't
need
to
reiterate
the
benefits
of
decentralization
I'm
sure
we're
all
kind
of
on
board
with
that.
But
you
know
some
of
them.
A
I
think
that
are
worth
reiterating
are
resilience
trustlessness
and
verifiability,
and-
and
this
is
also
a
theme
in
the
ipfs
world-
we
have
these
HTTP
gateways
and
you
know
everyone
thinks
okay,
I
have
a
sid
I,
just
grab
it
from
a
Gateway,
but
are
you
verifying
you
know?
Are
you
doing
verified
retrieval,
or
are
you
just
trusting
the
Gateway
I
think
most
cases,
people
who
are
using
gateways
aren't
doing
verified,
retrieval
and
I.
Think
that's
also
for
historical
reasons,
because
we
didn't
have
verified
responses.
A
We
didn't
have
the
ability
to
get
raw
blocks
up
until
recently
as
part
of
the
Gateway
standard,
and
so
now
it's
specified,
and
now
we
have
this
ability
and
so
that's
a
sort
of
like
stressing
the
reality
of
ibfs
as
an
evolving
protocol,
and
so
this
leads
me
to
the
second
part.
Okay,
what
is
ipfs
and
I'd
love
to,
like
you
know,
ask
you
like
what
do
you
think
ipfs
is
in
like
two
to
three
words
feel
free
to
shout
out.
A
Great
I
mean
I
think
there's
many
different,
varying
definitions.
It's
interesting
to
me
always
that
it's
a
distributed
file
system
comes
up
because
I,
don't
know,
I
mean
file
system
is
in
the
name
ipfs,
but
to
me
it
was
always
a
way
to
move
around
bytes,
rather
than
actually
being
the
file
system.
So
obviously
it
is
a
file
system
and
unixfs
was
sort
of
the
first
use
case
of
ipfs,
but
I
think
that
kind
of
leads
to
a
lot
of
confusion
amongst
people
coming
to
ipfs
because
they're
like
well.
A
A
The
first
one
is
content
addressing
and
the
second
one
is
peer-to-peer
networking
and
the
best
analogy
that
I
could
come
up
with
last
night
for
content
addressing
is
like
Ikea
article
numbers.
You
see
it
doesn't
matter
which
country
you
go
to
or
which
specific
Ikea
shop
you
go
to.
The
article
number
is
going
to
be
the
same
right,
so
you
can
always
get
the
Ikea
Kira
still
anywhere
in
the
world
right
sure
you
don't
have
to
go
to
a
DHT,
but
you
know
just
replace
DHT
with
the
Ikea
database.
A
There's
also,
this
notion
of
you
know,
breaking
away
from
the
client
server
model
and
I
think
you
know
when
we're
breaking
away
from
the
server
client
model,
we're
moving
into
this
kind
of
like
distributed
peer-to-peer
web
and
the
best
analogy
that
I
could
come
up
for
that
is
that?
Well,
you
know
in
a
client
server
model,
you
have
very
sort
of
specific
roles
that
every
Network
participant
takes
you're,
either
a
client
or
a
server,
obviously
and
I.
A
Think
a
similar
sort
of
shift
happened
in
like
the
hotel
industry,
where
you
know
you
used
to
have
tourists
and
hotels
or
hotel
chains,
and
now
we
have
Airbnb,
and
you
know
you
have
one
Airbnb
account.
You
have
one
sort
of
identity
on
Airbnb
and
that
allows
you
to
be
both
a
host
or
a
guest
depending
on
whether
you
know
you're,
hosting
or
you're
traveling,
so
we're
seeing
that
with
itfs2.
And
of
course
some
of
you
are
familiar
may
be
familiar
with
the
cathedral
in
the
bazaar.
A
One
of
the
attributes
Eric
describes
in
in
great
open
source
is
that
any
tool
should
be
useful
in
the
expected
way,
but
a
truly
great
tool
lends
itself
to
users
that
you
never
expected
and
I
think
it's
probably
evident
already
at
this
point
that
ipfs
camp
that
there's
a
lot
of
weird
quirky
use
cases
for
ipfs
and
I.
Think
that's
what
makes
it
so
exciting.
A
You
know
if
we
take,
you
know
again
the
two
core
pillars:
content
addressing
and
peer-to-peer
networking
the
possibilities
of
you
know
things
that
you
can
do
is
just
these
two
concepts
are
really
kind
of
like
endless,
and
so
this
will
be
a
theme
throughout
the
day.
This
top
down
this
tension
between
top
down
and
bottom-up
Innovation.
This
is
about
pragmatism.
This
is
about
what
you
can
do
in
the
real
world
right
and
a
lot
of
that
is
bottom-up.
A
Innovation
and,
of
course,
we
have
some
of
the
great
engineers
in
the
other
spaces
who
are
doing
you
know
thinking
about
things
from
first
principles.
They're,
you
know
very,
they
love
purity
of
ideas.
They
love
great
abstractions,
but
you
know.
Sometimes
you
know
they
live
in
the
cathedrals,
we're
in
the
bazaar.
You
know
we
we've
got
to
meet
somewhere
and
I
think
this
is
where
we
meet
so
again.