►
From YouTube: IPFS All Hands 🙌🏽📞 Aug 20, 2018
Description
This week we talk about the WHATWG HTML Spec, the Underlay Workshop, as well as demos around libdweb, and ipfs-cluster-ctl!
https://github.com/ipfs/pm/issues/680
A
A
A
So
if
somebody
would
not
mind
taking
over
this
job
and
just
sitting
here
and
corralling
people
in
20
different
people
to
talk,
that
would
be
great,
there's
a
link
to
the
issue
in
github.
That
is
an
overview
of
your
responsibilities,
mainly
it's
just
make
sure
you're
here
on
time
and
post
a
notice
to
IRC
beforehand,
but
let
me
know
on
slack
or
IRC
or
github
if
you'd
like
to
take
over
and
just
does
it
make
you
as
the
person
for
next
week.
B
Basically,
this
is
a
list
for
used
by
API
such
as
navigator
register
protocol
handler
that
lets
a
website
to
register
redirect
based
for
arbitrary
protocols.
That
way,
you
can
support
like
ipfs
:,
/
/
without
I,
like
actual
native
support,
still
the
URL.
If
you,
if
the
browser
finds
that
URL
on
the
page,
it
will
work,
it
will
redirect
to
HTTP
POST
reader.
So
it's
like
an
intermediary
step,
but
if
you
introduce
our
protocols
into
the
HTML
spec,
that
may
opens
a
lot
of
doors
because
we
gain
legitimacy.
B
B
So
I'll
post
I
posted
the
link
to
the
issue
to
the
chat
and
I
will
add
it
to
the
notes
and
it's
just
a
quick
shout
out
that
we
we
have
intended
to
implement
from
the
chromium
team
and
we
have
already
one
positive
review
and
basically
I,
don't
think
anyone
opposes
that
there
was
a
huge
blocker
for
like.
Historically,
if
you
had
wanted
to
change
this
list,
there
was
like
an
opposition.
The
spec
body
was
waiting
for
vendors
to
implement
change
and
vendors
did.
B
A
C
C
If
you
know
what
freebase
was,
which
has
now
become
Google
knowledge
graph,
it
was
a
public
open
database
where
people
were
building
semantic
information,
they're
building
like
a
graph
graph
database
of
knowledge
for
the
web,
that
Google
then
acquired
and
absorbed
and
made
private
a
few
years
ago
and
turned
into
Google
knowledge
graph.
This
is
a
project.
C
This
is
an
initiative
or
project
or
collaboration
to
basically
reboot
that
as
an
open
data
set,
that's
designed
to
be
locked
open,
it's
not
certain
to
use
the
IP
FS
right
now,
the
current
sort
of
prototypes
that
they're
building
are
using
IP,
LD
and-
and
so
this
workshop
was
like
an
initial
visioning
workshop
to
sort
of
think
about
the
data
models
involved,
so
it
wasn't
even
getting
really
into
the
functional
details
of
how
it
work.
It
was
just
looking
at
the
data
models
and
considering
the
edge
cases
and
structures
that
are
necessary
there,
but
yeah.
C
It
was
really
cool
workshop
that
happened
at
the
MIT
Media
Lab,
and
the
project
is
a
collaboration
between
the
MIT
press
and
protocol
labs
and
I'm
no
I'm
blanking
on
the
other
participants.
That's
that's
bad!
So
it's
a
really
interesting
project.
You
should
read
up
on
it
and
if
you
have
thoughts
on
how
knowledge
for
the
web
should
be
structured
or
cannulating,
you
should
jump
in
and
chime
in.
C
C
And
is
especially
if
you're
using
IPL
D
this
project
as
part
of
this
as
part
of
these
prototypes
that
are
using
IPL
T,
they
are
working
with
IPL,
D
and
json-ld.
So
it's
this
sort
of
this
spanning
this
link
data
IPL
the
kind
of
space
which
is
something
that's
been
hovering
there,
as
someone
should
take
a
stab
at
that
for
like
two
years
and
they're
stabbing
at
it.
C
D
But
it
means
that
a
ipfs
node
running
in
a
web
extension
like
an
IP
address.
Ipf
estimate
node
like
running
as
an
embedded
node
in
companion,
can
use
mdns
and
it
can
listen.
It
can
create
a
TCP
socket,
listen
on
it.
Make
connections
to
other
nodes,
discover
them
on
the
local
network
and
generally
do
stuff.
D
You'd
normally
only
be
able
to
do
with
a
GFI
TFS
node
running
in
node.js.
So
it's
pretty
damn
awesome
that
that
library
exists
and
I've
just
plugged
a
hut,
a
few
bits
and
pieces
together
to
make
it
work.
But
what
that
video
shows
is
two
nodes
running
in
two
different
browsers,
discovering
each
other
over
mdns
connecting
to
each
other
and
then
adding
some
data
to
one
of
them
and
cutting
it
out
from
the
other
one
and
the
communication
happening
just
using
p2p.
D
So
we
remove
all
of
the
other
like
discovery
and
transports
from
the
node,
so
the
only
discovery
in
transport
that
it
can
use
are
the
new
ones
that
are
provided
by
Lib
D
web.
So
it's
all
good
stuff
take
a
look.
It
might
take
a
while
and
I
do
a
lot
of
talking
there
as
well,
so
I'm,
sorry
that
rad,
that's
about
all
I've,
got
to
say
I
think
if
anyone
has
any
questions
and
I
will
take
them
now.
A
Okay
I,
so
so
I
may
not
totally
understand
all
the
ins
and
outs
of
this
properly
because
I
know
there's
some
sort
of
issues
with
with
broader
participation
and
discovery
in
the
DHT,
which
is
ipfs.
Does
this
sort
of
pave
the
way
for
improving
some
of
that
or
are
there
other
major
functionalities?
Just
helps
pave
the
way
for
in
Jos
ipfs,
in
communicating
better
with
go
ideas.
D
So
this
basically
opens
up
the
ability
for
Jess
ipfs
to
talk
to
other
nodes
and
like
I,
should
probably
say
that
live
D,
where
there's
a
possibility
that
it
might
as
a
library
make
it
into
Firefox
and
if
it
gets
into
Firefox,
then
that
will
be
super
rad
because
it
might
also
make
it
into
other
browsers.
And
then
it
means
the
way
one
step
closer
to
a
fully
peer-to-peer
web
with
I
professed
nodes
running
in
the
browser.
C
D
C
Question
is
there
anything
we
should
be
doing
too,
because
this
is
the
Lib
D
web
effort
and
the
work
you're
doing
with
it
is
part
of
this,
and
what
and
the
stuff
that
that
Lydell
was
just
updating
about.
Is
this
us
making
solid
stabs
in
this
effort
to
like
upgrade
the
protocols
of
the
web
and
update
upgrade
the
browsers?
And
it's
there's
a
lot
of
there's
a
lot
of
like
organizations
and
people
and
perspectives
involved?
D
B
Just
to
talk
quick
remarks
on
that,
one
thing
is
that
we
probably
could
make
a
blog
post
on
like
block
IPF
Sao
generally
about
Liberty
web
I.
Think
when
we
polish
protocol
handler
and
the
discovery
and
transport
to
the
point
when
it's
like
one
command
and
runs,
we
probably
could
make
a
blog
post
just
so
that
it's
something
we
can
link
to
and
Mozilla
really
appreciates
when
people
are
talking
about
using
the
web,
because
there's
that
gives
a
momentum
to
invest
more
time
in
that.
B
We
are
right
currently
changing
the
way
we
advertise
advertise
our
services
over
and
DNS,
and
that's
just
a
quick
shout
out
from
Mayans
that
if
we
search
it
out
so
that
the
same
spec
is
implemented
by
go
ipfs
and
JSA
PFS,
we
basically
have
JSOC
EFS
running
in
ipfs
companion,
announcing
itself
to
nodes
on
the
local
network
and
that
node
can
be
ipfs
in
other
companion
running
on
someone
else
web
browser
or
it
can
be
going
PFS
running
on
the
local
node
and
basically
it
won't
matter
which
implementation.
It
is
right,
that's
the
angle.
B
D
Yeah
I
guess
they
want
the
one
cool
thing
that
I
didn't
mention
was
yeah
like
Lyta,
said
the
possibility
for
I,
like
ipfs
nodes
in
a
web
extension
to
be
able
to
interrupt
with
local
or
discover
and
talk
to
local
NGO,
ipfs
nodes
or
JSI
VFS
demons
on
the
local
network.
It's
just
that
would
be
amazing.
E
So
we
are
in
the
process
of
getting
the
next
version
of
ipfs
cluster
out.
It's
gonna
be
0
5
0
and
the
main
feature
is
that
you
can
add
content
directly
using
cluster.
So
normally
you
do
ipfs
ad,
and
you
can
now
do
I
PFS
cluster
CTL
ad
and
fremont
works
exactly
like
ipfs
from
the
user
point
of
view.
The
difference
is
that
it
will
allocate
that
content
to
several
beers
at
once
and
it
will
send
it
once
it
is
chunked.
E
Until
now
the
workflow
will
be,
you
add
two
ipfs
first
and
then
you've
been
in
cluster
and
then
it
needs
to
copy
what
you
paint
to
to
the
places
where
the
pin
was
allocated,
and
now
you
can
just
add
directly
and
the
moment
you've
been
all
the
content
is.
Is
there
in
the
destination
view,
there's
a
video
that
I
recorder
and
yeah?
It's
just
that.
Just
to
let
you
know,
I
will
hopefully
come
back
when
we
do
the
final
release
and
tomorrow.