►
From YouTube: IPFS All Hands 🙌🏽📞 Nov 12, 2018
Description
IPFS Newsletter: https://tinyletter.com/ipfsnewsletter
A
B
B
Okay,
sorry
about
that,
so
let
us
continue.
So,
as
you
heard
before,
we
are
changing
our
format
and
it's
still
30
minutes,
but
we
will
be
going
with
announcements.
Then
we
will
have
our
main
talk
and
after
our
main
talk
we
will
leave
some
time
for
questions
and
answers.
So
let
us
begin
with
announcements.
If
you
have
an
announcement,
please
do
raise
your
hand.
Do
we
have
announcements.
B
D
So
we
call
lib
diva,
it's
admiral
on
Mozilla
github
account
it's
developed
all
the
main
tree
right
now,
and
it's
set
of
very
small,
primitive
API
said
you
know
implementing
something
like
ipfs
node
in
the
browser
itself.
The
first
one
is
a
protocol
API.
It
just
allows
you
to
define
an
implementation
for
your
protocol.
D
So
in
this
example
that
you
see
on
the
screen,
it
allows
you
to
define,
give
a
protocol
and
it
allows
you
to
basically
get
a
request
and
stream
the
response
which
can
be
whatever
you
want
it
to
be,
and
it
has
a
correct
origin,
meaning
that
origin
will
be
hostname
and
the
schema
combination.
So
if
you
have
many
different
things
on
the
ipfs
network,
there
will
be
always
related
through
the
standard
web
content,
security
policies
and
there's
an
example
how
it
actually
works.
D
D
It's
a
little
bit
more
tricky
because
we
need
to
go
through
the
users
homes
and
make
sure
that
we
don't
give
away
access
to
the
whole
filesystem
and
there's
also
quota
management
stuff.
So
a
lot
of
that
stuff
overlays
with
bunch
of
other
search
mechanisms,
but
it
seems
to
perform
way
better
than
say
indexeddb,
so
we'll
see
how
it
will
go.
D
D
D
Api
is
that
process
itself
contains
many
bunch
of
other
steps,
and
one
of
them
is
defining
a
thread
model
which
is
pretty
much
what
I'm
busy
with
trying
to
get
out
and
once
we
have
that
in
place,
I
think
it
should
be
just
matter
of
actually
learning
zone
just
inches
of
firefox.
There
are
some
of
the
additional
work
that
were
planning
on
doing
the
API
is
Wednesday.
D
There
are
few
things
that
are
still
open
and
kind
of
are
not
as
high
on
my
list
right
now,
but
I
would
welcome
anyone
to
come
in
and
help
I'm
happy
to
mentor
with
that
work
as
well.
One
of
them
is
Ben
for
the
protocol
API.
We
don't
actually
have
support
for
headers.
That's
because
headers
are
actually
tied
to
the
HTTP
protocol
so
for
the
custom
protocols,
there's
no
such
notion
as
header.
D
So
if
you
want
to
come
up
with
something
that
would
require
some
changes
into
how
Firefox
works,
they're
not
too
difficult
changes,
but
they
are
changes.
Nevertheless,
there
are
a
few
other
low-hanging
fruits
that
somebody
can
tackle.
If
you
want
to
come
and
help,
the
best
thing
would
be
to
go
at
Mozilla,
leave
the
web
on
github
and
just
open
an
issue
or
there's
a
pointers
to
everything
else.
You
can
get
in
touch
with
us.
So
that's
a
brother
overview
and
I'm
happy
to
answer
any
questions
you
might
have.
D
C
D
I
should
have
mentioned
so
Leo
and
Alan
show
like
things
they
did
amazing
job
at
integrating
that
stuff
in
choose
a
branch
of
the
fire
IP
of
his
companion.
So
you
can
actually
see
some
of
the
stuff
were
already
working
in
the
ipfs
line.
There's
also,
a
dad
version
that
is
being
worked
on
said
actually
runs
freighter,
which
is
Twitter
cloning
that
network
over
that
over
the
add-ons
that
use
lead,
Evette
and
I
think
there
are
some
attempts
to
get
it
into
the
secure
scuttlebutt
as
well,
but
I
don't
know
how
far
along
they
are.
D
E
D
That
would
be
great
too.
In
fact,
I
am
I'm,
happy
to
open
the
shred
model,
worked
more
open
to
it.
If
you
can
come
and
contribute
that
to
I
think
that
would
be
in
general
useful
in
terms
of
understanding.
Personally,
that's
not
something
I
have
ever
done
before.
So
it's
been
you
and
been
learning
how
to
like
see
all
the
ways
things
can
be
abused.
That
I
could
not
imagine
so
I
could
definitely
up
there
too,
and
that
doesn't
necessarily
require
any
code.
D
Writing
Martinez
thinking,
critical
thinking
in
terms
of
how
would
we're
exposing
can
be
abused
by
bad
players
and
those
can
be
bad
web
extensions
or
it
could
be
malicious
content
hosted
in
ipfs
Frances
documentation
wise.
That
would
be
another
great
help
or
making
small
examples.
There
is
also
a
bunch
of
work
needs
to
be
done
in
the
test.
Infrastructure.
D
Wise,
so
part
of
the
problem
is
Mozilla
has
test
infrastructure
that
is
tightly
coupled
with
the
Firefox,
build
steps
and
all
of
this
stuff,
and
we
didn't
want
to
bring
that
for
this
project,
so
it
was
easier
to
develop,
but
that
also
meant
that
we
don't
have
any
test
for
infrastructure.
So
we
start
kind
of
building
our
own
and
there's
something
in
place,
which
is
not
perfect,
but
that
could
use
some
help
or
just
adding
more
tests
would
be
also
helpful.
D
D
F
I'm
curious
about
the
threat
model,
and
it
seems
to
me
like
that,
the
extension,
if
you
give
it
full
access
to
tcp/udp,
that's
pretty
dangerous,
like
in
terms
of
like
people,
malicious
people,
so
I,
don't
I,
think
like
if
you,
if
the
extensions
were
like
there's
a
gatekeeper
in
Mozilla,
really
goes
over
the
security
on
this.
The
extensions
that
are
submitted
is
this
going
to
require
Mozilla
to
change
the
way
it
accepts
extensions
for
these
particular
EPO.
So.
D
D
It
is
that's
part
of
it
so
and
often
threat
model
as
I
understand
it
now
is
basically
deciding
listing
Ward
sinks
can
what
harm
it
can
do
and
then
sort
of
trying
to
place
mitigations
in
place
to
prevent
the
larger
issues
and
for
some
issues
we
might
say:
well,
that's
kind
of
a
user
choice
in
terms
of
installing
a
ban
on
and
I
think
big
chunk
of
that
will
fall
into
that.
But
tcp
api
is
not
actually
as
interesting.
D
I
think
the
protocol
handler
is
one
one
of
the
most
interesting
ones,
because
there's
a
lot
of
ways:
malicious
add-on
or
malicious
content.
Ipfs,
for
instance,
could
get
you
try
to
impersonate
popular
websites.
For
instance,
somebody
could
write
a
age
LS
protocol
handlers
that
will
try
to
trick
you
into
sick.
Is
that
your
own
amazon.com
and
steal
your
credit
card
number
or
some
variation
of
that?
So
one
of
the
mitigation
there
would
be
to
make
the
UI
of
the
address
bar
distinct
enough.
H
G
The
work
we
do
have
it
again
like
thinking
few
years
ago
and
I
came
back
three
years
ago
and
last
year
and
now
see
that
this
is
possible
and
if
we
already
have
them
all
working,
that's
amazing,
so
great
work.
First
first
question
is
I
think
we
we
never
talked
about
this,
but
one
of
the
topics
it
comes
often
when
you
talk
about
the
peer-to-peer
and
transports
is
not
only
ECP
and
UDP,
but
actually
like
the
fact
that
we
cannot
like
mess.
We
the
keys
that
get
used
by
web
RTC
or
engine
soon
future
quick.
G
We
actually
have
a
quick
government
that
you
didn't
go
for.
We
could
appear
today,
but
we
already
know
that,
like
for
browser
to
girl
nodes
or
to
just
know,
you
know
Jeff's,
we
will
always
have
to
face
the
challenge
of
doing
WebSockets
and
then
some
cryptic
channel
and
then
some
sweet
Mac
sure.
Even
even
if
you
that's
over
a
secure
connection,
because
again,
we
cannot
give
to
the
transport
the
keys
to
use
to
authenticate
the
connection
and
therefore
it
indicates
the
nodes.
G
D
Maybe
I
I
don't
have
a
answer,
because
that's
not
something
I
have
sought
before
I'm.
Not
exactly
sure,
if
said,
would
be
helpful
in
turn
because
likely
be
web
is
con
is
scoped
to
the
web
extension
stuff.
It's
not
gonna
be
available
to
the
web
pages
themself,
even
though
you
could
expose
some
of
those
things
and
I
don't
know
if
there
would
be
a
benefit
for
you
to
like
do
that
from
the
web
extension
space
rather
than
just
using
direct
protocol,
because,
like.
F
D
Know
so,
probably
quick,
would
it
make
sense
to
like
expose
some
API
for
that?
To
be
honest,
I
haven't
really
looked
into
creeks
that
much
so
it's
hard
for
me
to
how
much
but
I'm
totally
open
to
like
talking
about
the
specific
API
as
well
another
like
Stanford
that
might
be
related
to
what
you're
asking
is
me
and
there's
also
Dietrich
Iola
at
Mozilla.
We
are
trying
to
form
a
w3c
interest
group
where
we
can
actually
talk
about
what
are
the
APS.
We
can
standardize
and
what
they
should
look
like.
D
I,
don't
think
TCP
or
UDP
would
be
that,
but
some
sort
of
message
channeling,
where
you
could
say
maybe
for
my
swarm
based
on
some
key
and
we
don't
necessarily
talk
about
the.
What
the
underlying
thing
is
choose
the
developers
I
mean,
but
we
can,
as
an
implementer
decide
what
those
things
could
be.
That
might
be
one
option
we
could
go
and
in
general,
I.
Think
bunch
of
people
from
the
IPS
are
part
of
the
group
where
these
discussions
are
happening.
So
we
hope
with
forming
an
interest
group.
G
D
No,
so
what
happened
is
fly
web
I
think
because
it
was
touching
UI
parts
and
back-end
parts.
At
the
same
time,
once
it
was
backed
out,
they
couldn't
make
out
both
of
them
because
one
of
them
would
break
so
they
backed
out
the
on
part
and
forgot
the
back
end
part.
So
it's
a
historical
accident
that
the
code
base
that
we
use
end
up
in
a
tree
yeah.
So
that's
the
state
of
it
fly
web
I,
don't
think
it's
coming
back.
Okay,.
D
So
we
were
in
talks
with
the
people
who
developed
it
and
got
some
advice,
and
actually
they
were
very
helpful
in
trying
to
get
things
working,
but
yeah
I,
don't
think
the
fly
web
as
a
project
is
coming
back
or
I,
don't
know
if
it
we,
it
is.
I
have
not
heard
anything
along
those
lines
once
some
of
the
things
that
we
will
be
trying
to
do
at
Mozilla
is
do
some
explorations
that
might
some
resemble
fly
web
or
it
may
not.
D
On
top
of
those
api's
and
other
things
that
we're
also
looking
into
so
we
have
a
gecko
in
the
Android
now
so
we're
and
it's
pretty
much
set
of
components
that
you
can
build
browsers,
which
and
we're
thinking
about
integrating
Lady
webs
are
too
so.
You
can
have
Android
apps
that
do
kind
of
your
stuff.
C
C
So
could
you
speak
to
where
this
puts
us
along
the
line
of
having
it
be
the
case
that
an
everyday
user
of
a
web
browser
has
the
web
support
automatically
just
because
they're
using
a
web
browser
like
they
don't
have
to
do
anything
special
in
order
to
be
able
to
have
these
protocols
working
when
they're
browsing
on
the
web
and
also
sort
of
with
that
there's?
How
has
the
reception
been
around
you
on
the
on
the
side
of
Mozilla
and
like
browser
manufacturers?
Oh
I,.
D
D
If
I
don't
go
through
the
cloud,
which
is
just
ridiculous
and
I
think
talking
about
those
problems,
these
people
really
resonate.
So
maybe
we
ought
to
do
something
about
it.
Where
browsers
can
just
do
allow
websites
to
do
those
things
and
I
think
it's
kind
of
more
of
a
singing.
What
if
we
did
something
like
what,
if
rather
I,
became
ipfs
note-
or
maybe
it
said
ipfs-
maybe
it's
that
or
some
other
purple
I,
don't
I'm
not
trying
to
take
Kemp's
but
kind
of
thinking.
D
How
we
can
make
the
web
be
how
we
can
make
sorry,
I'm
not
answering
really
well,
but
I
I'm
singing.
We
need
a
high
level.
Api
is
that
are
part
of
web
that
enables
applications
that
were
people
are
trying
to
build
today
with
ipfs
or
that
not
necessarily
is
this
low-level
api
s--
and
that's
part
of
one
of
the
things
that
we
want
to
pursue
through
the
conversations
between
protocol
implementers
and
browsers
through
the
w
stranger
at
scope.
D
D
Time
is
kind
of
right,
because
people
are
becoming
more
conscious
about
many
things
that
they
were
not
before,
and
I
think
this
is
the
right
time
to
kind
of
push
towards
a
decentralization
more
and
I
think,
there's
an
opportunity
for
mozilla
or
any
other
browser
to
do
stuff.
Sir
and
I
think
Mozilla
is
actually
uniquely
positioned,
being
aligned
with
a
user
needs,
more
than
corporate
needs
to
be
tackling
that
so,
hopefully
said,
would
help
making
more
progress
in
the
direction.
B
B
B
A
Announcements
just
for
all
the
people
who
are
watching
us
on
live
streaming
that
it
should
be
working
properly
again
and
all
of
our
links
should
go
to
the
right
places.
Yeah
surprising
amount
of
work
to
figure
out
what
was
going
wrong,
but
it's
all
good
now
and
so
and
the
next
time
we
ever
run
into
that
again.
It's
been
added
to
our
call
dialogues
and
our
how-to.
So
if
you
have
questions
go
to
the
thank
you
and
all
of
the
instructions
on
how
to
fix
it
will
be
there
and
Molly.
G
But
it's
about
like
meets
chlorine,
the
oh
yeah.
So
this
is
a
time
of
the
quarter
where
we
check
in
with
our
teams,
and
we
check
what
we
had
planned
for
the
quarter
and
we
kind
of
like
score
all
far
and
we
come
on
the
week
set
that
fast
and
we
try
to
do
a
projection
like
a
best
guess,
given
the
time
that
is
left
and
given
the
new
ally
responsibilities
that
we
might
have
received
since
the
quarter
started
and
also
the
things
that
we
identified,
that
are
blockers
that
were
unknowns
in
the
beginning.
G
G
You
go
from
like
the
ad
is
going
from
point
A
to
point
B,
a
lot
of
people
like
fine,
a
little
bit
off
route,
and
so
now
it's
the
time
to
really
like
check
in
and
see
if
it's
more
available
to
adjust
course
and
go
back
and
in
or
if
it's
still
valuable
to
go
into
this
new
direction,
because
you
found
something
more
important
if
there
is
much
something
more
important.
That's
totally
fine,
as
long
as
like
you
make
sure,
like
you
check
in
with
the
team
and
to
commit
to
that.
G
If
you
do
a
record,
if
you
have,
for
example,
a
completed
project,
call
it
completed
score
right
now
is
you're,
not
four,
because
that's
how
far
you
haven't
gotten
till
now
and
then
you
want
to
say,
I'm
not
going
to
work
on
this
anymore
until
doing
of
the
quarter,
because
I
want
to
focus
on
a
p0
rather
than
focusing
on
this
p3,
for
example,
just
just
put
the
same
score.
Let
people
know
okay,
I,
think
I'm
going
to
move
the
needle
is
not
going
to
move
here
and
then
not
use.
G
Your
weekly
things
from
your
video
working
looks
just
like
you're
checking,
if
that
makes
sense
to
everyone
get
feedback
or
or
if
there
are
other
stakeholders
in
the
community
that
you
know
it's
we're
waiting
on
that
feature.
I'll
serial
consider
like
tagging
them
on
a
comment
on
an
issue,
but
that's
it
any
any
questions
on
this
like
there
was
an
issue
that
kind
of
explains
this
issue:
seventy
set
by
five
seven,
five:
five
on
the
team
management,
repo.