►
Description
Opera launch: https://blog.ipfs.io/2020-03-30-ipfs-in-opera-for-android/
Meeting Notes: https://github.com/ipfs/team-mgmt/pull/1132
About IPFS GUI and Browsers Weekly: https://github.com/ipfs/team-mgmt/issues/790
IPFS Mirror: https://ipfs.io/ipfs/bafybeif32c2xsyysbn5trysuhvtydexdijxilcdlmmjbhju6fn5uycliku/
For more information on IPFS
- visit the project website: https://ipfs.io
- or follow IPFS on Twitter: https://twitter.com/IPFS
Sign up to get IPFS news, including releases, ecosystem updates, and community announcements in your inbox, each Tuesday: http://eepurl.com/gL2Pi5
A
A
This
is
me
sharing
my
screen.
Please
not
if
you
can
see
my
screen
perfect,
alright,
so
the
first
is
sort
of
like
a
shout
out
and
also
highlight
of
how
ipfs
adoption
is
progressing.
Its
room
project
is
using
ipfs
and
ens
for
hosting
their
website,
so
ET
room,
dot,
h,
ET
h
is
e
NS
name,
which
you
can
load
in
any
web
browser
that
has
IP,
has
companion
installed
or
has
a
support
for
DNS
resolving.
If
not,
you
need
to
add
that
link
suffix
and
just
for
the
purpose
of
demonstration.
A
There
is
this
IP
NS
colon,
slash,
slash,
link
which
later
we'll
talk
about
opera
and
why
this
link
is
important.
But
for
now,
if
you
open
this,
you
will
notice
that,
on
my
machine,
I
have
active
as
companion
installed,
this
better
version
which
supports
subdomains.
So
it
automatically
what
happened
here
is
IP
first
companion,
redirected,
ET,
room
that
eth
to
local
gateway
and
that
gateway
resolve
DNS
link
using
eth,
dns
resolver
and,
in
the
end
everything
ended
up
on
the
sub
domain,
which
gave
not
bad
user
experience
on
the
address
bar.
A
B
Feel
like
we
need
the
the
animated,
the
cinematic
version
of
all
of
the
hijinks
that
take
place
in
order
to
actually
make
that
happen.
I
I
do
take
issue
with
that.
It's
like
a
good
experience.
Hugh's
experience
at
the
end,
give
me
there's
a
big
padlock
with
anteye
through
it
not
secure.
So
there's
some
work
to
be
done,
but
still
form
a
community
about
so
standpoint.
Huge
and
shouting
it
was
massive
yeah.
C
A
So
close,
so
the
moment
day
will
hard-code.
The
problem
here
is
Firefox
does
not
trust
the
local
host
name,
because
it
may
be
hijacked
by
DNS
resolver.
What
chromium
is
doing
the
hard-coded
that
to
local
host
IP,
so
they
marked
out
a
secure
context
in
case
of
Firefox.
It's
not
the
case
yet,
but
it
will
be
so
if
anyone
wants
to
follow
this
issue.
There's
like
last,
like
two
weeks
ago,
there
was
a
patch
which
gave
us
both
like
secure
context,
hard-coded
IP,
but
it's
not
there.
A
Yep,
that's
the
announcement.
Another
quick
announcement,
slash
update
is
subdomain
in
IPS
companion.
It
was
scheduled
to
land
last
week,
but
they
did
not
cuz
I
continue
to
find
new,
bugs
and
edge
cases,
and
this
one
is
interesting,
so
I
thought
I
would
share.
So
it's
about
sub
resource
quirks.
So
it's
easy.
When
we,
when
you
open
an
entire
website,
it
gets
redirected
to
a
sub
domain
and
that's
it
like
everything.
It's
loaded
from
the
same
origin.
A
But
what
happens
if
you
got
a
regular
HTTP
web
page
and
that
web
page
has
some
ipfs
resources
and
a
sub
like
a
sub
resource,
for
example
you're
on
a
page,
and
this
image
could
be
loaded
from
ipfs
dot,
io
or
some
other
public
gateway.
So
what
happens
in
that
mixed
environment
when
you
have
a
regular
website,
but
someone
embedded
image
from
ipfs
gateway
and
we
want
IDF
s
companion
to
leverage
the
fact
that
we
have
a
local,
node
and
load
every
IP
FS
resource
from
local
node
instead
of
a
public
gateway.
A
This
is
what
companion
is
doing
right
now
if
it
finds
ipfs
resource.
On
a
page,
let's
say
it's
a
public
forum
and
someone
embedded
an
image
from
a
public
gateway.
It
will
redirect
that
specific
image,
not
the
entire
website,
which
is
not
on
ipfs,
but
that
specific
image
will
be
loaded
from
ipfs.
The
problem
is
like
most
of
websites
are
secular
context.
They
are
loaded
from
HTTP
and
if
we
try
to
load
sub
resource
from
insecure
context,
that
request
will
get
blocked,
and
here
where
it
gets
interesting.
A
So
I
made
a
test
page
to
illustrate
this.
This
is
I,
have
a
previous
beta
version
of
companion,
which
doesn't
ship
with
the
fix
yet,
but
it
demonstrates
the
problem
so
here
this
is
a
hug
hug
and
B
it's
HTTP
page,
and
that
page
has
an
embedded
image
from
our
public
gateway.
So
you
can
see
here
the
image
did
not
load
and
why
did
it
did
not
load,
because
the
image
location
is
a
public
gateway?
A
However,
ipfs
companion
will
redirect
that
to
a
sub
domain
and
that
sub
domain,
as
we
discussed
before,
like
moments
before
it's
not
secure
context.
Yet
so
it's
like
our
Rube
Goldberg
machine,
which
ends
up
with
image,
not
working.
However,
I
found
a
way
we
basically
would
use
path
of
gateway
for
sub
resources
for
now
in
Firefox,
I
think,
there's
a
separate
bug
in
chromium
where
they
may
be
a
little
bit
over
the
top
with
like
they
are
on
the
side,
safe
side
on
the
security
and
they
kind
of
hard
called
HTTP
protocol.
A
When
you
use
proxy
setup
so
I
think
there
is
either
a
bug
or
a
feature:
I
need
to
research
it
a
little
bit
more,
however,
the
same
fix
which
fixed
this
should
fix
it
in
chromium.
But
this
is
like
just
to
illustrate
its
if
we
want
to
leverage
IP
first
companion,
to
till
it's
like
limits
of
what's
possible.
We
we
need
to
address
this
before
we
ship
subdomains,
because
I
had
a
lot
of
hack
pots
where
I
had
images
from
ipfs
and
those
stopped
working
and
I
to
fix
that
first.
A
B
I
will
start
but
I
thinking
you
for
all
the
hard
work
you
did
in
setting
up
this
page,
for
example,
that
you
just
showed
us
that
led
to
the
launch
of
upper
browser
yesterday
for
Android
day
before
yesterday,
technically,
but
it
really
got
started
out
into
the
Play
stores
across
all
the
regions
by
Tuesday
morning.
Pacific
time
had
shown
up
for
me
in
East
Coast
us
in
the
place
water
but
yeah
opera
for
Android,
a
number
57
them
I
launched
with
a
cuvettes
protocol
support.
B
It
can
load
ever
since
back,
but
still
has
a
couple
of
bugs.
We
also
discovered
a
late-breaking
bug
where
the
Settings
panel
is
not
available
and
they're
gonna
ship,
a
an
update
that
looks
like
in
the
next
week
or
two
that
will
fix
that.
So
you
can
actually
then
choose
your
gateway
as
we
advertised
in
a
blog
post.
But
overall
the
reception
has
been
really
really
good,
interesting
character.
B
They
have
an
estimated
around
quarter
of
a
million
or
a
quarter
billion
users,
they're
saying
like
think
they're,
public
numbers
or
like
225
million
worldwide,
or
something
that
across
all
markets
products.
They
have
some
interesting
products.
Like
TV,
which
is
a
chunk
of
their
revenue,
or
they
have
applications
and
services,
and
then
they
have
a
numbers
on
smartphones,
desktop
dumb
phones
and
different
variations
of
some
of
these
devices.
B
Opera
Mini
has
deep
penetration
in
markets
Africa,
so
it's
really
interesting
and
then
the
company
itself
also
sold
to
a
investment
conglomerate
based
in
China
a
couple
years
ago,
so
just
a
unique
character.
The
source
is
not
open
to
the
browser
as
well,
but
they're,
very
supportive
of
web
three
use
cases
and
stocks.
The
developer
needs
which
really
I
profess
support,
was
a
easy
decision.
I
think
therein
to
make
it
just
smoothen
the
developer
pathway
there.
B
Let's
make
it
so
that
the
web
3
apps,
could
load
easily
in
browser
very
interesting
as
well
that
they're
choosing
mobile
first,
which
speaks
to
a
developer
demand
from
that
segment
for
them,
and
that's
something
that
in
other
efforts
key
all
around
nascent
support
for
mobile
use
cases
were
close
business
cases
are.
You
can
see
more
and
more
of
this
over
the
next
year,
but
overall
positive
replies
in
social
launched
a
bunch
of
different
press
coverage.
B
They
called
it
basically
like
their
their
kaboom
browser
and
that
really
got
a
lot
of
play
in
the
world
in
some
larger
tech,
press
and
business
press
as
well
so
and
I
profess
has
mentioned
in
almost
all
of
those
in
their
official
press
release.
I
profess
was
the
first
thing
mentioned
in
their
list
of
features,
so
that
was
a
and
try
to
get
some
fixes
into
that
press
release
around
things
like
the
consistent
use
of
address
or
addressing
there's
some
work.
B
A
It's
awesome
just
just
a
quick
highlight
that
they
launched.
However,
their
rollout
on
the
Android
may
be
delayed
depending
on
the
region
you
are
in.
So
if
it's
you,
you
can
install
it
from
the
Play
Store
on
Android.
However,
the
version
dude
that
you
installed
may
not
have
it
like.
Yes,
that
was
the
case
in
the
beginning
of
the
week.
However,
maybe.
B
A
B
B
A
Awesome
awesome,
it's
so
very
exciting,
especially
given
that
it's
the
first
time
the
like
mainstream
web
browser
supports
content.
Address
content
addressed
resources
as
the
first-class
citizen,
because,
just
like
I
shared
before,
like
you,
you
have
this
ipfs
:
/
CID
link
and
you
don't
specify
location
and
your
browser
will
just
get
it
and
that's
pretty
powerful.
The
fact
that
its
delegated
ipfs
client,
all
that
stuff
is
like
important
but
I,
think
that's
the
way,
we'll
look
back.
That
would
be
like
the
first
step
into
like
moving
away
from
location,
addressing
how
super
awesome.
A
F
F
I'm
doing
a
bunch
of
work
related
to
like
removing
no
Global's
stuff
like
that.
So
basically,
our
code
base
has
a
lot
of
expects,
a
lot
of
stuff
that
it's
only
available
in
node,
so
I'm
doing
some
changes,
so
the
browser
integration
is
easier
with
bundlers
and
all
that
stuff.
So
I
got
stuck
a
little
bit
on
the
data
store
that
we
had
that
Jesus
used
level
DB.
F
So
I
did
some
work,
trying
to
move
the
leveldb
ripples
to
know
the
global
free
environments,
but
I
was
getting
stuck
a
lot
because
there's
so
many
repos,
some
interaction
and
abstractions.
So
basically,
I
have
been
wanting
to
do
a
new
database
for
a
long
time.
So
yeah
I
did
it.
You
have
the
the
repo
there
it's
it
uses
directly
index
TV
number,
so
there's
no
rubber
around
index
to
be
like
leveldb
does,
and
actually
we
didn't
need
it
like
for
no,
it
might
be.
F
F
It's
a
little
bit
faster
I'm,
still
seeing
like
level
DB
or
level
J's,
which
is
a
browser
version
of
level.
Dvd
I
think
it's
still
in
lines
a
little
bit
better.
So
if
you
run
it
like
three
or
four
times
level,
TV
starts
to
be
a
little
bit
faster
than
the
new
data
store,
but
I
should
be
able
to
work
around
it
and
optimize
the
code
base,
but
mostly
it's
easier.
It's
smaller!
It's
like
maintaining,
maintaining
it.
It's
it's
gonna
be
much
simpler
because
it's
just
a
few
lines
of
code.
F
A
D
F
F
G
F
One
remark
yeah
I
saw
this
is
mostly
for
like
core
developers
or
contributors,
that's
already
dealt
with
repo
or
anything
related
to
data
source.
I
saw
in
a
lot
of
places
like
the
data
stores
being
used
without
calling
the
open
method,
because
some
of
the
I
don't
even
know
how
why
this
was
like
all
over
little
ways.
But
you
need
to
call
open
on
the
database
at
least
on
this
one
on.
F
If
you
don't
call
open,
you'll,
get
an
error
explaining
I
added
specific
for
this
use
case,
because
in
a
lot
of
places,
the
open
method
was
not
being
called,
and
at
least
for
this
one.
You
need
to
call
it
or
it
doesn't
work
to
get
an
error
telling
you
latest
or
needs
to
be
open.
So
if
you
see
that
you
already
know
what
you
need
to
do,
you
need
to
call
open
in
a
data
store.
A
A
The
grant
to
add
native
protocol
handler
for
browser
extensions,
so
this
will
be
a
generic
API
I
believe
we
discuss
it
a
few
times
before
so
here.
I
will
just
give
a
quick
update
that
we
have
a
discussion
with
earlier
about
some
details
to
flesh
out
maintain
milestones
mostly,
for
example,
one
of
one
of
those
is
the
API
for
registering
redirect
based
protocol
handler,
which
is
not
supported
by
chromium.
Yet
so
there
is
this
API
which
lets
Firefox
users,
Firefox
extension
users
to
register
redirect
based
protocol
handler.
A
A
So,
unless
we
do
this
as
a
part
of
this
grant,
it
won't
happen
and
I
feel
it's
probably
a
good
milestone
to
first
deliver
feature
parity
with
Firefox,
so
gets
to
the
level
of
redirects
being
supported
in
both
places
and
being
able
for
extension,
being
able
to
register
protocol
handler
and
then
build
this
native
handler,
which
is
capable
of
returning
bytes
on
top
of
this.
So
this
is
like
a
quick
update,
we'll
probably
continued
discussion
in
this
long
long
long
long,
long,
follow
request.
However,
I
just
wanted
to
highlight
that
the
discussion
is
happening.
I
So
pops
up
options
in
the
browser
are
pretty
limited
at
the
moment
and
they
usually
evolve
some
some
middle
server.
That
has
a
bunch
of
like
weird
logic
in
there
to
make
that
happen.
So
this
past
weekend,
I
stopped
out
the
initial
work
for
a
pub
sub-base,
pure
discovery
module,
and
it's
mostly
done
I'm
just
working
on
some
integration
with
a
relay
server.
But
ultimately
what
this
does
is
it
just
requires
that
a
JSL,
p2p,
node
or
JSI
PFS
node,
run
pub/sub,
and
then
it
subscribes
to
a
global
namespace.
I
So
it's
great
for
the
future.
One
of
the
things
that
I'm
gonna
be
working
on
before
this
gets
released
is
actually
being
able
to
namespace
your
discovery
so
that
you
would
be
able
to
be
like
application
specific
namespacing
for
for
peer
discovery,
instead
of
just
binding
to
the
to
the
global
pre
discovery.
Space.
A
This
is
super
cool,
especially
the
name
spacing
path,
because
the
way
mdns
works.
It's
basically
shouts
to
everyone
on
the
local
network,
hey
I'm,
here
those
are
my
addresses
and
in
case
of
the
web,
that's
sort
of
like
a
local
network
is
the
origin,
so
I
believe
in
the
joyous
ipfs
that
like.
If
we
talk
about
safe
default,
we
probably
would
not
put
everyone
on
the
same,
like
namespace
I'm,
not
sure
how
that
would
scale
if
it
gets
a
lot
of
usage.
I
That
we
hosts
for
like
will
end
up
deploying
a
relay
server
that
listens
to
this,
that
it
will
probably
have
like
maybe
its
own
namespace,
and
so
this
kind
of
like
handles
like
the
Stardust
or
WebSockets,
our
server
of
like
okay.
This
is
everybody
who's
intended
to
be
connecting
to
this
point
and
then
potentially,
you
could
also
subscribe
to
to
multiple
things,
but
that
would
be
something
I
need
to
add
to
is
being
able
to
subscribe
to
multiple
namespaces
and
respond
to
those
queries
appropriately.
That
would
be
that
would
be
helpful,
I
think.
B
A
Ups
up
its
the
world
yeah
and
my
question
was
what
makes
more
sense
to
implement
support
for
listening
on
multiple
namespaces
or
just
taking
a
namespace
like
a
template
in
a
constructor.
And
if
someone
wants
to
listen
on
multiple
namespaces,
they
just
initialize
the
same
module
multiple
times.
Is
there
like
a
performance
consideration
or
is
it
just
like
I
mean.
I
D
A
I
Not
currently,
and
partly
because
like
live
DP
configuration
is
daunting
at
the
moment,
but
we
have
an
initiative
to
clean
that
up
and
make
that
like
a
much
more
functional
approach
to
configuring,
Lippa
DB
and
that
would
provide
a
lot
more
flexibility
to
be
able
to
say:
okay,
I'm,
just
gonna
instantiate,
this
pub/sub
module
like
five
times
and
then
pass
that
into
the
discovery
services
and
then
call
it
a
day.
So
that
would
be
an
option
in
the
future.
Yeah.
A
F
I
Yeah,
that's
what
I'm
gonna
get
to
at
the
next
spot
like
this,
the
module
is
pretty
close
to
being
ready.
The
major
thing
I
have
right
now
is
I'm.
Configuring
I've
got
a
repo
where
I'm
just
building
out
a
pre-configured
relay
server,
so
that
people
will
be
able
to
like
deploy
it
from
a
docker
container
and
in
there
I
have
the
test
suite
that
uses
this,
and
then
it's
running
like
easier
to
actually
test
across
the
browsers
to
make
sure
that,
for
your
discovery
is
working
so
that
relays
are
rich
doing
a
lot
of
integration.
I
I
Main
issue
is
that
it
does
not
emit
or
he
does
not
a
support,
remove
listener,
and
so
that
creates
that
created
some
problems,
because
low
p2p
actually
is
using
remove
listener
instead
of
off
in
a
min
rate,
only
supports
off,
and
so
that
was
something
that
I
just
had
to
add
and
remove
listener
method
to
resolve
that
so
there's
some
considerations
there
of
it
basically
only
applies
to
like
its
emitter.
Ii
only
supports
node,
ten
events,
syntax.
B
Because
triplicates
so,
but
one
thing
I
found
here,
is
that
I
I
understand
with
general
leave
the
impact,
because
I
asked
you
all
about
it
every
week
here
in
this
meeting.
The
other
ones
as
well,
but
with
the
impact
of
some
of
this
work
is.
But
if
you
could
do
me
a
favor
and
for
your
okrs
there
it's
a
list
of
issues,
but
it
isn't
clear
what
the
impact
of
the
some
of
those
issues
are
and
why
this
work
needs
to
be
done.
So
for
each
one
of
your
cars
for
Rosco.
J
A
A
C
A
B
A
A
C
J
Not
it's
fine!
Oh
hello,
yeah
hi!
Everyone
is
this
elder.
I,
wasn't
sure
if
this
was
the
birthplace
to
join
I,
just
wanted
to
clerk
in
and
hear
the
meeting
in.
You
know
with
ipfs
Pharrell
now
and
had
a
great
talk:
I,
don't
the
other
day
so
yeah.
That's
me,
Helder,
nice
to
meet
you
all
nothing
much
to
end
at
this
moment,
but
it
was
a
great
pleasure
to
be
part
of
the
meeting.
So
thanks
for
all
you've
been
doing
for
ipfs,
it's
a
great
project
that
I
have
a
good
feelings
towards.
A
B
A
A
About
to
share
that
that
you
are
very
a
very
modest
person,
because
you
are
the
person
behind
ipfs
deploy,
which
is
this
very
cool
command-line
tool
you
created
initially,
it's
a
super
useful
command
line
tool
for
quickly
publishing
website
to
ipfs,
with
things
pinning
with
pinning
services.
So
if
someone
is
watching
this
and
it's
using
this
and
now
you
know
who
created
it.
J
B
It's
already
on
github.
The
idea
is
it's
not
we're,
not
even
gonna
ship
it
as
a
browser,
but
it's
a
place
where
people
can
experiment
and
we
can
show
what
the
face.
That's
that
browser
would
look
like.
We
talked
a
lot
about
this
in
the
past
around
having
an
experimental
vehicle.
It's
where
we
can
say,
for
example,
implement
the
browser,
design
guidelines
or
even
let
me
be
a
person's
back
and
play
with
an
accessibility
accessible
way.
So
where
we
can
take
some
of
the
ideas
we
have
be
able
to
demo
them
in
a
way.
B
That's
fast
and
easy
top
of
the
browser
did
this
a
couple
of
weeks
months
ago,
where
they
released
a
similar
idea
where
technology
wired
up
to
in
browser
nodes
and
with
electron
it
makes
this
really
really
easy.
So
this
is
a
place
where,
if
you're
interested
well,
then
I'm
source
take
a
look
back
on
it.
It's
a
place
where
we
can
play
with
different
developer
patterns,
user
patterns,
design
patterns,
things
like
location
bar
security
privacy.
B
Hopefully,
we
use
it
as
a
tool
for
experimenting
and
prototyping
some
of
the
ideas
of
that
and
use
it
as
a
tool
for
communicating
to
browsers
so
that
you
can
actually
have
a
visceral
and
tangible
experience
of
what
these
technologies
might
look
like
how
they
might
behave.
What
so
the
visual
communication
challenges
are,
but
it
takes
Alain
for
spinning
yourself.
D
C
C
A
A
A
If
you
don't
have
any
ad
hoc
thoughts,
we
can
always
make
it
bit
shorter.
This
week
it
was
super
action-packed
I
feel
like
it
was
a
lot.
However,
we
finished
15
minutes
earlier,
so
either
I
don't
know
it's
the
efficiency
of
doing
something
over
and
over
again.
I
think
all
right
folks,
I
think
I'll
give
you
back
your
15
minutes.
Thank
you
for
everyone,
who's
watching.
Thank
you
for
everyone
who
shared
projects
or
thoughts
and
ideas
or
updates
and
see
you
next
week,
I.