►
From YouTube: Default Ports and Subdomain Gateways - IPFS GUI and In Web Browsers Weekly, 2019-07-17
Description
Meeting notes: https://github.com/ipfs/team-mgmt/pull/999/files
IPFS Mirror: https://ipfs.io/ipfs/bafybeignzfe3sk3dkjhv7mlin6dq3e2bbqdv7ircts6cuqm4kqbyjtvbzu/
B
Welcome
to
the
Kohl
top
of
the
agenda
is
seamless
experience
around
port
8080
or
5001
being
taken,
so
port
8080
being
the
gateway
pool
by
default
by
business
and
port
5000
among
being
the
API
ripe,
or
by
default.
You
know,
IP
advanced
Lydell.
You
added
this
vision
that
you
care
to
get
it
unpack
it
a
bit
further,
usually
added
yeah,
adding
more
to
a
gender
right.
Now,
no.
A
I'm
just
press
the
wrong
button
yep,
so
the
story
behind
this
is
that
we
will
I.
Give
us
note
start
be
that
coati,
FS
or
IP
first
desktop.
It
tries
to
open
gateway
port,
which
is
8080
and
API
port,
which
is
5001,
and
the
problem
is
that
88
is
very
popular
port
and
especially
if
the
person
is
a
developer,
there's
a
very
high
chance,
the
party's
already
taken
by
something
else,
some
HCP
server,
it's
unfortunate,
but
that's
like
a
default
port
for
gateway.
A
The
problem
is
that
sort
of
impacts,
the
onboarding
of
new
users
when
a
person
wants
to
try
a
TFS
right
now,
it's
high
probability
that
person
is
so
developer.
That
means
there
is
high
probability
of
port
8080
being
taken
and
that
person's
installs
ipfs
quite
give
us
already.
First
desktop
they
start
and
the
first
experience
is
all
it
crashed.
It
says,
forties
taken
and
usually
the
message
is
so
the
idea
is
we
should
optimize
this
golden
path,
at
least
for
users
of
ipfs,
desktop
I.
Think,
there's
a
fair
assumption.
A
B
Improving
the
new
user
experience
for
new
users
by
picking
a
more
sensible
default
port
and
then
this
problem
of
like
we
can
choose
random
ports.
But
then
we
have
lots
of
code
elsewhere
as
a
is
doing
best.
Effort
heuristics
to
be
like
I'm
gonna
check.
Port
5001
for
HTTP
api
and
changing
that
were
relying
more
heavily
on
a
random
port
would
make
that
process
harder.
A
Yes,
so
mainly,
we
don't
want
to
change
defaults
because
for
most
of
users
that
if
it
will
work
fine,
however,
if
the
port
is
already
taken,
we
should
optimize
that
path,
because
it
leads
to
frustration
like
we
should
avoid
situation
when
someone
tries
to
use
ipfs,
and
the
first
welcome
screen
is
the
for
cystic.
There
might
be
a
reasonable.
B
A
So
that's
the
question:
should
we
either
way?
Even
if
we
switch
to
two
unique
port,
there
will
be
still
a
situation
when
someone
has
already
something
chronic
there,
it
will
be
rare,
but
so
I
think
my
point
is
we
should
for
the
GUI
apps
such
as
IP
first
desktop,
it
should
be
seamless
for
user
and
there's
the
path
for
checking
if
the
thing
that
is
listening
on
the
port
is
actually
ipfs.
C
I
really
like
them,
the
midterm
solution
was
ultimately
changing
the
port
to
be
something
that
98
thousand
other
developer
tools,
don't
use,
but
I
think
one
of
the
frustrations
that
developers
have
is
that
even
if
we
get
there
first,
it's
a
port
race
and
then
there
other
tools
break.
So
our
part,
that's
great,
that
we
got
the
port,
but
then
there
other
workflows,
I
professed
as
a
pain
for
that.
C
So
the
question
that
I
like
to
ask
around
things
like
this
is
what,
if
we
added
1
million
to
5
million
new
developers
who
are
using
ipfs,
we
still
think
8080
is
the
right
thing
or
let's
say
we
only
out
of
the
hundred
thousand
I
profess
developer.
So
we
still
think
it'd
be
80
he's
right
and
from
that
understanding
a
word
growth
is
going
to
come
from.
If
we
continue
winning,
then
maybe
we
want
to
have
something
that
will
set
us
up
for
a
long
term
and
play
nicely
with
all
the
only
tools.
A
Sort
of
sort
of
related
to
that
is
a
concept
of
familiarity
right
now
we
know
that,
like
8080,
it's
a
gateway
wrench
in
jeaious
world,
it's
like
1990,
it's
bit
better
and
the
API
part
in
some
separate
namespace,
which
is
like
5,000,
something
so
at
least
there's
like
a
visual
indicator,
which
is
which
and
I
think
I
finally
can
show
you.
What
I
wanted
to
show
you
before
I
think
is
that
right
now,
I
have
go.
A
B
Really
down
with
picking
a
random
pool
in
the
link
on
where
we
find
that
the
false
port
is
unavailable,
I
think
my
thinking
was
coming
from
as
a
web
app
there's
no
good
way
for
me
to
discover
a
local
IPS
API
running
on
the
system
yep
and
as
it
stands,
our
recommended
path
for
that
is
to
use
a
companion
window.
Type
EFS
has
caveats,
but
the
the
path
is
like
have
a
randoms
web
app
have
direct
access
to
your
yeah.
B
It's
still
pretty
ropey,
it's
just
a
terrible
security
model,
it's
terrible
UX
model,
so
I'm
kind
of
okay
with
like
a
friend
of
ports,
I
think
if
you
install
desktop
today
and
the
options
are
like
crash
or
install
successfully,
but
you
are
now
running
things
on
different
ports.
I
think
we
probably
need
to
go
for
running
things
on
random
ports,
not
crash,
but
we
need
to
highlight
to
the
user
that
your
setup
is
not
the
same
as
everyone
else's,
so
certain
certain
things
might
not
work
as
expected.
B
You
will
have
to
go
and
configure
companion,
whereas
another
user
won't
have
to
configure
companion
brothers
drop
it
in
and
it
will
work
out
of
the
box.
I
guess
it's
all.
The
best
we
can
do
right
now
is
be
robust
and
signpost
people
to
the
ramifications
of
that
difference
that
make
them
both
like
right
now,
companion
and
Justins.
If
you're
not
only
on
a
table,
just
work
out
the
box,
that's
great
and
if
they
stop
put
things
that
run
importance,
there's
no
way
to
companion
to
know
that.
A
B
If
you're
writing
a
tool
that
has
system
API
is
like,
we
could
discover
things
on
mdns,
but,
alas,
companion
is
not
one
of
those
and
that's
I,
guess
the
thing
that
we
care
about
most
at
this
point
is
like
from
a
GUI
I
could
ask
GUI
perspective
is
like
desktop
works
out
of
the
box.
Companion
works
out
the
box,
and
you
have
a
good
experience
when
you
read
the
documentation,
the
things
that
it
tells
you
to
run
work
for
you.
A
For
like
for
companion,
companies
sort
of
like
an
easy,
easy
case,
because
we
control
both
ipfs
companion
and
IP
has
desktop,
so
we
can
figure
it
out.
A
protocol
where,
like
desktop,
opens
the
page
and
companion
takes
that
as
a
hint.
However,
we
should
think
about,
like
third-party
developers.
Yeah
don't
have
that
privilege
and
that's
like
sort
of
I
think
it's
rather
than
specific
to
API
port,
like
specific
reports,
it's
more
about
local
discovery
and
the
problem
of
having
multiple
nodes
on
single
machine.
It's
like
at
some
abstraction
level.
It
collapses
into
a
single
problem.
B
I,
yes,
although
I'd
be
wary
of
that
level
of
abstraction,
can
make
it
hard
to
fix
anything
yeah.
There
is
the
ongoing
discussion
of
like
what
does
a
cluster
of
discovered
local
nodes
on
one
machine.
You
look
like
like
how
would
that
process
function?
That
would
be
super
cool
feel
like
there
are
smaller
battles
to
fight
between
here
and
there
I
think
I
think
the
problem
resides
in
like
web
pages
and
web
extensions
talking
to
a
demon
that
has
assigned
itself
random
ports.
I
think
I
think
there
is
a
path
for
local
tooling
to
discover.
A
I
think
and
yeah
it
would
be
useful
to
have
Hugo
cuz
I'm,
pretty
sure
he
sort
of
solve
that
for
for
IP,
NS
in
web
browser,
prototype
and
and
the
idea
was
to
announce
like
DNS,
using
DNS
discovery
service
to
announce
like
ipfs,
I,
KN
s,
local
web
service
and
that
service,
like
act
as
a
bridge.
Everyone
talks
to
that
service
as
a
rendezvous
point,
and
you
are
always
sort
of
you
know
that
if
that
service
is
there,
you
will
get
the
idea.
First
note
I
just.
B
A
If
the
thing
that
is
listening
on
a
port
is
a
TFS
or
not,
and
then
telling
user
that
the
port,
the
default
port,
is
taken
and
do
you
and
do
you
want
to
start
a
BFS
on
a
different
port?
Or
do
you
want
to
quit
and
figure
it
out
on
your
own
I?
Think
that's
like
the
safe
default
that
would
both
make
less
technical
and
more
technical
users.
Happy.
B
We
had
a
related
request
the
other
day
on
the
issue
tracker,
for,
let
me
so
I
started
using
desktop
and
I
realized
that
I'm
using
IPS
a
lot.
It's
taking
a
pull
up,
storage
space
on
my
main
drive.
Can
I
rehome
it
can
desktop
be
like
move
my
ipfs
home
to
this
external
drive,
the
actual
request
was
like
ask
me
up
front
where
I
want
to
store
things,
and
so
then
there's
two
possible
solutions.
B
It's
like
have
a
kind
of
we
had
this
in
desktop
previous
version
that
there
was
a
step
during
install
that
asked
you
a
bunch
of
questions
and
to
streamline
things
we
removed
it,
but
that's
for
simplicity.
We
have
traded
off
some
user
understanding
or
certainly
I
would
quite
like
it.
If
it
was
easy
to
get
started,
didn't
ask
me
these
questions
and
then
allow
desktop
to
read
to
move
and
ipfs
home
directory
that'll
be
pretty
good.
I.
Think.
A
B
What
I'm
saying
is
you
start
as
a
new
user?
You
start
using
opulence
desktop
and
it
puts
a
bunch
of
the
whole
flock
blocks
into
a
block
store
in
your
home
directory
and
to
have
a
button
in
desktop.
That's
like
I
would
like
to
move
this
not
like
I,
want
to
forget
my
old
identity
and
create
anyone
with
nothing
in.
A
A
A
So
this
is
a
media
loaded
from
subdomain
gateway
provided
by
CloudFlare.
There
is
a
content
identifier
in
the
very
beginning:
it's
a
cid
v1
in
base
52
and
there's
there's
a
short
domain
name,
which
has
like
the
name
of
namespace
and
there's
like
actual
domain
and
there's
a
path
under
that
cid.
So
the
way
this
works.
A
That's
a
short
story,
and
the
thing
is
like
it's
easy
to
reason
about.
When
we
have
an
actual
domain,
what
happens
on
local
machine
when
you've
installed
ipfs,
you
run
local
gateway,
it's
possible
to
have
subdomains
and
the
local
host
host
name
and,
however,
we
don't
want
to
have
both
some
domain
based
and
path
based
gateways
on
the
same
host,
name,
I
feel
we
should
pause
there.
It.
B
A
B
B
A
It's
like
I
would
not
say,
secure
and
insecure.
It's
just
a
matter
of
when
you
want
origin,
isolate
like
when
you
want
to
isolate
stuff
and
when
you
don't
care
about
that
type
of
isolation.
So,
if
you
load
web
sites
from
ipfs,
you
care
about
origin
based
isolation,
because
web
sites
then
can
use
local
storage
cookies
and
they
cannot
edit
each
other's
cookies,
storage
and
stuff.
A
The
only
thing,
I
sort
of
see
as
a
path
to
move
forward
is
to
detect
leg
on
the
our
gateway,
should
detect
requests
to
localhost
name
without
any
subdomain,
and
then
redirect
that
to
just
IP
and
that's
sort
of
a
trick,
because
the
IP
is
interpreted
as
a
different
host
name
and
that
way
the
origin
sandbox
created
by
web
browser
will
be
separate,
and
then
we
have.
We
can
have
both
sub
domain
based
gateway
and
path
based
gateway
on
the
local
host
machine.
A
B
Fair
to
say
that
path
based
gateways
that
we
have
today
would
essentially
be
legacy
like
this.
There
isn't
a
good
reason
to
use
it.
We
would
want
to
migrate
people
to
subdomain
based
gateways
because
they're
just
there.
They
just
have
this
security
flaw
that
we
can't
work
around
yeah.
It's
like
borrowing
pushing
forward
work
on
sub
origin
specification,
yeah
music,
which
is
a
good
protocol
labs.
All
the
ocean
kind
of
strategy,
yeah.
A
So
it's
like
the
matter
of
what
are
you
using
ideas
for
generally
the
sub
domain
base?
Gateway
can
be
used
for
everything
at
and
solves
those
problems
that
we
had
with
the
path
of
based
one
but
I
believe
a
lot
of
people
will
still
want
to
continue
using
that,
mostly
because
they
don't
want
to
change
all
the
setups.
They
have.
A
They
often
put
ipfs
behind
nginx
and
then
use
the
path
based
addressing
on
our
gateway,
like
the
main
gateway
at
EFSA.
Oh
I
believe
we
will
want
to
at
some
point
just
redirect
everyone
to
the
sub
domain
based
one
at
the
DNS
link,
but
for
this
transitional
period
and
also
as
an
option
for
people
who
care
about
just
having
half
addressing
without
and
manipulation.
A
We
still
need
to
support
the
tough
ones
and
there's
like
a
discussion
between
me
and
Steven
on
the
PR
on
go
IDF
s
repo,
and
we
want
to
add
support
for,
like
native
support
for
sub-domain
gateways,
to
go
ipfs
right
now.
If
people
want
to
run
their
subdomain
gateway,
they
need
to
add
some
additional
code
at
the
nginx
level,
which
sits
in
front
of
Goa
Kiev
s.
What
we
want
to
do
is
to
make
super
simple
for
people
to
host
their
own
gateways
with
some
domain
support,
so
you
just
run,
go
ipfs.
A
A
A
Mean
like
the
way
it
will
work
if
you
share
the
old-school
IDF
si
o
/
IDF
is
something
that
from
desire
that
will
redirect
2d
web
link
mm-hmm,
but
we
we
could
support
like
subdomains
on
IDF
SEO,
but
those
would
redirect
to
do
web
link.
That's
a
separate
problem.
We
just
want
to
move
away
from
ipfs
io.
C
C
C
So
I
guess
that's
why,
in
as
much
as
I
understand
the
compatibility
concerns
and
I'm
glad
there
are
some
of
the
options
for
handling
that
I
really
do
feel
it's
important
that
the
understood
goal
is
to
not
support
path
based
addressing
as
long
as
there
is
no
way
to
mitigate
the
security
issue.
Oh
like
totally,
and
the
plan
eventually
hit
like
hacker
news
level
slammed
this
yeah.
A
And
actually,
we
sort
of
already
started
planning
her
around
that
and
the
idea
is
that
it
will
be
like
a
switch
that
you
are
not
able
to
have
both
on
on
localhost
IP.
You
are
not
able
to
have
both
subdomain
and
pop
address
the
Gateway,
and
if
you
make
a
request
like
half
addressed
requests,
it
will
be
redirected
that
will
be
like
HTTP
301
to
the
sub
domain.
So
that
way
links
will
not
be
broken,
but
you
won't
be
able
to
abuse
that.
A
B
Yeah,
there's
definitely
like
different
user
stories
here,
where
we
were
talking
a
lot
about
the
one
where
another
user
wants
to
stand
up
a
public
multi-tenant
ipfs
gateway,
but
what
what
tends
to
keep
surfacing
is
that
most
people
want
to
run
an
app
or
community
or
single
use
cases.
Justices
gave
way,
and
also
better
schooling
around
I
just
want
to
I.
Don't
want
people
to
be
able
to
request
any
CID
through
my
HTTP
game.
B
A
Other
thoughts,
I
just
have
a
quick
thing
to
show
ya,
just
a
quick
plug
about
immigration
to
see
it.
If
you
want
there's
a
an
issue
in
ipfs
ipfs
and
in
the
very
first
comment
created
by
kyle,
there's
a
link
to
a
big
bird's
eye
view
which
I
sort
of
went
over
recently
and
updated
a
little
bit.
So
you
can
see
all
the
PRS
and
issues
related
to
the
effort,
there's
of
course
I.
Think
more
than
that.
However,
it's
like
a
good
entry
point.
If
you
want
to
see
where
we
are
I'd,
be
right.
F
A
B
If
else
that,
otherwise
we're
going
to
mention
quickly
removing
special
cases
and
hacks
around
web
UI,
so
ofs
team
is
talking
about
releasing
more
more
regular
release
cycle,
but
at
the
minute
that
is
blocked
on
a
longer
really
focal.
While
we
get
our
testing
story
to
be
more
robust,
but
if,
in
the
general
direction
of
momentum,
there
is
release
more
and
more
predictable
frequency.
B
In
the
meantime,
in
many
months,
leading
up
to
that,
Riedel
and
I
have
slowly
been
adding
hacks
to
places
and
rolling
out
newer
versions
of
the
web
UI
in
in
ways
in
the
ways
we
could.
So
this
is
the
most
obvious
one
is
the
open
web
UI
button
in
companion
will
open
a
CID,
a
version
of
web
UI
that,
if
it's
gonna,
if
it
used
to
bundle
it
and
now
it
just
redirects
to
delicious
the
IDE,
but
it
has
to
do
some
tricks
to
make
that
work.
So
in
go
ipfs.
B
There
is
just
a
list
of
blessed
see
IDs
that
can
be
load
info,
that
it
can
be
loaded
by
the
api
port,
which
means
they
have
unrestricted
access
to
the
API,
and
that
list
of
see
IDs
is
all
the
released
versions
of
web
UI.
No
other
web
app
gets
that
blessing,
and
but
this
makes
it
means
that
we
can
only
release
new
versions
of
UI
by
PR,
Ingo,
ipfs
and
waiting
for
their
next
release.
B
So
the
proposal
is
to
remove
this
workaround.
So
I
have
companion,
simplify
things
so
that
it
doesn't
have
its
own
internal,
blessed
list
of
CI
DS.
It
just
defers
to
port
5000
one
slash
web
B
Y,
and
that
will
get
you
the
version
of
web
UI
that
go
out
if
it's
shipped
with,
even
if
that's
a
bit
old
and
then
as
an
extension
to
that
idea.
B
To
get
back
to
this
feeling
of
most
people
should
use
the
stable
version,
but
some
people
may
want
to
use
the
more
fleeting
edge
version
of
web
UI
to
include
in
go.
Ipfs
is
blessed
list
of
addresses
an
IP
NS
address
for
open-air
slash
web
UI,
DFS
tayo
and
then
have
a
toggle
in
companion
that
lets
you
say.
I
would
like
to
when
I
say,
open
web
UI
from
the
companion
drop-down
I
would
like
to
use
the
stable
version,
which
is
whatever
your
god
is
s
or
Jessica's
s
says?
B
E
E
Yeah
we
had
a
couple
of
conversations
already
a
good
way
that
we
can
like
a
great
way
box
to
use.
The
local
demon
would
be
to
find
a
way
to
connect
the
browser
and
the
local
daemon,
the
local,
even
being
the
the
best
of
us,
and
one
of
the
way
would
be
something
like
that.
Instead
of
like
shipping
with
where
you
buy
inside
just
about
the
open
web
UI
in
the
browser,
and
when
you
basically
ask
ask
the
browser
to
load
the
web
UI,
you
kinda,
you
have
a
way
to
send
some
information.
B
So
right
now
desktop
bundles
a
release
of
web
UI,
essentially
just
for
first
first
usage
performance
I
can
few
up
in
the
desktop
app
and
then
you're
like.
Please
wait
while
I
go
and
fetch
the
web
dy
from
ipfs
there's
I
mean
there's.
The
ongoing
question
of
my
desktop
could
actually
just
be
the
drop
down
menu
and
have
it
when
you
click
files.
It
just
opens
a
browser
and
then
you
could
have
a
query.
String
parameter
tells
you
what
API
put
to
use
happen.
B
Yes,
that
I
think
right
now
that
pile
of
logic
and
bundling
isn't
causing
us
any
problems.
So
what
with
what
we're
seeing
today
is
the
companion'
having
its
own
list
of
blessed
see.
Ids
is
running
afoul
of
it'll
load,
a
version
of
web
UI
that
go
ipfs
hasn't
blessed
and
it
strips
the
origin,
headers
and
fat,
then
I'm
sure
what
I
can't
remember.
A
It's
just
aside,
the
additional
context
is
that,
apart
from
regular
course
that
everyone
knows
on
is
happening
on
the
web,
go
ipfs
has
some
additional
hardening
when
you
request
from
the
different
origin
that
the
local
host
different
origin
than
the
api
port
on
localhost.
So
it's
like
hardened
chorus,
which
even
if
we
true
trick,
use
this
trick
of
removal
or
origin
header
still
the
code
that
is
on
the
go
idea.
Fest
side
will
return
forbidden,
HTTP
error,
it's
a
very
low
level,
the
the
gist
is
that
go
idea.
A
Fest
goes
extra
mile
on
top
of
course,
and
it's
just
a
losing
battle
to
try
to
work
around
and
build
additional
layers
and
keep
the
like
multiple
lists
of
blessed.
Where
do
I,
when
we
can
just
agree
that
now
that
go
ipfs
will
have
regular
releases
and
web
UI
recycle
slowed
down
and
we
are
sort
of
like
stable
right
now.
It
should
be
fine
to
just
delegate
web
UI
loading
to
go
ipfs.
F
Some
a
bunch
of
them
ellen
has
been
kind
enough
to
merge
for
me
already
in
github
spots,
but
I
wanted
to
just
run
by
people
some
ideas
for
how
we
would
surface
it
better
on
the
actual
websites
and
I'm
very
open
to
suggestions
here
so
I'm.
The
main
ipfs
website
once
that
I
see
that
might
work
naturally,
is
here
where
we're
offering
options
for
a
deeper
look
at
IP
FS.
Maybe
there
was
a
second
button:
that's
a
explore.
F
Proto
school,
tutorials,
I,
think
of
I
mean
I,
don't
think
I'm
gonna
pull
any
traffic
from
the
white
paper
because
I
think
of
these
as
different.
It's
this
you
care
about
these
things,
but
that's
one
plot.
Just
with
the
way
things
are
set
up
right
now,
it's
not
protocol.
It's
not
really
an
implementation.
They're,
an
app.
F
B
B
F
Right
any
other
ideas
on
standard
ipfs
do
so
Jae
sik
Festa
IO
I.
Do
it's
not
accepted
yet,
but
I
put
in
a
TR
so
that
when
you
go
when,
if
you
click
tutorials
protocol
tutorials
will
be
listed
alongside
the
ones
so
just
kind
of
live
in
a
repo
there
that
you
have
to
download
anything
explained,
then
you're
not
gonna
have
to
download
them.
C
F
B
B
E
D
F
B
There
is
one
more
item
on
the
agenda.
We
need
a
KO,
we
need
a
key
result.
I
am
desktop.
Ok,
we
can
see
through
eg.
We
keep
shipping
it
fixing
highest
priority
stuff,
but
not
low
priority.
For
irony.
Retinas
check.
Stop
you.
Web
UI
are
stable,
usable
eg,
we're
not
regressing
them
as
we
shipped
you
stuff.
This
quarter.
A
Sort
of
I
would,
on
top
of
like
keep
keeping
shipping.
It
like
thinks
that,
like
the
discussion
about
remove
removal
of
web
UI,
all
the
overhead,
all
the
hugs
and
unnecessary
optimizations
so
like
part
of
maintaining
heat
and
would
be
removal
of
of
standards
over
complicated
I
would
like
to
primary.
Somehow
in
the
key.
Our
who
added.
B
A
C
Encourage
you
to
add
the
wording
that
you
would
like
to
that
kr
this.
This
group
is
in
control
of
your
destiny
with
regards
to
how
we
continue
working
on
and
shipping
these
things
but
which
which
team
is
looking
after
it
right
now,
is
the
make
sure
that
they're
still
shipping
and
not
regressing
just
has
my
name
on
it.
Next
in
the
project
operations
Brett,
oh
so
I
would
love
your
assistance
in
writing.
Vkr
that
meets
your
needs
best.
Isn't
it
okay.
B
Well,
so
this
this
discussion
is
around.
We've
moved
focus
to
things
like
making
sure
we
do
a
good
job
of
the
sausage
managers
endeavor,
so
fixing
the
per
fishes
and
having
the
features.
We
need
to
support
exchanges
and
do
a
good
job
of
running
the
activist
gateway
and
empowering
others
to
run
high
reliable
gateway
firms.
B
So
that
has
bumps
down
the
amount
of
time
and
energy
that
we
will
have
to
work
on
desktop
and
the
joy
over
the
next
three
to
six
months.
But
we
haven't
checked
in
with
Henrique
about
what
he
would
like
to
focus
on
over
the
next
three
to
six
months.
So
it
may
be
that
and
really
focusing
on
desktop.
But
that
is
a
conversation
for
perhaps
now.
But
it's
worth
noting
this
that
my
still
and
Enrico
so
curcumin.
C
This
is
really
to
the
whole
group
right,
like
I
mean
this
is
after
that
we
need
to
focus
on
docs
and
performance.
Instability
services
are
up,
and
but
we
everybody
acknowledges
that
there
are
dependencies
still
on
things
like
like
web
UI
and
desktop,
and
there
are
really
strong
dependencies
on
on
working
and
not
not
progressing
and
making
sure
this
is
so
functional
from
the
brave
work
that
I
hope
to.
C
B
So
it's
worth
us
it's
worth,
making
sure
that
we
have
like
if
there
are
things
that
we
do
want
to
achieve
over
the
next
three
to
six
months,
that
we
are
clear
about
what
is
in
what's
on
our
flight
path
and
then
also
to
wonderful
community
members.
I.
Definitely
like!
Oh,
yes,
don't
worry,
we're
not
gonna
like
trample
all
over
the
work.
You're
doing
it's
that
stuff
that
you
were
interested
in,
like
that's
plenty
of
room
for
that.
There's.
C
B
One
piece
of
work
is
so
sweetly
done.
It
could
go
through
the
best
weather
issues
on
what
your
own
death
thought
and
just
give
them
like
a
triage
in
the
sense
of,
if
we
can
say,
like
that's
a
high
priority,
but
when
you
know
Help
Wanted,
this
one
doesn't
make
any
sense,
because
now-
and
this
one
is
stuff
that
we're
still
planning
to
do
in
the
next
three
months,
so
that
we
can
reduce
how
many,
how
much
time
with
them
that
we
throw
at
it.
G
For
a
minute
sure,
did
you
want
to
talk
about
that
right
now
we
can
basically
I
resubmit
that
PR,
we're
gonna,
add
identicards
to
the
peer
table.
So
when
you
have
a
CID
for
a
peer
right
next
to
it,
you'll
see
a
little
identic
on
that's
unique
to
every
your
ID,
so
it
makes
it
easier
to
identify
peers
that
are
being
used
as
relays,
and
things
like
that.
B
B
The
teaser
for
Enriquez
demo
next
week
here
in
the
current
format,
webui
is
focused
around
visualizing,
the
files
that
are
in
your
NFS,
your
mutant
file
systems,
view
of
your
IP
of
s,
repo,
which
is
just
a
tiny
sliver
of
what
can
possibly
be
in
there.
Anything
that
you
buy
PFF
tab
doesn't
show
up,
yep,
don't
necessarily
show
up
unless
they
and
Rhys
taken
this
idea
and
run
with
it.
So
rethinking
this
file
browser
so
that
it
can
not
only
show
like
arbitrary
ipfs
addresses.
B
So
today,
if
you
faced,
it
might
be
a
pet
address
into
the
Explorer
bar
at
the
top
opens
in
the
IP
of
the
Explorer
and
shows
you
the
underlying
data
model,
which
is
fun
but
way
too
low-level.
As
a
starting
point
like,
if
I
have
an
idea,
distressed
chances
are
one
of
CSS
content
as
the
IP
of
Ithaca.
So
he's
he's
done
a
large
PR
that,
like
changes,
the
focus
to
be
more
to
be
more
flexible
so
that
you
can,
like
average
content,
you
can
visualize
your
pins,
which
it's
gonna
be
round.
B
It's
so
clear
tuned
in
this
case,
so
the
great
web
GUI
and
in
and
then
it's
all,
she
wrote
if
there
is
any
more
business
you
have
one
last
second
interject.
Otherwise,
let's
goodbye
for
me,
you
could
buy
from
all
of
you.
This
has
been
the
ipfs
in
web
browsers
and
GUI.
T
weekly
call
see
you
next
week.