►
From YouTube: IPFS Web and GUI Biweekly 2020-08-11
Description
Meeting notes: https://github.com/ipfs/team-mgmt/pull/1171
About IPFS GUI and Browsers Biweekly: https://github.com/ipfs/team-mgmt/issues/790
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
Cool
welcome
to
gui
and
in
web
browsers,
bi-weekly
call
for
11th
of,
is
it
august,
2020
yeah?
So
that's
us,
and
probably,
as
always.
The
first
one
item
is
me
giving
an
update
on
pinning
services.
As
always,
so
I
won't
be
screen
sharing.
I
will
quickly
just
mention
that
pinning
service
api
spec
been
has
been
stable
for
over
a
week.
Now.
A
Well,
no
big
changes
are
planned.
There's
one
open
issue
about
disambiguating
those
providers
fields.
So
we
got
this
concept
of
provider
hints
which
aims
to
mitigate
the
problem
of
people
behind
the
retries,
restrictive
nuts
or
a
strange
network
topologies
when
peers
from
the
outside
are
not
able
to
connect
to
them.
A
Some
people
won't
have
a
relay
setup.
We
don't
have
a
like
relays,
are
not
that
popular
and
not
enabled
by
default.
So
there
will
be
some
peers
which
are
behind
nuts
corporate
nuts,
and
if
that
person
wants
to
pin
content,
they
will
send
the
request
to
the
pinning
service
using
this
new
api
and
the
pinning
service
will
probably
not
have
the
data
behind
the
cid
they
want
to
pin.
A
A
When
someone
wants
to
pin
something
they
provide
known
sources
and
then
in
the
response
they
get
a
list
of
multi-others
of
pinning
service
nodes
which
the
client
can
connect
to
and
that
way
pinning
service
has
a
direct
connection
to
known
source,
so
that
solves
the
weird
nut
problem.
The
problem
is,
we
use
providers
in
both
pin,
object
and
pin
status
object
and
there
was
proposal
to
disambiguate
it
so
in
the
notes,
there's
a
link
to
straw
polls.
A
So
that's
like
a
perfect
bike
shot.
If
you
care
about
how
things
are
named
with
the
straw
poll
is
open
till
thursday
and
will
either
open
pr
or
not
open
pr
depending
on
the
results.
So
that's
my
announcement
and
the
next
item
on
the
agenda
is
changes
to
web
ui.
So
we
will
have
a
new
web
ui
release
at
some
point,
but
we
had
some
nice
redoings
of
welcome,
screen
connectivity
area
pages.
I
think
jessica.
Do
you
want
to
talk
about
this.
B
B
Oh,
come
on,
oh
give
me
a
sec,
my
I
had
paused
to
anyway,
once
you
disconnect
it
no
longer
shows
you
the
welcome
information
boxes
crap.
B
This
is
the
worst
demo
ever
was
not
the
last
thing
in
my
buffer.
So
I
just
like
I
hit
up
in
her
and
then
just
like
something
completely
random
happened,
which
was
not
what
you
wanted
all
right,
so
idea
being
that
there
we
go
all
right.
So
if
you
cannot
connect
to
the
ipfs
api
notice
that
the
welcome
information
that
was
like
what
is
ipfs
that
appeared
below
no
longer
appears
there,
those
three
items
are
themselves
components
that
the
main
components
directory.
B
So
it's
totally
modular,
which
means
in
the
case
of
say-
and
you
have
to
sort
of
kit-bash
this-
to
make
it
work,
but
so
say
I'm
connecting
right.
So
I'm
connected
to
ipfs,
I'm
in
my
status
page
and
then
I
disconnect
you
get
this
entire
guidance
here
again,
including
the
api
reset,
the
api
config
here
there
are
also
a
couple
of
nice
things.
The
status
page
will
I'll
take
this
back
up
the
status
page.
B
If
you
click
on
it,
you
know
if
you're
like
anywhere
you're
in
files,
and
you
click
on
the
up
down
here.
It'll,
take
you
straight
to
the
status
and
it'll
have
all
of
the
advanced
stuff
expanded
which
then
you
can
also
go
to
the
api
config
edit
in
the
settings
itself.
So
it's
a
lot
more
intuitive
raphael
fixed
a
little
bit
of
a
bug
with
this
not
really
expanding
correctly.
I
am
about
to
attempt
to
fix
the
issue
that,
if
you
hit
this
edit,
it
should
take
you
all
the
way
up
to
there.
B
So
that'll
be
another
improvement.
The
other.
The
only
open
question
on
this
is
whether
surrey
now
the
the
right
now
the
status
screen
is
the
only
one
if
you
disconnect
suddenly
that
will
within
a
reasonable
amount
of
time,
throw
you
an
error
message,
the
other
ones.
B
If
you
are
if
you're
connected,
and
then
you
drop
your
connection
like
I'm
in
files
right
and
I
stop,
it
can
be
a
very,
very,
very
long
time,
basically
just
waiting
for
me
to
have
to
have
to
refresh
or
go
to
another
page
before
you
actually
get
that
information.
So
the
open
question
on
this
issue
is
whether
we,
it
would
be,
as
I
understand
it,
nice,
but
not
totally,
critical,
to
have
better
detection
and
refresh
on
loss
of
connection,
so
that
every
page
behaves
a
little
bit
more
closely
like
that
status.
Page.
B
If
you,
if
you
lose.
Your
connection,
sounds
like
that.
C
B
Yeah
yeah
yeah,
because,
right
now
it
is
a
component.
An
overlay
would
be
a
good
way
to
do
it
because,
like
you,
absolutely
cannot
ignore
that.
That's
a
good,
that's
a
good
call,
the
detection
and
refresh.
I
would
need
somebody
or
the
detection
and
addition
of
the
overlay.
I
would
need
somebody
else
to
do.
Unfortunately,
that's
a
little
bit
outside
of
outside
of
my
abilities
and
then
also
one
open
question:
how
about
we
just
change
this
text
to
read:
welcome
to
the
future
of
the
internet.
B
C
C
Like
there's
so
much
text
and
yeah
stuff
going
on,
maybe
we
can
like
make
something
more
obvious
to
the
user.
Like
start
using
files
start
messing
around
here,
so
the
user
can
learn.
B
B
Cool,
I
will.
I
will
fix
the
the
internal
link
problem
to
take
people
down
to
the
api
picker
on
settings,
and
then
I
will
kick
that
over
to
you
lytle.
If
you
have
time
to
play
with
the
detection
logic,
if
not
we'll
have
something
that
we
can
at
least
include
in
2.11.
A
A
Yes,
cool
yep,
so
just
sort
of
like
a
to
get
a
feeling
of
the
land
I
created
a
milestone
for
211
and
because,
basically,
we
wanted
to
like
cut
this
release
this
week,
and
then
I
realized
like
try
to
figure
it
out
what
needs
to
happen
to
avoid
painful
regressions,
which
will
make
people
open
issues
and
then
block
our
precious
time.
So
I
created
this
milestone
and
included
open
pr
by
iraq,
lee
which
I
really
will
review
today.
A
A
A
Over
vpn,
please
don't
expose
your
api
ports
to
the
internet.
I
assume
those
people
have
the
basic
off
or
some
other
means
of
protecting
their
api,
but
we
got
a
bunch-
probably
that's,
not
the
full
list.
I've
added
them
to
the
milestone,
but
those
like
marked
as
bag
and
related
to
gateway,
port
or
api
port.
A
Everything
is
like
a
bunch
of
those
issues
will
probably
be
closed
by
a
single
pr
where
we
refine
the
way,
one,
how
we
find
the
api
when
it's
not
found
on
the
default
port,
or
it's
not
provided
through
like
by
user.
Through
that
input
that
we
have
in
the
interface
the
edge
cases
are
when
api
is
on
catch
all
ip,
which
is
four
zeros.
A
They,
if
you,
if
web
ui
assumes
gateway
port,
is
at
four
zeros
images,
embedded
or
links
to
the
gateway
in
web
ui
will
be
interpreted
by
the
web
browser
by
us
resource
on
the
local
machine
instead
of
that
remote
node
that
a
person
connected
to
so
we
need
so
we
need
to
create
improved
heuristic
iraqi
already
switched
web
ui
from
from
that
redux
bundle
to
ipfs
provider
which
by
default,
tries
if
no
api
address
is
provided,
it
will
try
to
access
api
on
the
current
origin.
If
I
remember
correctly,.
A
Yeah
yeah,
so
that's
the
very
basic
logic
which
guaranteed
that
if
a
person
changed
the
default
api
port
and
web
ui
was
loaded
from
the
gateway
port,
it
would
automatically
work
because
it
tried
the
same
origin
as
the
web.
Ui
itself
was
loaded,
so
we
probably
need
to
improve
that.
I'm
not
sure
if
it's
in
ipfs
provider
or
it
will
be
in
the
user
land
in
the
web,
ui
itself
to
handle
those
additional
use.
A
Cases
where
there
is
either
that
catch
all
four
zeroes
in
go:
ipfs
configuration
and
the
user
loaded
web
ui
from
one
origin,
and
we
is
trying
to
access
api
at
the
different
origin,
but
a
bunch
of
those
issues.
I'm
sharing
right
now
will
be
closed.
If
we
simplify
that
logic,
I
also
like
reopened
one
issue
where
we
in
ipld
explorer.
A
We
have
this
link
to
view
on
the
ipfs
gateway
and
it
was
like
pointing
at
the
local
house
or
something,
and
we
changed
that
to
point
at
the
public
gateway,
because
most
of
people
will
have
ipvs
companion,
so
assumption
was
that
will
redirect
to
a
local
gateway.
But
the
problem
is
like.
A
A
That's
like
a
long
story,
but
the
gist
here
is,
apart
from
the
user
interface
improvements
and
performance
improvements.
By
directly,
we
have
the
only
issue
we
have.
Is
this
use
case
where
people
are
using
api
port
at
a
remote
origin,
and
we
need
to
tackle
that
before
cutting
the
release,
because
that
would
like
introduce
a
bunch
of
regressions.
A
A
If
you
feel
there's
something
that
we've
missed
and
should
be
added,
feel
free
to
just
like
add
the
issues
to
this
milestone,
so
we
don't
miss
it,
but
those
related
to
api
and
gateway
gateway.
Ports
are
effectively
just
one
issue.
I
I
just
like
did
not
do
a
proper
triage
and
did
not
merge
them.
D
Can
I
make
a
note
about
the
one
of
the
issues
that
are
part
of
the
milestone,
so
the
buffering
issue,
that
is
to
not
buffer
files,
while
we
add
them
to
the
ipfs
while
there's
a
pull
request,
it's
actually
awaiting
on
the
js
ipfs
released
to
be
cut,
because
some
of
the
changes
that
fixes
it
on
ipfs8
has
not
been
released
in
the
previous
version.
So
we
need
to
wait
for
a
version
to
actually
be
able
to
take
advantage
of
it.
D
A
I
believe
it
should
be
this
week,
but
yeah
next
is
probably
one
yeah.
I
don't
think
it's
like,
even
if
it
does
not
happen
for
some
reason.
It's
it's
not
a
problem.
I
think
the
the
most
important
part
was
already
merged
the
the
async
refactor
that
you've
did.
This
switch
to
http
client
with
performance
improvements,
sort
of
like
a
maybe
not
deadline,
but
the
the
window
that
we
want
to
align.
A
It
is
when
we
merge
the
feature
branch
with
pinning
services
that
sort
of
like
solved
deadline
for
for
those
improvements
to
land.
We
want
to
either
land
performance
improvements
before
or
with
that.
So
it's
not
a
big
problem.
If
js
ipfs
did
not
like
do
not
release
on
time,
we
may
cut
a
release
of
web
ui
without
that
and
do
that
later,
but
would
be
nice
to
like
the
sooner
the.
A
A
Cool
that
was
me
checking
the
agenda.
We
are
at
the
end
of
the
agenda
so
last
occasion
to
grow.
It
add
your
own
items.
I
will
go
quickly
over
highlights:
we've
been
busy
with
a
backlog
of
issues
and
other
stuff,
so
not
very
eventful
week,
but
we've
released
new,
it's
ipfs
library,
maybe
I'll
share
my
screen.
Just
so
people
don't
stare
at
me.
People
will
stare
at
my
notes.
A
And
maybe
that
proof
of
concept
that
we
had
at
share
ipfsio,
but
that's
about
it.
Since
then,
we've
created
a
framework
agnostic,
ipfs
provider,
so
ipfs
3x
bundle
makes
a
bunch
of
assumptions,
and
it's
probably
most
useful
in
in
react
and
redux
environments.
Ipfs
provider
is
something
you
can
use,
no
matter
which
framework
you
choose
and
it's
more
flexible.
A
A
I
believe
a
tidally
wiki
plugin,
which
supports
ipfs,
is
using
ipfs
provider
for
lazy
loading
and
it's
lazy,
learning,
http,
client
and
more
and
more
projects,
including
web
ui,
is
now
depending
on
ipfs
provider.
So
it's
good.
We
are
dogfooding
that
and
the
second
deprecation
is
not
really
done
by
us,
but
alex
from
js
ipfs
did
that,
but
it's
something
that
we
use
and
we'll
probably
have
to
remove
over
time
from
our
projects,
and
we
had
this
split.
We
had
two
npm
packages.
A
A
So
if
your
project
was
running
tests
against
ipfs,
but
you
did
not
want
to
like
install
it
doubly
or
something
that
package
included,
go
ipfs
binary
in
your
node
modules,
subdirectory,
so
alex
unified,
like
he
backported
all
the
features
from
goeiqfs
depth,
namely
ability
to
find
the
path
to
the
binary.
And
now
that's
provided
by
the
go
ipfs
package,
and
that
means
go.
Ipfs
dab
is
deprecated.
A
If
you
want
to
include,
go
ipfs
in
your
project,
use
just
go:
ipfs
use,
ipfs
provider,
I
don't
believe
yeah
redux
bundle
soon,
soon
or
currently,
being
deprecated
and
go.
Ipfs.
Dev
is
also
like
duplicated.
A
I
believe
we
also
duplicated
them
on
npm,
but
it's
worth
mentioning
here
in
case
someone
is
using
them
right.
Folks,
that's
all
my
highlights,
hopefully
in
when
we
meet
in
the
two
weeks,
we'll
have
web
ui
release
compiler
release,
oh
desktop
release,
whatever
we'll
release
a
bunch
of
stuff,
so
stay
tuned.
D
I
I
want
to
bring
something
up,
so
I,
as
I've
been
trying
to
refactor
a
bunch
of
things.
I
keep
running
into
things
that
I
find
somewhat
unclear
what's
happening
in
web
ui,
and
I
got
myself
into
trying
to
refactor
some
of
the
logics
there
and
I
could
use
some
help
in
trying
to
check
my
assumptions
whether
what
I
interpret
as
my
interpretations
of
things
are
actually
correct.
D
So
can
you
one
of
you
please
help
me.
Maybe
we
can
schedule
an
hour
or
so
so
we
can
go
through
everything
and
try
to
make
sure
that
it
is
the
case
and
if
I
shouldn't
be
doing
that,
that's
also
another
thing.
Maybe
you
should
tell
me
that
stop
doing
this.
A
Yeah,
I
honestly
would
like
to
have
like
an
hour,
ideally
with
me,
raphael
you
and
maybe
ollie
or
hack,
and
I'd
like
to
like
include
them,
because
they
were
from
the
beginning
of
the
project
and
they
may
be
more
familiar
what
some
decisions
were
made.
If
you
are
able
to
schedule
something
this
week
or
or
maybe
later
with
most
of
us,
I'm
happy
to
join.
D
Cool
yeah
I'll
schedule,
a
meeting
then
or
I'll
try
I'm
not
sure
about
people's
availability.
I
I
might
have
it
may
not
be
this
week
because
I
want
to
have
a
proposal
at
hand,
so
I
can
actually
go
through
it
and
then
I'll
do
that.
Yep.
A
Right
folks,
I
won't
keep
you
around
for
much
longer,
see
you
in
two
weeks.
Thank
you
for
joining
fingers
crossed
for
all
our
real
upcoming
upcoming
releases.
To
go
smoothly.