►
From YouTube: Web connectivity state, future transports, visualization - IPFS GUI and Browsers Biweekly 2020-04-21
Description
Meeting Notes: https://github.com/ipfs/team-mgmt/pull/1152
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
A
B
Everyone
I
Jacob
nice,
to
meet
you
as
well.
We
we
haven't
been
introduced,
yet
it's
a
pleasure,
so
I
work
at
moxie
same
as
last
crash
quiz,
a
old
friend
of
mine,
too
I'm
excited
to
be
here,
I'm,
a
very
UI
and
UX
oriented
developer,
I
always
like
to
provide
nice
features
in
terms
of
UX
and
really
pretty
in
terms
of
UI.
Since
I
have
a
friend
that
has
some
disabilities
visually,
he
has
a
80%
impairment,
visual
visibility
and
he
always
tells
me
about
these
problems
when
using
websites.
B
So
I
have
a
lot
of
attention
to
accessibility
because
of
that
so
I
guess
that
helped
in
this
specific
project,
because
I
believe
we're
lacking
a
bit
of
accessibility
and
I
want
to
bring
that
to
the
table,
improve
that
and
about
myself:
I'm,
really
passionate
about
animations
and
making
fast
and
nice
websites.
So
it's
I
guess
that's
my
profile
in
too
long
didn't
read:
format!
B
A
C
A
A
A
A
A
Cool
the
next
one
is
gosh
I'm
constantly
sharing
my
screen
and
not
sharing
it.
Maybe
I'll
just
keep
it
like
this,
alright
bit
chaotic
today
for
some
reason,
so
there's
this
web
platform
qubit
or
community
group
under
w3c,
and
they
have
a
forum
where
some
discussions
about
extending
the
web
platform
happens
some
proposals
and
there
is
a
topic
of
filling
the
remaining
gap
between
WebSockets
WebRTC
and
this
future
spec.
That
is
proposed
called
web
transports.
A
A
And
in
that
discussion
we
had
a
pretty
good
feedback
from
the
web
community,
namely
the
problem
with
building
dhts
I
got
highlighted
by
ferris,
and
then
we
as
a
browsers
and
community
special
interest
group
from
ipfs
and
the
p2p
projects
prepared,
like
a
short
summary
of
where
we
are
right.
Now.
What
are
our
primary
use
cases
and
how
we
try
to
address
those
use
cases
but
which,
like
each
transport,
each
API
has
its
own
limitations.
A
E
E
C
A
C
E
A
F
F
The
second
milestone
the
duplicate
appear
in
four:
it's
basically
the
couplet
in
all
these
pr's,
because
basically
we
got
rid
of
the
pure
info
in
all
the
code
base
and
the
API
and
basically
I
think
we
are
like
at
7
out
of
18
yards
merge
now.
Oh
the
other
one
are
ready
for
the
view
in
Jacob
has
been
reviewing
them
over
yesterday
today.
F
So
I
think
we
are
in
a
good
shape
here
to
get
this
milestone
all
merged
by
next
week.
Meanwhile,
I
will
speak,
I
will
start
in
oh
yeah
and
other
important
thing
about
muslin
to
basically,
as
we
are
removing
the
period
from
leap
HP.
Basically,
relief
info
was
with
what
we
were
using
to
provide
two
lipid
fee,
which
chalices
should
know
it
listened
to,
and
so,
as
we
know,
don't
provide
a
period
flow
anymore.
In
a
little
bit,
V
dot
create,
but
basically
we
are,
we
have.
We
are
providing
two
configuration.
F
The
list
of
addresses
that
appear
wants
to
listen
on,
however,
since
we
have
been
working
feeling
about
doing
another
thing
for
a
while,
which
is
more
than
supporting
the
with
no
drives
used
to
support
the
announced
in
no
announced
addresses
last
week
and
I
I
got
a
little
bored.
What
stunts?
No
friends,
basically
I
decided
to
create
an
address
management
which
will
end
all
these
listen
announced
in
announce.
Traces
I
also
seem
to
eject
about
it
and
yeah.
F
So
basically,
I
will
start
this
week
in
the
back
data
store,
which
will
add
the
persistence
to
the
address
book
and
wrote
a
book
and
up
fully
by
our
next
call.
I
aim
to
have
the
old
milestone,
PR
master
on
to
PRS
merged
and
the
third
milestone
ready
to
review
and
yeah.
That's
it's
any
questions
or
comments.
B
F
It
singles
I,
don't
think
so,
because
I
think
this
will
be
only
internal
changes
from
ipfs
land,
so
I
will
I'm
also
working
on
integrating
there's
these
new
breaking
changes
from
just
a
bit
to
be
into
jessica,
fest,
and
so
basically
it's
our
ipfs
consumes
liberty
later.
So
I
guess
it
will
not
change
nothing
so
far,
but
if
it
changes
it
will
be
a
long
time
until
it
gets
to
the
web
UI.
So
we
will
sink
later.
If
we
need
all.
D
F
Basically,
our
goal
is
to
have
the
format
the
for
initial
milestones,
which
is
the
p0
for
this
quarter,
and
for
that
I'm.
Almost
it's
the
off
the
word,
maybe
a
little
bit
more
now,
because
basically,
what
I
need,
except
for
getting
your
if
you
merge
from
s1
to-
is
to
introduce
the
persistence
for
the
mouse
country,
which
I
think
it
will
be
not
that
hard.
F
Considering
how
things
are
and
from
the
address
book
and
brought
a
book
there.
I
did
before
and
then
the
key
book
formats,
one
for
which
also
a
bit
because
Shane
and
persistence
for
the
keys,
which
will
be
a
little
more
tricky.
But
I,
don't
expect
that
any
major
problems
and
then
the
goal
is
to
firm
minjae
congested.
The
BGP
to
do
are
see
maybe
in
three
four
weeks,
and
it
will
include
all
these
and
some
other
stuff
that
Jacob
will
be
working
on.
F
D
D
C
Yeah
sure
I
mean
I
can
try
to
share
my
screen,
so
you
can
all
look
at
it.
Pr
alongside
me.
Actually
there's,
probably
some
stuff
I
can
show
you
too
so,
can
y'all
see
that
yes,
okay,
cool
cool,
all
right,
so
this
sort
of
started
out,
as
fixing
like
two
little
isolated
things
that
were
hanging
on
there,
ipfs
CSS
repo
and
is
now
like
kind
of
turned
into
a
major
release.
C
So
we
have
improved
the
icons
reference
page
so
that
you
actually
list
the
name
of
the
icon
in
question
and
I
also
went
through
and
reconciled
on
the
missing
icons
where
we
had
like
one
in
the
stroke
version
and
one
in
the
glyph
version.
So
now
we
have
all
of
these
icons.
Some
point
in
the
future.
I
want
to
clean
them
up.
C
They're,
not
super
consistent
between
each
other
and
I
think
it
could
actually
stand
a
some
concentrated
visual
effort,
but
but
we
have
them
all
and
now,
which
is
great,
also
replaced
the
basic
teal
color,
with
a
slightly
darker
teal
color
for
accessibility
purposes.
That
now
means
that
we
get
a
button
where
you
can
have
white
letters
on
a
teal
background
and
that
passes
or
CAG
I
think
Double
A
also
added
support.
We
had
a
we
had
a
community
contributor
who
came
in
and
added
support
for
SAS
s,
which
is
cool.
C
However,
it
was
just
static.
It
wasn't
automatically
generated
from
the
core
variables
file
that
lives
at
the
base
level
of
this
project.
So
added
that
so
that
you
know
there
tries
to
have
both
a
CSS
and
just
like
regular,
plain
old
CSS.
If
you
don't
want
a
CSS
support
and
you
don't
get
your
color
variables
but
as
I
understand
correctly,
I
think
electron,
like
out
of
the
box,
doesn't
play
well
with
s
CSS.
Just
the
box
I
think
you
actually
have
to
add
some
modules
to
make
that
work
so
I'm,
so
I
left
both
them.
C
It
does
update
to
the
latest
version
of,
inter
which
is
a
fairly
substantial
change
to
that
thought,
got
a
lot
of
things
that
are
just
a
lot
easier,
particularly
if,
like
a
western
character
based
language
is
not
your
usual
language,
so
things
like
the
lowercase
L
and
the
uppercase
I
are
a
little
bit
easier
to
tell
apart
also
reconciled
the
font
so
that
we've
got
300
through
800
weight
for
both
of
the
core
fonts
Lyle
I
need
to
check.
In
with
you,
you
found
something
that
I
think.
C
Is
you
left
a
comment
on
this
about
some
like
extra
files
and
I
think
something's
getting
weird.
These
are
set
up
as
sub
trees
and
I
think
what
you're
actually
seeing
is
like
an
earlier
version
of
inter
UI,
which
was
what
injury
used
to
be
called.
So,
if
you
don't
mind,
I'd
like
to
chat
with
you
about
like
what
you're
seeing
in
your
end
and
then
just
a
whole
bunch
of
sort
of
random
random
little
fixes,
so
that
should
be
almost
done.
I
know,
I
was
waiting
on.
C
You
know
I
think
if
we
do
this
as
a
major
release
to
your
point,
Lydell
that
that'll
that's
a
little
bit
safer
for
folks,
so
I
am
inclined
I
think
all
he
had
a
quick
glance
at
a--
and
he
mentioned
what
the
deal
was
with
the
subtrees,
which
I
was
not
aware
of,
but
I
think
I
think
once
we
get
the
sub-tree
reconciling
figure
it
out
Lytle.
If
you
smoke
tested
it
and
you're
happy
I
think
we
should
just
push
this
out
the
door
and
move
on
to
other
stuff,
yeah
I.
A
As
expected,
so
I
think
that's
a
good
sign.
I
would
not
spend
too
much
time.
It's
possible.
I
had
some
like
leftovers
from
the
old
build.
It
was
late
when
I
tested
it
and
like
when
I
reported
those
additional
funds.
I
just
mentioned
it,
because
if
we
don't
use
them,
it's
like
1
megabyte
overhead
yeah,
yeah.
C
And
I
did
I
did
add
some
fonts
that
are
not
explicitly
declared
in
any
of
the
CSS
right
now,
because
I
intend
to
account
for
them
in
future,
because,
right
now
we
don't
have
full
support
on
both
base.
Fonts
for
300
wait
all
the
way
up
to
800
wait.
We
had
some
holes,
so
I
went
in
and
I
did
add
those
even
suits
so
they're
not
explicitly
in
the
CSS
rules
right
now,
but
they
will
be
at
some
point.
What
you're
seeing
I
think
was
actually.
F
C
I
do
think
I
think
there's
room
for
those
just
for
consistency,
also
just
like
if
you're,
if
you're
coding.
On
top
of
this
you
know
you
don't
expect
to
have
like
a
hole
for
your
500
font,
and
it's
just
like
blows
out
in
a
weird
way.
So
also,
if
somebody's
like
explicitly
on
there
and
like
if
somebody's
starting
with
this
kit
and
explicitly
declaring
like
a
500
weight,
we
don't
want
it
to
be
missing
for
them.
So
even
then
we're
not
using
it.
It's
still
like
a
gracious
thing
to
do.
C
I'm
just
a
little
concerned
that
I
think
the
things
that
you
listed
in
there
we're
actually
stuff
like
because
one
of
them
was
inter
UI,
which
is
the
old
name
which
shouldn't
even
be
in
there
at
all,
so
I
think
there's
something
weird
going
on
with
the
subtree.
So
if
you
don't
mind
doing
like
a
clean
and
a
rebuild
on
it
and
seeing
if
those
are
there,
it
might,
it
might
just
be
stuff.
That's
left
over
on
your
local,
but
I
want
to
make
sure.
A
B
B
Yes,
yes,
so
I
took
into
consideration.
The
review
Marcin
gave
me
and
I
reduced
the
number
of
thoughts
and
I
aggregated
them
by
a
distance
of
I
believe
is
10
of
these
small
dots.
So
if
you're
an
area
close
to
10
dots
it
aggregates
into
a
single
group,
so,
for
example,
if
we
over
here
in
Spain
in
Portugal,
you
can
see
that
there's
a
lot
of
connections
coming
from
Spain,
Portugal
and
France,
so
it
actually
just
tells
oh
I
need
to
fix
this,
though,
when
I
want
to
over
over
here.
B
So
it
tells
us
that
they
are
five
visible
connections
and
14
plus,
and
then
it
highlights
in
the
table
as
well.
Let's
see
if
I
can
see
any
more
French
connections
here
there
he
is
so
it
depends
on
the
part
of
the
map
it
is
I
can
adjust
these
parameters
easily
I
think
it
depends
a
bit
on
the
area.
Maybe
I
can
add.
Some
criteria
like
for
coordinates
for
specific
coordinates
like
America.
We
can
probably
just
aggregate
east
and
west
of
America
I,
don't
know
how's
how's
the
best
approach.
C
B
Mean
I
just
set
up
a
quick
possibility
of
my
out
of
my
head
just
to
provide
you
like
a
list
of
possibilities.
We
can
add
them,
but
if
you
just
want
to
make
like
10
dots
around
just
aggregate
10
dots
around
it,
it's
easier
for
all
the
world
instead
of
just
specifying
for
each
part
of
the
world.
But
if
it
provides
useful
information
for
the
user
I
guess
we
can
do
it
by
state
and
by
country
and
everything.
C
A
Super
super
super
helpful
visual
to
understand,
because
previously
we
had
just
like
a
lot
of
dots,
which
was
one
on
top
of
another,
which
gave
a
pretty
good
picture
that
how
they're
like
how
dense
stuff
is
in
Europe
or
each
side
of
us
or
in
Asia
the
problem
was
you
and
still
it
kind
of?
Is
it's
hard
to
quickly
eyeball
the
like
the
order
of
magnitude
like
that?
You
have
ten
dots
here
and
you
have
a
hundred
dots
there
when
you
put
them
one
on
top
of
another.
They
look
the
same.
A
You
lose
after,
like
ten,
you,
basically
it
looks
the
same
and
I
feel
we
need
to
figure
it
out
the
way,
either
by
the
size
or
the
or
or
some
other
way.
Just
like
I
said
maybe,
and
if
it's
more
than
a
hundred,
maybe
we
put
a
number
because
selected
dot
itself
is
big
enough,
so
you
can
put
a
number.
So
that's
the
certainly
something
we'll
figure
it
out
going
forward.
It's
super.
C
Exciting
I'm
gonna
link
I
hate
to
do
this,
but
I'm
gonna
link
to
the
New
York
Times
coronavirus
map,
because
they're
doing
good
things
with
sizes
of
circles
and
numbers
inside
of
them
and
I
think
a
lot
of
us
have
been
stared
at
that
map
way
too
much
lately,
but
it
is
also
kind
of
rapidly
establishing
a
visual
precedent
for
a
lot
of
people
right
now.
I'll
throw
that
into
the
issue.
B
D
All
the
research
findings
are
linked
there
and
are
in
a
good
book,
but
reading
it
relates
to
us
and
browsers
and
to
the
extent
that
we
haven't
really
like
our
first
release,
was
ipfs.
Support
in
Opera
on
Android
was
our
first
mobile
support
but
I
think
as
we
move
forward.
We
need
to
be
able
to
have
a
clear
picture
of
how
IP
this
addresses
move
between
applications
on
mobile.
What
the
end
user
flow
is
for,
how
that
is
possible?
How
do
people
get
see
IDs?
D
D
What
a
node
on
mobile
might
be
understanding
of
what
differently,
either
like
light
client
approaches
versus
full
node
approaches
and
finding,
where
the
sweet
spot
is
there,
where
we
can
give
as
much
decentralizing
user,
centered
user
data
ownership
capability
on
mobile,
but
without
burning
down
the
CPU
a
battery
and
eating
up
everyone
bandwidth.
So
hopefully,
some
of
this
research
will
help
us
narrow
down
from
a
protocol
implementation
standpoint,
similar
requirements
and
then
for
the
next
10,000
apps
built
on
mobile,
that
use,
PDP
or
ipfs,
make
them
not
suck
and
hopefully
be
awesome.
A
E
A
Well,
if
I
want
to
share
a
black
night
someday,
that
would
be
useful,
but
all
right
so
I
will
talk
about
my
remaining
items.
Apologies,
that's
I!
Don't
have
a
proper
set
up
to
share
so
I
believe
I
mentioned
about
native
protocol
handler
API
link
to
the
grant
grant
is
in
the
notes.
The
thread
discussion
thread
about
filling
the
remaining
gap
between
web
sockets,
WebRTC
and
Web.
Transport
is
there
as
well?
A
If
you
want
to
retweet
d-trix
tweet,
it's
also
there
and
I
have
a
question
which
I
hope
to
back
up
by
a
demo,
and
now
I
will
explain
to
you
how
I
would
demo
it
and
you
need
to
use
the
power
of
your
imagination
to
think
how
it
would
look
like
if
Lydell
show
that
to
me
to
make
his
case.
So
we
have
a
browser,
extension
called
ipfs
companion
and
we
have
ipfs
desktop,
which
is
a
GUI
application
for
managing
local
aquifers
node.
A
So
now
we
are
in
the
process
of
releasing
goai
TFS
0.5,
that's
the
actual
demon
which
is
orchestrated
by
Agatha's
desktop
right.
So
we've
go
ipfs
0.5.
We
will
support
subdomain
gateways,
which
I
talked
so
many
times.
No
one
wants
to
hear
about
it
anymore,
however,
while
I
was
implementing
that
I
also
implemented
HTTP
proxy
mode,
so
HTTP
proxy
is
a
service
which
you
can
use
for
asan,
like
intermediate
node.
A
The
HTTP
proxy
does
not
support
encryption
like
encryption.
That
TLS
needs
something
that
you
need
to
put
on
top
of
the
proxy.
You
need
to
put
nginx
in
front
of
it,
set
up
certificates
and
you
are
not
able
to
get
a
certificate
for
a
domain
name
that
you
don't
own,
so
that
effectively
means
even
though
you
can
load,
let's
say:
ipfs
dot,
IO
or
lip-lip
p2p
dot
IO
website,
which
is
backed
by
DNS
link.
You
can
fetch
that
from
HTTP
proxy
that
will
be
on
the
HTTP
colon
slash,
slash
protocol.
A
So
that's
why
that's
the
main
blocker
why
we
don't
have
like
don't
keep
the
original
domain
name
in
the
address
bar
and
need
to
have
that
dot,
IGFs
that
local
host
port
suffix,
however,
the
the
en
s
ET
room,
naming
service
system,
they
have
this
custom
top-level
domain
name
called
that
eth
and
it's
impossible
to
get
certificate
for
that
hostname.
If
you
use
that
TLD.
A
That
means
there's
no
problem
of
HSTs,
headers
being
present
and
breaking
and
breaking
the
connection,
because
if
the
browser
sees
the
HSTs
header,
it
will
force
TLS
encryption
and
if
we
use
HTTP
proxy,
that
proxy
won't
be
able
to
support
their
GPS
I
feel
that's
pretty
long
description,
so
I'll
just
summarize
it
once
again,
we
may
be
able
to
keep
ETH
addresses
in
location
bar
with
companion.
The
question
is:
should
we
do
that
by
default
or
out
of
the
box,
when
someone
has
IP
first
desktop
and
IP
first
companion
installed,
or
should
we
maybe
that's?
A
D
Iii
think
we
should
definitely
bring
it
up
again.
I
think
we're
only
starting
to
walk
down
the
road
of
asking
these
types
of
questions.
What
the
long-term
ramifications
are
from
a
from
a
user
experience
standpoint
I
have
real
concerns
about
creating
URLs
that
are
not
that
the
U
is
not
there
that
they
are
they
break.
You
know
that
that
fundamentally
defeats
the
purpose
and
goal
of
the
BFS
project.
D
However,
I
understand
the
value
that
short
term
usability
improvements
like
readable,
URLs
or
you,
readable,
RLS
or
readable
addresses
may
have
in
the
acceleration
and
adoption
of
the
project,
also
as
a
experiment
in
different
ways
to
be
able
to
marry
naming
systems
with
the
address
ability.
So
I
would
encourage
us
to
keep
poking
up
this
from
an
experimental
standpoint.
I,
don't
think
that
I'm
worried
also
about
machinery
that
we
build
up
around
user
experience
that
we
eventually
must
turn
out.
D
If
this
does
not
materially
contribute
to
a
permanent
part
of
our
stepping
stones
that
we're
laying
down
on
the
roads
to
a
universally
addressable
I've
confessin
distributed
web,
then
we
should
be
very
cautious
about
encouraging
developers
and
designers
and
businesses
to
to
build
on
to
this
Rube
Goldberg
machine
if
we're
going
to
pull
it
out
from
under
them.
Eventually,
so
there's
some
trade-offs,
probably,
but
we
can
use
language
like
experiment
and
things
like
that
to
people
say
and
and
writing
about
it.
I'm,
quoting
you
about.
D
Okay,
we
blog
post
this
to
say
we
introduced
this
feature
here
are
the
constraints
here
are
the
benefits?
We
don't
know
what
it
really
looks
like
yet
many
people
are
experimenting
this
area
so
delivering
something
like
this
in
a
way
that
makes
it
very
clear
that
this
is
not
the
end-all
be-all,
but
we
something
to
learn
from
and
play
with
talk
about
the
pros
and
cons.
I
would
really
encourage
that.
D
A
It
sounds
like
starting
with
opt-in
experiments.
We
have
like
an
experiment,
experiment,
section
in
components,
preferences
added
there
as
an
opt-in,
as
probably
it
will
make
a
demo
hopefully
next
week.
So
that
will
be
clear,
because
it's
super
easy.
When
you
see
it,
it's
super
hard
to
explain
it
or
maybe
I'm,
just
not
the
best
person
to
explain
it.
A
Yeah
and
I'd
also
totally
agree
with
the
fact
that
the
local
Gateway
URLs
are
not
Universal.
However,
there
are
still
better
than
a
URL
which
looks
like
a
thing
that
should
work
and
does
not
work,
because
right
now,
when
you
enter
IP
FS
that
ETH
it
will
break.
However,
if
you
have
a
companion
it
will
like
it
will
fix
that
load
and
load
it
from
a
local
gateway.
The
address
bar
will
change.
It
will
be
IDF
as
that
ETH
that
IP
NS
localhost,
but
at
least
user
knows
hey
it's
from
the
local
host.
A
F
A
F
But
from
infra
perspective
we
are
bearing
on
basically
deploying
the
relay
server
instead
and
use
in
recommend
people
to
use
the
pub/sub
one,
because
it's
more
performant
we,
while
in
the
Stardust
we
don't,
we
are
actually
doing
a
double
encryption
because
of
needing
to
use
Sakai,
oh
and
then
the
less
s.
While
if
we
do
the
up
subversion,
it's
way
simpler,
both
in
implementation
and
in
performance
as
well
and
yeah.
A
Cool,
thank
you
so
my
so
effectively
for
now.
If
someone
wants
to
use
chess
ipfs
in
a
web
browser,
they
either
need
to
use
the
older
version
with
WebSockets
tar
or
they
need
to
set
up
their
own
WebSocket
starter
right.
They
kill
me
the
only
WebSocket
Stardust
instance
that
there
is
is
the
one
provided
by
match.
I
think
right.
F
Yes,
I
think
so
as
well.
They
can
use
the
WebRTC
star
wound,
which
we
don't
have
as
well
deployed,
but
that
it's
in
a
more
advanced
stage,
because
we
Jakob
already
open
the
issue
for
the
infra
team,
but
we
are
still
waiting
on
them
to
deploy
the
newest
version
with
a
sink.
Wait,
stop
so
for
now.
People
also
need
to
use
their
own
server
to
to
use
it,
and
hopefully,
in
the
near
future,
the
relay.
A
F
A
A
A
A
A
F
That
the
WebRTC
star
server
is
also
on
the
way
to
getting
deployed.
We
already
have
the
issue
an
inference.
That
would
be
the
short-term
thing
to
have
deployed
so
that
people
could
use
as
a
public
server
and
after
that
it
will
be
the
the
relay
server.
I
I,
don't
I,
don't
know
expectations
for
the
time
line.
12
ever
the
web,
RTC
is
deployed,
but
we
already
have
the
the
auto
doctor
set
up
for
automatic
deploys
and
we
already
open
it's
the
issue
for
inference.
A
A
We've
seen
this
problem
with
various
projects
on
the
on
the
github.
When
I
was
like
looking
for
people
who
are
building
stuff,
nearly
everyone
was
using
the
default
servers
like
the
default
web.
Sockets
are
ones,
and
this
might
be
probably
better
like
healthier
situation.
When
people
learn
to
self
house
from
the
get-go.
F
Yet
ago,
the
goal
is
to
people
to
deploy
their
own,
because
the
public
is
I
would
say.
The
public
is
mostly
for
demos
and
to
for
people
to
experiment,
but
when
people
start
to
really
creating
their
own
project,
they
should
also
run
their
own
server,
because
one
of
the
things
that
we
say
from
the
beginning,
in
all
the
tutorials
and
examples
for
these
reports
is
that
we
don't
guarantee
that
a
developer
leti
of
the
server's.
A
D
I'm
and
liking
that
more
and
more
people
are
piling
into
that
thread.
Sharing
perspectives
it'll
be
very
interesting
to
see,
ultimately,
the
long-term
impacts
of
these
types
of
threads.
This
present
cult
to
understand
and
may
have
nothing
specific
material,
but
I
think
it
definitely
helps
us
to
be
very
clear
about
one
what
our
use
cases
are
and
to
what
our
web
platform
gaps
are.
So
I
think
it
was
great
that
we
spent
invested
the
time
and
articulating
that
well
thanks
everybody
who
jumped
in
and
helped
them
get
feedback
on
the
post.
A
A
Let
me
quickly
check
if
there
is
its
turn
I
think
I
missed
I'm
a
mess.
This
way
the
screen
sharing
thing
totally
distracted
me
I
think
we
covered
all
all
the
topics
we
had
on.
The
agenda
expect
more
interesting
work
in
the
web
UI
land
in
the
following
weeks:
I
try
to
figure
out
how
to
demo
ETH
experiment
next
week
in
two
weeks,
it's
a
Buick
recall,
in
sum,
if
someone
did
not
notice
it,
it's
biblical.
So
I
have
more
time
to
fix
my
screen,
sharing.