►
From YouTube: WebUI E2E in go-ipfs, browser design & native impl. - IPFS GUI and Browsers Weekly 2020-01-15
Description
About IPFS GUI and Browsers Weekly: https://github.com/ipfs/team-mgmt/issues/790
IPFS Mirror: https://ipfs.io/ipfs/bafybeieo5gqa6t3y7hlmwybaexz2dd57vxjq7hzoejyam65bxqew2c2opa/
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
All
right
welcome
to
go
anymore
browsers.
It's
15th
of
January,
2020,
I,
won't
say
welcome
to
20s
cuz
I've
been
told,
that's
a
problematic
issue
when
a
decade
starts
or
ends
there.
People
have
like
arguments
about
that
on
the
internet,
so
this
week
is
bit
shorter
on
the
agenda.
We
got
some
topics
discussed.
First,
one
is
sentient
short
update
on
end-to-end
tests,
I
think.
A
If
you
go
yesterday
on
the
jetty
of
SPR-
and
it
seems
there
are
bugs
in
the
existing
CI
setup,
so
Hugo
is
taking
another
look,
I
unpack.
The
context
then,
can
transfer
that
to
him
yesterday,
so
I
moved
to
go
ipfs
and
actually
I
got
it
working
today.
So
I
did
not
even
like
update
this
to-do
list
before
this
call,
and
it's
still
in
the
draft,
but
it's
actually
like,
maybe
I,
that's
a
good
thing.
I
was
like
commenting
without
sharing
screen,
so
it's
probably
ridiculous
for
people.
A
Let's
see
now
cold
cold
cold,
no
go
back.
Yes!
So
there's
a
PR
enjoy
DF
as
we
power
to
to
basically
add
basically
run
those
existing
end-to-end
tests.
We've
got
in
a
keifa's
web
UI
repo
again
the
head
of
go
ideas,
so
the
setup
is
that
we
got
a
special
like
docker
image
with
browsers
it's
mostly
because
we
ran
tests
in
real
chromium
and
it
requires
some
additional
libraries
and
then
we
got
like
separate
web
UI
tasks,
which
runs
in
parallel
with
other
interrupt
tests.
A
A
For
some
reason,
however,
if
upstream
is
green
and
then
if
we
switch
to
log
to
this
developer
version
of
Goa
TFS
and
that
one
breaks,
that
means
we've
got
a
regression
and
that's
basically
it
the
electrical
caching
solutions
provided
by
psycho
CI.
So
it's
pretty
fast
and
actually
I've
run.
It
I
run
here
and
runs
like
under
seven
minutes
or
something
and.
A
You
can
see
like
it
runs
here.
It
runs
in
test.
It
builds
web
UI
and
then
run
is
using
the
stable
production
version,
which
is
right
now
with
zero
422,
it
all
green,
and
then
we
run
tested
again,
but
this
time
we
run
it
against
this
specific
commit
from
this
specific
branch
for
which
entire
tests
were
executed.
A
The
same
thing
that
we
will
have
for
JH,
HFS,
and
probably
it
will
land
like
this
week
depending
on
fix,
is
provided
by
Hugo
for
J's,
because
I
don't
want
to
like
open
too
many
peers,
so
probably
just
merge
all
the
fixes
for
both
I
think.
That's
it
for
them
to
enters.
The
cool
thing
is
that
we
would
found
some
bars
in
existing
setup,
so
we
are
harden
our
CI
and
I
like
I,
like
the
fact
that
Ingo
idea
files
I
did
some
timing
and
interrupts
tasks.
A
A
A
Yeah
and
cool
thing
is
that
we
are
like
forbidding
people
from
creating
when
they
initialize
idea.
First,
we
thought
that
it's
impossible
to
Esha
initialize
repo
with
small
key
sizes.
Now
so
you
are
not
able
to
like
use
like
1k,
let
heard
thousand
24
bits
you
need
to
have
at
least
like
2000
bitter
keys
and
in
CI
we
we
use
a
special
variable
to
lift
that
limitation.
Just
so
the
test
runs
faster,
but
took
some
time
to
figure
out
why
it
does
not.
B
A
So
like
in
both
Jas
and
go,
those
PRS
are
still
in
a
draft
state
and
I
just
want
to
sleep
them
both
when
I
merge
all
the
sup
stream,
because
we
had
like.we
fix
some
things
in
the
edger
at
the
first
decontrol
web
UI
itself.
It's
like
now
like
start
several
random
port,
so
you
can
run
multiple
ones
on
the
same
hardware
and
stuff
like
that.
So
it's
just
a
matter
of
merging
upstream
stuff
and
then
slipping
those
PRS
and
asking
like
Stephen
in
Goa,
TFS,
probably
and
Alain.
B
A
B
The
the
next
next
issue
that
I
and
we
talked
about
doing
a
brave
post
kind
of
like
a
more
technical
dive,
something
sprained
specific
acknowledges,
braves,
1.0
acknowledged
as
to
what
ipfs
work
shipped
in
that
release,
and
I
think
that
that's
great
if
we
can
get
that
out
before
the
end
of
the
month,
that
we
also
have
a
kind
of
set
up
before
each
denver
of
a
thing
to
point
out
and
talk
about
there.
So
I
wonder:
I
was
thinking,
maybe
for
this.
B
A
A
A
B
B
Like
Galia
does
a
work
for
a
long
time
doing
native
implementations
of
technologies
for
web
browsers
if
limiting
web
standards-
and
we
approached
them
in
December
around
just
have
a
early
conversation
with
what
native
implementation
in
chromium
would
look
like.
They're
super
experienced
long
and
long
involved
in
standardization
and
had
landed
major
major
things
like
CSS
grid
they're,
really
committed
to
open
source.
The
open
web
is
a
platform
and
very
experienced
in
landing
code
in
browser
vendor
repositories,
so
they're
part
of
browser
vendor
teams
they've
committed
access
to
the
depositories.
B
So
when
we
met
in
summer,
I
basically
give
them
kind
of
like
a
intro
into
just
30
doing
at
the
background
of
PL,
where
we're
coming
from
in
terms
of
in
source
by
default
and
the
values
of
the
organization
and
a
quick.
You
know,
look
at
my
professors
and
technologies
and
record
a
review
of
the
work
that
we've
done
so
far.
I'm
working
with
browser,
vendors
and
I
think
we're
gonna
meet
again
and
kind
of
intro
into
kind
of
how
a
Galia
works
and
what
their
approaches
are,
how
to
evaluate
projects
to
take
on.
B
So
the
idea
of
chromium
first
supports
the
two
partners
that
we've
already
been
working
really
closely
with
brave
and
opera
in
in
a
positive
way.
It
also
makes
that
a
possible
choice
if,
if
Microsoft
as
they're,
now
I
think
they're
about
to
launch
the
edge
chromium
as
their
default
browser
is
pretty
soon.
So
it
makes
the
conversation
for
them
if
they
enter
crypto
space
and
more
and
more
strongly
or
want
to
build
ads.
B
B
If
we
take
that
approach
first
and
then
you
know
conversations
with
other
browser
vendors
s
as
as
usual
still
ongoing,
but
interesting
opportunity
with
worth
highlighting
there
and
then
just
also
a
quick
update
on
the
on
the
browser
design,
guidelines
I
think
slide
all
for
doing
a
massive
pass
on
it.
Jim's
gonna
integrate
your
comments.
B
This
would
not
be
limited
to
browsers
specifically,
but
would
include
browsers
and
web
flows,
especially
as
like,
let's
say,
if
you're,
on
a,
if
you're
on
a
mobile
device
that
has
an
IP
FS
protocol
handler
or
share
handler
of
some
kind
or
so
do
registers
to
be
able
to
be
a
handler
for
those
types
of
URLs.
What
those
flows
would
look
like
patterns
for
spinning
up
and
down
transient
demons
on
device.
B
What
the
you
know,
UX
and
interaction
patterns
for
when
you're
and
I
propose
client
only
say
only
using
HP
API
or
not
actually
running
a
full
node
daemon,
but
even
reading
a
you
know
lighter
later
I
profess
light,
as
we've
seen,
both
in
the
Burney
work,
a
textile
work
and
in
three
blocks
all
looking
at,
like
various
various
levels
of
what
a
node
looks
like
in
a
mobile
app,
both
in
and
out
of
rouser
context
for
those
native
apps.
Do
they
do?
B
Should
they
register
as
an
api
profess
protocol
handler
on
the
OS
like
there's
all
these
questions
that
need
to
be
figured
out
and
mobile,
that
of
which,
like
mobile
browser,
is
a
subset,
an
important
subset,
but
but
part
of
so
there's
some
and
probably
follow
along
work.
That's
going
to
happen
after
after
this
completes
might
wanted
to
give
an
update
there
and
I
think
those
were
the
only
things
that
I
had
on
the
agenda.
A
Can't
let
now
like
I
need
to
like
fake
the
same
reaction,
honest
candid
reaction.
No
I
was
like
saying
that
I'm
really
stoked
about
having
like
those
browser
designs,
especially
like
I,
really
like
how
scoped
those
are,
and
it's
probably
a
good
approach
to
like
better
produce
multiple
but
like
very
specific
scoped
documents
and
design
guidelines.
Then,
like
one
huge
butter
fork,
yeah.
B
We
want
to
make
it
as
easy
as
possible
for
per
browser
vendors
for
design
teams
and
security
teams
and
people
doing
evaluation
of
security
UI
in
browsers
to
be
able
to
understand
what
the,
how
much
of
the
browser
service
area
that
it
affects,
how
much
of
user
actions
will
be
impacted
and
where
that
matters
and
ultimately,
for
the
way
that
desktop
browser
designed
today.
That
really
is
as
nearly
scenario
as
possibly
as
closely
down
to
what
a
user
can
actually
do.
And
that's
really.
The
URL
bar
today.
B
A
I
have
something
that's
not
on
an
agenda,
but
I
thought
it's
a
fun
topic
too,
and
and
that
review
process
on
Chrome,
Web
Store
turns
out.
We
got
updated
to
2.10
in
December
and
I
was
like
tracking
how
long
it
takes
for
so
there's
a
like
an
issue
about
why
we
are
in
permanent,
in-depth
review
on
Chrome
Web
Store.
So
every
time
we
publish
we,
the
pub
new
version
is
not
immediately
available
to
our
users.
A
So
historically,
we
always
took
multiple
days,
but
for
some
time
it
looked
so
it
was
suspicious.
It
always
took
seven
days
and
it
was
the
same
I
believe
in
the
last
case.
So
I
start
to
suspect
this
simply
like
a
manual
delay,
hard-coded
somewhere,
because
every
time
it
takes
seven
days
before
a
new
version
is
published
and
it's
both
for
our
stable
channel.
This
one
is
the
stable
Channel.
We
also
have
a
better
channel
and
it's
the
same
there,
which
is
kind
of
a
bummer,
because
it
makes
it
impossible
for
people
to
try
like
quick.
A
A
A
I
believe
it
won't
improve
unless
we
create
a
dedicated
person
for
for
chromium
for
Google
Chrome.
Basically,
that
is
using
manifest
to
be
free.
But
let's
that's
like
a
separate
topic
and
it's
not
even
available
in
a
stable
channel,
there's
like
a
early
preview
canary
but
not
all
80
ice
out
there
and
it's
hard
to
tell
how
the
final
version
will
look
like.
So
time
will
tell
all
right,
I,
believe
that's
it
for
this
week
and
the
last
words
Deitrick
know
good.