►
From YouTube: GraphiQL Working Group - 2022-07-12
Description
GraphQL is a query language for APIs and a runtime for fulfilling those queries with your existing data. GraphQL provides a complete and understandable description of the data in your API, gives clients the power to ask for exactly what they need and nothing more, makes it easier to evolve APIs over time, and enables powerful developer tools. Get Started Here: https://graphql.org/
B
C
B
D
B
D
Yeah,
are
you
in
berlin.
D
B
C
That
was
great
yeah,
I
think
yeah
well,
do
we
want
to
review
anything?
Does
anyone
have
any
proposals
or
new
cool
ideas
to
talk
about.
D
Yeah
we
can
quickly
create
an
agenda
together,
feel
free
to
go
into
the
document
and
suggest
things
I
send
it
in
zoom.
Let's
agree
on
that,
and
then
we
can
dive
into
things.
A
Yep
yeah
and
I've
got
some
notes
here
specifically
about
I'm,
not
yeah
about
the
prototype,
the
point,
the
the
plan
for
validating
and
testing
and
how
we
get
people
involved.
A
A
I
haven't
heard
back
from
julian
since
you
asked
him
to
have
a
look
tim
that
was
a
few
weeks
ago,
but
prior
to
that,
he
and
I
I
tried
to
get
him
to
explain
the
alpha
colors
thing
to
me
and
I
still
don't
get
it
so
I
we
can
get
into
it.
But
my
primary
concern
is
just
about
extensibility,
like
people
writing
their
own
themes
and
how
that
how
the
alpha
value
situation
is
working.
So
I
want
to
add
that
to
the
agenda
as
well.
D
Good,
please,
I
think
that
for
the
next
one
I
should
actually
invite
julian.
We
used
to
have
him
even
in
one
a
few
months
ago.
It's
crazy
it's
already
july.
I
think
that
ricky,
we
talked,
I
think,
in
june
january,
about
potentially
having
a
new
design
one
day
and
now
we're
here
having
an
implementation,
nearly
done.
C
Yeah,
it's
exciting
yeah
how
how
much
things
have
changed
in
such
a
good
way
so
quickly,
yeah
yeah.
C
C
Obviously
so,
but
then
eventually,
we'll
have
like
I'd
love
to
help
with
docs
for
graphical
react
and,
of
course,
graphical
itself,
and
maybe
I
I
guess
what
I'm
getting
at
is
a
doc
site.
So
if
we
want
to
talk
about
a
doc
site,
I
don't
know
if
we,
if
it's
okay,
to
treat
this
working
group
as
a
monorepo
working
group.
Or
is
it
just
a
graphical
and
react
component
focus?
You
know
that's
a
question,
but.
C
B
C
The
yeah
anyways
but
yeah
it
was
yeah,
so
it
is
a
little.
It's
always.
D
Confusing
there's
a
discrepancy
between
what
the
working
group
does
or
what
we
talk
about
and
what
the
ripple
in
entails
right,
but
that's
fine.
We
can
deal
with
it.
B
C
C
Oh
weird,
the
the
most
intensive
work
right
now
is
going
on
in
graphical,
like
the
end
user,
product
of
graphical,
and
so
let's
yeah,
we'll
keep
the
working
group
focused
on
like
product
working
group,
basically
and
I'll
trail
on
about
doc's
site
stuff.
Maybe
after
you're
done
talking
about
how
about
that.
D
Okay
sounds
good,
so
then,
let's
get
started.
B
D
D
I
mean
john's
prototype,
I
would
say
john
the
expert
on
it.
How?
How
long
do
you
think
we
need
to
talk
about
it.
D
D
C
B
A
Yeah,
so
I
made
just
over
the
last
couple
of
days.
I
made
some
pretty
like
there's
no
radical
effect,
but
I
made
some
big
changes
to
the
way
that
I'm
deploying
it.
So
previously
it
was
running
on
ladle
the
just
that
root.
A
Url
was
all
the
components
and
everything-
and
I
really
it
was
just
a
sort
of
a
first
stab
at
like
getting
the
thing
on
the
screen,
and
so
people
could
play
with
it,
but
I've
since
taken
what
I've
been
calling
just
like
the
reference
implementation
of
the
prototype
and
deployed
it
just
independently.
So
I
updated
the
readme
in
the
repository
and
now
when
you
go
to
that
root,
github
pages
url,
you
just
see
the
prototype,
there's
no,
there's
no
ladle.
A
So
you
want
to
run
lately
and
and
have
the
development
environment
and
see
all
the
bits
and
the
pieces
just
get
the
repository
locally
and
run
it.
That
way.
It
obviously
makes
a
ton
of
sense
like
there's
no
reason
for
ladle
to
be
exposed
at
that
root
ul.
But
it's
just
again.
I
was
trying
to
like
just
go
super
fast.
So
I'll
put
all
that
information
onto
the
discord
channel
right
after
we're
done
here.
I
don't
know
that.
There's
anybody
following
this
thing
at
all,
so
it's
I.
C
A
Well,
I
mean,
I
think,
there's
a
few
people
that
are
that
are
looking
after
it,
but
I
don't
think
anybody's
actually
using
the
thing
right,
because
we
don't
have
a
plan
like
there's
no
plan
for
how
we
validate
it.
The
truth
is
that,
like
the
the
prototype
is
a
prototype
it,
it's
it
just
shits
the
bed
on
massive
schemas
right,
because
the
react
components
aren't
oriented
for
performance
it
just
it's
real
quick,
so
we
should
probably
find
a
way
to
formalize
the
validation
plan.
A
A
No,
no,
no,
no
just
validating
the
experience,
that's
what
I
mean
like
like.
If
we
had
a
user
experience
researcher
here,
this
is
work
that
they
would
be
doing
right.
So
it
would
be
nice
to
get
somebody
that
could
run
a
program
like
that,
but
I
think
it's
probably
a
stretch
unless
anybody
knows
somebody
that
wants
to
sort
of
get
that
get
that
work
in
their
portfolio
for
an
open
source
project.
A
But
I
guess
really,
you
know
the
cadence
that
I
usually
run
for
these
sort
of
prototypes
is
there's
an
initial
design.
We
know
that
it's
kind
of
full
of
holes.
We
know
that
we
we
know
that
it's
directionally
really
well
done,
but
it
just
needs
to
get
filled
out.
So
you
build
the
initial
version
of
a
prototype.
A
You
have
a
specific
subset
of
people
that
are
that
are
testing
against
it
right
and
that
this
could
just
be
like
15
minutes
of
playing
around
and
like
conversationally
describing
their
experience
so
that
we
can
find
some
clear
misses
that
need
to
be
done,
and
then
it
goes
back
into
design.
We
fill
the
holes,
we
update
the
code
base
and
then
we
just
sort
of
keep
that
cadence
going.
I
don't
know
how
quickly
we're
going
to
be
able
to
step
that
up.
A
If
this
everybody
on
the
call
probably
knows
if
this
were
a
private
company,
we
would
probably
be
doing
this
multiple
times
a
day,
we're
not
in
that
situation.
Right
now.
I
don't
know
that
it's
necessary
for
us
to
be
moving
at
that
sort
of
at
that
kind
of
a
clip.
C
I
honestly,
if
anyone
is
watching
these
videos,
as
they
recorded
and
uploaded,
if
your
company
would
like
to
sponsor
someone
to
help
thomas
review
work
and
introduce
you
know,
prs
and
help
with
design
and
whatever
like
please,
because
then
things
could
could
could.
Maybe
maybe
we
that
could
help
with
a
a
a
balanced
acceleration
pace
and
also
all
review
help
is,
is
welcome
by
anyone
if
people,
if
anyone
submits
a
review
of
these
pr's
like
it's
useful
feedback,
probably
so
yeah.
D
Yeah,
probably
not
too
many
people
are
watching
the
recording,
but
we
should
also
yeah.
We
can
talk
about
that
later,
how
we
can,
or
we
talk
about
it
now
how
we
can
involve
the
community
here.
I
saw,
I
think,
ricky.
You
tweeted
that
right
that
john,
that
there
was
the
prototype,
did
we
get
any
anyone's
actually
looking
into
it
or
so
far
no
responses
from
the
community.
C
Oh,
the
oh,
the
there's
lots
of
comments
on
the
prototypes
and
people
in
discussion,
but
you
know
it's
I
don't
know.
I
think
some
people
did
end
up
reviewing
some
pr's
and
have
shown
some
interest.
You
know
and
they're
looking
into
things,
but
mostly
just
like
cool
looks
nice.
You
know
like
and
that's
that's
like
that's
validation.
That's
well.
D
A
A
Yeah
so
one
one
thing
that
one
sort
of
initial
stab
at
at
just
advertising
and
getting
people
who
have
just
getting
people
sort
of
just
a
small
like
again
like
a
handful
half
dozen
people
that
are
interested
in
paying
some
level
of
frequent
attention
to
this
was.
I
was
just
going
to
go
back
to
the
tim
to
your
discussion
on
github
right,
where
a
lot
of
this
is
starting,
and
I
was
just
going
to
call
all
those
people
out
and
say:
hey
because
there
was
early
on
there
was.
A
A
That
has
like
all
kinds
of
great
features
and
ideas,
and
people
came
in
from
that,
and
so
I
was
just
gonna
just
gonna.
Add
all
those
people
in
a
response
in
that
discussion
and
and
just
say,
like
here's,
here's
the
plan.
If
anybody
is
willing
to
commit,
you
know
20
minutes
once
a
week
or
whatever
we
decide,
the
cadence
is
to
sort
of
revalidating
their
ideas.
A
That
was
my
thoughts
for
just
getting
an
initial
group
of
people
that
have
shown
interest
in
the
past
right,
not
just
interest
but
have
strong
opinions
about
a
lot
of
these
things.
I
think
that's
great,
because
I'm
not
sure
how
many
people
that
are
in
that
discussion.
That's
on
github,
are
on
the
discord
and
and
paying
any
attention.
So
that
was
just
that.
Where
I
was
that's
where
I
was
thinking,
we
might
source
interested
folks.
D
That
make
sense,
do
you
need
any
support
to
ping
dp,
the
people
john,
or
could
you
already
go
ahead
and
go
to
all
of
these
places
and
and
start
pinging
fox.
D
Awesome
there's
another
thing:
there's
a
trick
how
we
can
get
more
views
on
tweets,
which
is
posting
an
image,
so
we
should
also
post
an
image
ricky
by
the
way
I
have
a
new
laptop
and
don't
have
access
to
the
graphical
twitter
account
anymore.
So
I
will
probably
annoy
you
again
yeah.
B
C
I
need
to
get
set
up.
That's
a
just
a
quick
reminder
here.
If
there's
anyone
from
graphql
foundation
and
linux
foundation
watching
I
I
I
don't
know
who
a
graphql
foundation
is
our
contact,
I'm
pretty
sure
chris
is
now
because
I
saw
brian
warner
left,
but
yeah
that's
internal.
In
fact,
I
won't
talk
about
it
because
it's
an
internal
security
detail
and
I
won't
mention
it.
D
Yeah
yeah,
but
we
can.
We
can
coordinate
that
two
together.
B
D
We'll
coordinate
perfect.
A
Yeah
yeah,
so
that's
the
that's
the
first
bit
about
how
we
begin
to
get
feedback
for
this,
which
I
think
is
great.
So
the
next
two
bits
current
design
needs
and
the
themeability
of
the
color
palette,
I
think,
are
kind
of
the
same
thing.
A
A
There
needs
to
be
no.
No,
it's
fine.
I
totally
agree.
There
needs
to
be
a
threshold.
There
needs
to
be
a
bar
that
set,
but
I
think
if
there
is
low
hanging
fruit
in
terms
of
extensibility,
the
color
palette,
the
themes-
that's
that's
something
that
we
absolutely
should
be
able
to
support,
especially
since
this
thing
is
run
with
big
companies,
small
companies,
it
would
be
really
nice.
A
I
think,
for
a
lot
of
people
for
companies
to
adopt
this
and
to
talk
about
it
more
often
if
they
were
able
to
just
really
simply
update
the
the
color
scheme
right
and
just
sort
of
brand
it
a
little
bit.
So
I
think
I'm
gonna,
I'm
gonna,
I'm
gonna
harp
on
this
alpha
colors
thing.
A
I
think
it's
too
much
to
ask
when
you
want
to
build
your
own
theme
or
just
sort
of
change,
a
few
of
the
colors
that
are
running
in
graphical,
to
ask
people
to
understand
the
what
what
julian
was
explaining
to
me
was
a
contrast
situation,
but
it
there
may
just
be
like
a.
I
may
misunderstand
something
to
me:
it's
not
a
contrast
situation.
It's
it's
a
depth
conversation!
It's
not
about
the
contrast,
because,
as
a
engineer
as
a
web
engineer,
contrast
means
something
very
specific
to
me.
A
A
People
have
to
build
out
alpha
palettes,
because
it's
very
difficult
to
test,
like
the
actual
contrast,
the
readability,
the
usability,
the
accessibility
of
those
colors
with
alpha
palettes,
if
there's
just
16,
colors
or
20,
colors
and
they're
all
flat,
it's
a
lot
easier
to
understand
how
that's
going
to
affect
you,
know,
readability
and
usability
in
the
future.
I
could
be
totally
wrong
about
this
and
again
I
haven't
heard
back
from
julian,
so
I
need
to
I
need
to
get
in
touch
with
him.
D
A
I
don't
know
how
much
time
he's
you
know,
he's
meant
to
be
giving
towards
this,
and
I
don't
actually
also
understand
how
complete
you
guys
understand
the
current
design
to
me
or
to
be
to
me.
It's
there's
still
plenty
of
work
to
do
there
right.
We
haven't
talked
about
the
font
like
we
need
that
the
the
core
figma
file
can't
have
that
font
in
it.
It's
beautiful.
I
love
tt
commons.
I've
used
it
in
the
past,
but
it
can't
be.
A
We
need
to
change
it
right,
so
there's
so,
let's
so
that's
the
theme,
ability
thing.
That's
the
font
thing,
I'm
happy.
B
A
Do
that
work
in
figma
and
I've
already
started
doing
that
work
in
sigma,
it's
very
chaotic,
because
I'm
trying
to
do
you
know
I'm
trying
to
juggle
a
couple
of
these
things
a
few
of
these
things
at
the
same
time,
but
there's
that
part
and
then
the
other
part,
and
it's
more
specific
to
the
actual
new
ui.
The
big
important
thing
that
we're
all
trying
to
test
out
here
and
that's
that
the
situation
with
arguments.
A
If
you
just
go
play
with
the
prototype
right
now,
you
can
see
that
the
arguments
thing
doesn't
really
like
doesn't
feel
very
good
right.
The
show
optionals
hides
like
using
arguments
we
have
to
discuss.
We
have
to
find
out
whether
or
not
users
want
us
to
when
you
activate
a
field
in
the
query
builder.
A
If
we
we
want
to
automatically
add
required
arguments
to
the
query
or
if
we
want
to
simply
in
the
ui,
for
the
query
builder
flag,
that
you
know
with
an
asterisk
that
an
argument
is
required
and
then
let
the
editor
bark
at
the
user
that
they
haven't
entered
a
requirement
argument
for
a
given
query.
That's
that
would
be
my
preference,
because
I
think
the
editor
does
a
really
good
job
of
yelling
at
you.
A
When
those
things
don't
happen,
and
I
think
it's
a
very
familiar
experience
to
what
you're
going
to
get
when
you
actually
are
in
your
editor
writing
code.
So
we
need
to
discover
there's
a
lot
of
discovery
that
we
need
to
do
around
arguments.
There's
an
inline
fragment
situation
that
is
not
addressed
at
all
in
the
sigma
file,
so
we
need
to
we
need
to.
We
need
to
figure
that
out.
A
Those
are
the
two
big
ones
in
terms
of
like
current
design,
for
the
query
builder
is
just
the
arguments
and
how
we're
going
to
handle
inline
fragments.
One
idea
that
I
have
for
inline
fragments
is
that
we
pull
fragment,
building
out
into
a
plug-in
like
a
sidebar
plug-in
right,
and
so
you
can
save
those
fragments
and
you
toggle
those
fragments
separately
to
add
them
into
the
editor,
which
would
then
spread
which
would
do
like
the
ellipses
on
spread
inside
the
editor.
I
can
show
a
quick
demo
if
you
guys
want
to
see.
C
Had
solved
this
prop,
did
it
not
graphical
explorer,
didn't
solve
this
problem,
because
that's
what
there's
a
lot
of
implementations
out
there,
like
you
know
for
branded
custom
implementations,
the
graphical
with
graphical
explorers
still
so
that's
the
one
thing
also
to
think
about
with
that
is
people
are
probably
going
to
expect
at
least
a
parody
of
what
that
provides,
which
I
think
we're
almost
there
yeah,
but
it's
like
also
defining
very
like
inline
variables
in
line
or
our
arguments.
I
think.
A
C
A
C
Explorer
does
what
I
mean
by
parity,
though,
is
like
just
in
terms
of
what
graphql
features
you
can
use
with.
It
is,
I
guess,
as
a
baseline,
so
we
should
just
I
mean
we
that
doesn't
mean
that,
like
they've,
they
had
a
lot
of
time
to
implement
all
of
that
as
well.
So
we
don't
have
to
have
all
of
that
right
away,
but
it's
something
to
measure
against.
I
guess
at
least
in
terms
of
what
graphql
features
are.
A
I
think
it's
one
it's,
I
think
it's
an
opportunity
to
really
improve
that
experience
right.
I
I'm
struggling
to
understand
how
people
are
going
to
use
the
query
builder
over
just
in
the
editor,
because
the
new
editor
is
just
like.
C
It's
going
to
be
great,
but
one
thing
to
note
sorry
is
that
one
graph,
sean
at
onegraph
told
me
that
they
had
a
lot
of
customers
that
were
really
trying
to
expand
on
the
analyst
experience
so
and
analysts.
They
have
tools
now
for
visually
building.
Sql
queries
so,
like
that's
why
he
said
there
was
a
case
for
he
was
hoping
this
with
2.0
back
then
this
is
in
like
2020.
C
We
were
talking
about
this
that
that
the
graphical
explorer
could
even
like
perhaps
take
over
the
query,
editor
experience,
optionally,
that
you
could
have
a
mode
where
you
could
configure
graphical
to
work
that
way.
Obviously
it
would
not
work
that
way
by
default,
but.
C
It
would
be
more
like
it
would
be
a
implementation
detail
in
the
background,
so
in
monaco
it
would
be
like
a
model
that
would
be
written
to
and
then
and
then
you,
if
you
wanted
to
switch
to
a
source
tab
so
to
speak.
Where
you
view
the
graphql
source,
then
it
would
print
to
asc
and
lo
or
well.
I
guess
it
would
have
to
print
the
ast
already
to
to
to
store
the
query
in
the
in
the
in
the
the
monaco
model
but
yeah
but
similar
to
how
we're
doing
now.
A
Yeah
so
there's
I,
I
think.
A
I
don't
know
how
much
work
we
want
to
do
in
code
to
explore
these
ideas.
I
think
there's
a
lot
of
time
waste
that
can
happen.
If
we
do
a
lot
of
this
in
code,
we
should
probably
be
doing
a
lot
of
this
in
figma,
so
we
just
need
more
people
that
know
how
to
use
figma
to
really
like
explore
a
lot
of
these
ideas.
The
core
set,
I
believe,
is
already
defined.
I'm
happy
to
shift.
I
think
the
prototype
is
in
a
great
place.
It's
usable
right!
A
Now
you
can
plug
in
your
own
schema.
You
can
add
headers.
You
can
do
all
that
great
stuff,
I'm
happy
to
jump
over
into
figma
and
start
crafting
sort
of
like
a
more
long-term,
viable
figma
file
that
everybody
can
use
without
without
restrictions.
A
A
Up
on
the
code
prototype
because
it's
equally
as
important,
but
we
just
need
to
get
a
solid
foundation
in
for
that
sigma
file,
I'm
happy
to
start
working
on
it.
You
know
as
much
as
I
can
I'd
love
to
get.
I
don't
know
tim
again.
A
I
don't
know
the
situation
with
julian
and,
like
the
team
that
you
have,
I
don't
know
how
much
effort
you
you
want
your
team
to
put
into
working
on
the
figma
file,
I'm
sure
there's
a
community
of
designers
out
there
that
that
we
can
get
to
rally
around
this,
which
would
be
really
interesting.
The
challenge
is
that
is
so
specific
to
graphql
and
finding
designers.
That
know
all
of
this
detail
and
all
this
nuance
with
graphql
is
difficult.
So
I
don't
know
where
to.
C
C
A
C
C
Studio
used
yeah
and
everything
yeah,
so
I'm
just
looking
at
the
spacex
example
the
the
open
source
like
project.
It
has
nothing
to
do
with
actual
spacex,
but
they
the
only
difference.
I
think-
and
I
think
maybe
this
is
reflected
in
the
design.
I
can't
I
just
can't
find
where
it
is,
but
is
that
they
they
they
allow.
They
don't
deal
with
the
fragments
issue
because
at
least
from
what
I've
seen
unless
you,
I
think-
maybe
if
you
add
the
fragments
in
actual
query,
but
they
don't
they
have.
C
You
can
specify
inline
variables
statically
with
like
an
editor
in
the
explorer
mode.
So
if
there's
like
a
default
variable
value-
or
I
think
it's
both
for
default-
wait,
no!
It's
not!
For
d.
A
B
A
B
D
C
Are
we
missing
in
terms
of
if
we
were
thinking
of
graphql,
parody,
like
in
terms
at
the
level
of
graphql
features
that
you
can
in
like
capabilities
that
you
can
do
in
our
editor?
We
could
just
add
that
and
then
it
would
have
enough
parity
for
people
are
used
to
using
explorer.
E
C
D
Yeah,
that
was
an
intentional
choice
that
we
do
not
do
that
because
we
believe
that's
a
repertoire,
but
we
can.
We
can
get
into
that
later,
like
it
seems
like
people
adding
that
stuff
in
the
query
would
be
sufficient,
because
we
as.
C
Yeah,
that's
the
thing.
Well,
then,
you
match
up
the
variables
with
because
as
you're,
if
you're
thinking
again
in
a
query,
oh
in
a
query
list
composition
mode,
you're
you're
specifying
these-
you
could
then
specify
the
replacement
variables
in
this.
So
you
just
have
like
maybe
a
drop
down
a
selection,
a
selection
box
of
some
kind
with
autocomplete
that
lets
you
choose
variables
that
you've
already
declared
to
apply
to
that
argument.
D
A
It's
kind
of
like
that,
so
early
in
the
design
process
tim
and
everybody
there
was,
there
was
a
the
ui,
it's
still
in
the
sigma
file,
but
it's
it's
not
really
in,
like
sort
of
the
final
version
that
got
presented,
there's
a
ui
for
all
the
variables
just
have
inputs
and
in
an
earlier
version
I
love
it.
I
think
it's
fantastic
because
it
unlocks
so
much
potential
with
like,
like
ricky's,
saying
having
history
of
of
arguments
that
you're
passing
in
and
being
able
to
really
like
just
do
lots
of
interesting
things.
A
I
had
it
in
the
variables
editor
here.
I
think,
if
you
go
to
the
discord
channel,
there's
a
screenshot
of
like
a
past
version
where
you
could
down
here,
you
could
select
an
editor
type
right.
You
could
select
an
input
editor
or
a
code
editor,
so
I've
I've
removed
that.
But
let
me
just
do
a
real,
quick
demo
here,
so
I've
selected
this
is
the
the
rick
and
morty
api
that's
available
here
and
I'm
just
gonna.
Look
for
a
character
is
a
required
id
argument
and
you
can
see
it's
here.
A
This
is
yelling
at
me
that
I
haven't
selected
subfield,
so
let
me
just
grab
a
subfield
to
get
that
to
go
away
and
then
down
here
in
the
variables
editor.
It
doesn't
automatically
fill
it
out,
but
I
can
control
space
and
it
tells
me
that
I
need
to
fill
out
an
id
right,
the
variable
and
then
I
can
put
that
in
and
then
I
can.
I
can
run
that
query
so
this
this
this.
This
is
great.
This
is
awesome,
but
this
input.
B
A
A
A
A
If
I
wanted
to
just
do
this
all
in
the
variables
editor,
I
can
wonderful
job
ricky.
This
is
great.
This
totally
works,
but
it
doesn't
understand
the
type
right
it's
gonna
bark
at
me.
If
I,
this
is
a
string.
So
if
I
put
in
a
number,
oh
validation
actually
might
be
broken
here
for
some
reason,
but
basically
this
isn't
telling
me
that
it's
not
a
string.
It's
just
it's
taking
a
number.
I
go
here
over
to
the
you
know
easy
variables,
which
is
what
I've
been
calling.
C
B
A
This
is
something
I
haven't
fixed
this.
It
should.
B
A
Know
but
there's
a
warning
about
it,
not
doing
something
I
don't
know.
I
haven't
really
look
into
it,
but
this
here
I
can
switch
between
my
tabs
right
to
get
my
variables
and
my
different
tabs.
This
would
be
pretty
relatively
simple
to
stuff
these
variables
into
right
to
save
these
as
some
sort
of
history
in
in
the
situation
that
that
ricky
was
was
explaining
earlier.
So
I
think,
there's
like
I'm
doing.
I
think
what
I'm
doing
is
I'm
doing
a
lot
of
work
in
code
to
just
because
it's
fun.
A
This
is
really
interesting
that
we
should
probably
be
doing
figma
to
validate
a
lot
of
these
ideas.
So
this
is
just
what
I
wanted
to
show
that
there
is
opportunity
here
for
addressing
some
of
that
parody.
D
Okay,
so
let's
tie
things
back
a
bit.
We
talked
about
many
things
here.
I
just
thanks.
Everyone
sharing
the
thoughts.
What
I
just
did.
I
went
through
the
notes
here
and
marked
or
underlined
the
points
that
we
should
talk
about
a
bunch
of
things
to
talk
about
yeah
so
that
as
well,
so
maybe
yeah.
I
will
just
give
my
perspective
on
this
anything
that
is
a
bit
more.
Let's
say:
design
specific,
for
example,
using
alpha,
colors
or
not
or
obviously
not
using
tt
comments.
D
Td
comments
was
just
because
of
the
design
system
that
we
had
there,
but
we
won't.
We
will
just
use
something
similar
that
is
open
like
intel.
That
discussion
should
happen
with
julian.
It's
quite
simple.
Julian
definitely
can
work
on
this
and
we
want
it's
always
with
any
project.
The
last
20
are
the
tough,
the
tough
part,
and
it
seems
like
we
got
a
lot
of
first
80
percent
done
with
that
graphical,
and
it
would
be
a
travesty
if
we
would
not
finish
the
last
party
and
of
of
course
we
want
that.
D
Oh
yeah,
and
I
can
tell
you
john,
that
definitely
julian.
We
are
happy
to
finance
that
and
and
let
julian
so
we
are
basically
paying
for
that
workers
late.
We
are
happy
to
do
that
and
what
I
suggest
is
that
it
will
probably
not
be
very
productive
to
dive
into
these
discussions.
Now.
D
D
Yeah,
so
I
think
that,
still
for
such
a
fairly
small
project,
having
already
a
few
people,
few
designers
from
the
ecosystem
and
so
on,
would
not
be
helpful.
Julian
has
enough
work
enough
time.
I
just
need
to
ping
him
and
he
might
not
have
looked
into
it
sometimes
need
to
ping
him
multiple
times
he
is
still
busy
for
sure,
but
we
can
make
that
happen
in
a
reasonable
amount
of
time.
C
D
Anything
I
think
that
was
like
the
alpha
palettes.
The
fonts
were
about
that.
Also
what
you
said
to
john.
I
really
liked
of
using
figma
as
a
prototyping
tool.
Instead
of
directly
implementing
everything
makes
total
sense.
I
think
those
would
be
perfect
agenda
items
and
I
think
waiting
another
month
for
the
next
working
group
is
a
bit
long
happy
to
organize
something
here,
that
we
can
talk
next
week
or
something
great
yeah,
and
then
we
have
a
bunch
of
topics
that
are
not
related
to
design
but
general
functionality.
D
Topics,
as
we
just
said,
it's
not
clear,
as
you
said,
not
clear
how
to
handle
inline
fragments
that
is,
in
the
sense,
also
design
related
because
it's
ux.
D
So
we
can
also
take
that
topic
to
next
call
with
julian
and
just
bring
it
up.
Maybe
john,
if
you
could
just
sketch
out
options
that
you
see
until
then
for
dealing
with
inline
fragments
and
if
the
number
of
options
that
you
can
come
up
with
is
zero.
That's
also
totally
fine
and
we
can
just
brainstorm
around
that.
A
A
People
either
people
either
use
fragments
or
they
or
they
don't
right
or
they
use
fragments,
and
they
in
graphical
currently
right.
They
they
they
do
them
in
the
editor,
and
then
they
express
those
fragments
inside
of
their
queries,
things
get
a
little
wonky
given
like
if
people
are
using
playground
right
now.
Maybe
graphical
is
the
same
way,
but
if
you've
got
like
multiple
operations
inside
the
editor,
you
get
that
when
you
hit
play
you
get
that
drop
down.
C
B
A
So
people
are
using
fragments
like
in
in
just
what,
like,
whatever
way,
is
available
now
right,
because
the
way
that's
available
now
is
what's
in
the
editor,
which
is
what
you
would
expect
from
a
traditional
editing
experience
moving
that
into
the
ui
is
a
is
a
is
a
different
thing.
It's
a
challenge
right.
A
We
want
to
make
sure
that
we
cover
it
and
and
and,
as
you
said,
tim
I'll
I'll
explain
with
some
screenshots
what
I've
done
currently
to
try
to
do
that
and
and
as
far
as
suggesting
alternatives,
I
think
everyone's
just
gonna
we're
gonna
need
to
hear
from
the
community
about
what
they
might
like
to
see.
D
In
the
editor,
and
also
I
wanna
make
a
quick
note
here
that.
B
D
We
have
these
more
tricky
topics,
then
the
overall
graphql
working
group
is
always
welcoming
us
to
carry
these
things
over
there
and
get
their
view
because
they
are
the
graphical
community
in
a
sense
they're
for
sure
not
all
of
them,
but
I
know
that
they
mentioned
the
fragments
topic
and
there
was
one
I
think
benji
brought
it
up.
Oh
surprise
with
our
idea
here
with
the
design
fragments
would
be
tricky,
so
that
was
no.
That
would
be
tricky.
We
just
didn't
think
about
the
proper
solution.
Yet.
A
D
Yeah,
so
that's
always
something
we
can
do,
but
let's
first
create
a
proposal
internally
or
maybe
whatever
options
we
have
and
then
we
can
still
present
that
to
a
broader
audience.
Before
we
go
deeper
into
implementations
there
yeah.
A
D
C
It's
a
prop
for
graphical
that
allows
you
to
specify
fragments
that
are
available
that
get
automatically
appended
to
the
query
on
execution
that
aren't
visible
in
the
query
editor.
This
is.
D
C
So
so
you
would
be
able
to
it's
a
it
can
be
a
string
or
a
array
of
fragment
definition
nodes,
and
so
then
you
would
be
able
to
take
this
and
be
able
to
have
a
pre-supplied
set
of
external
fragments
that
could
be
potentially
inserted
in
different
places
for
autocomplete
or
for
whatever
purpose.
So
that's
I
know
there's
several
companies
using
that
already
that
found
that
to
be
a
suitable
way
to
kind
of
bring
in
their
own
domain
logic,
it's
a
little.
C
It
was
it
was.
That
was
a
little
controversial
though,
as
you
might
imagine,
because
you
know
we
like
fragments
to
usually
in
the
graphical
community,
we
like
queries
to
be
fully
described.
Just
descriptive,
it's
a
it's,
a
client-driven
domain
language.
So
you
you,
you
want
to
see
everything
there.
You
don't
want
magic,
fragments
as
gatsby
calls
them,
but
it
helped
with
that
case.
So
that
might
be
one
option
to
get.
You
could
get
it
from
there.
C
A
C
Yeah
and
with
monaco,
we
can
do
the
definition.
Look
up
for
the
fragment
type,
so
thing
too.
D
B
D
I
know
I
think
we
have
some
actionable
steps
here.
I
will
make
sure
that
we
schedule
a
meeting
for
including
julian,
because
he
is
currently
the
one
who
came
up
with
the
design
and
we
should
just
talk
with
them.
I
think
that
would
be
most
productive,
anything
design
related,
and
I
think
that
is
also
a
good
opportunity
to
bring
up
the
challenges
around
fragments
because
julian
as
a
designer,
we
sometimes
need
to
give
him.
D
Let's
say
some
time
to
like
ramp
up
on
these
topics
as
he's
not
a
programmer
great
and
then
john,
you
would
go
into
all
the
issues.
Actually,
let's
have
some
action
items
written
down
here
so.
D
Items
so
tim
create
car,
john,
let's
say:
notify
community
about
new
changes.
D
Another
thing
what
I
wanted.
What
did
I
want
to
say
yeah,
please.
D
Exactly
what
we
could
do
on
the
on
the
fragment,
fragments
and
what
you
already
did,
I
would
also,
I
would
say,
tim
ricky,
coordinate
twitter.
I
also
wanted
to
do
another
tweet,
I
think,
on
my
old
laptop.
I
could
actually
do
a
tweet
here,
just
after
the
call
with
the
design
so
yeah
and
to
encourage
people.
I
know
that
at
the
graphql
in
austin
there
were
some
people
like
the
joey
from
paypal
who
was
like.
When
can
I
use
it
and
those
people?
D
We
should
go
to
now
and
say:
hey,
you
can
use
it
and
by
the
way,
yeah.
If
we
literally
get
no
response,
we
can
literally
just
put
a
five
minute
thing
into
the
graphql
working
group,
so
everyone
sees
it
and
like
hey,
please
try
it
out
with
your
schema.
A
If
you
guys
got
you
know,
your
presentation
was
great.
I
can
see
how
it
would
have
like
roused
people
that
were
there
into
wanting
to
get
involved.
A
A
little
pressure
on
them
to
get
involved
with
sort
of
reigniting
a
lot
of
the
momentum
that
existed
earlier
this
year,
but
seemed
to
like
fall
off
since
then.
B
D
Yes,
unfortunately,
not
enough
yeah,
true.
D
There
last
since
last
call,
but
I
will
make
sure
that's
an
extra
item
for
me.
Okay,
anything
else
on
your
prototype,
john,
that
you
would
like
to
talk
about,
or
do
you
feel
good
with
this.
D
Good
this
is
important.
You
have
actually
worked
a
lot
on
this,
so
that
is
I'm
happy
that
we
talk
about
it
so
yeah,
okay,
I
will
put
the
action
items
to
the
bottom,
so
yeah
good.
So
then,
let's
talk
about
what's
happening
today
and
thomas
who
is
doing
a
lot
of
implementation
work.
Maybe
thomas
you
can
give
us
a
quick
overview
where
we're
at
and
what's
basically
left
before.
We
can
do
the
big
merge
and
the
big
graphical
to
release.
E
The
status
quo
is
basically
that
we
said,
like
we
looked
into
doing
actual
pre-releases,
with
change,
set,
ditch
that
because
that
would
mean
we
have
to
do
it
for
everything
in
the
monorepo,
which
would
like
block
a
language
server
and
this
code.
E
So
we
basically
have
again
a
branch
where
we
collect
all
the
changes
that
have
been
reviewed,
which
I
keep
rebased
regularly
on
the
main
branch
so
that
we
don't
run
into
huge,
merge
conflicts
in
the
end
and
yeah.
It's
basically
me
opening
the
the
pr
status
that,
oh,
that,
with
the
changes
that
I
have
still
like
patched
up,
there
are
some
things
that
are
still
missing.
I
didn't
implement
the
the
search,
yet
I
guess
in
the
in
the
doc
explorer.
E
So
I
need
to
like
the
design
of
the
search.
So
I
need
to
move
that
over.
I
need
functionality.
It
will
change
for
v2
until
we
get
the
new
docker
explorer.
What.
C
A
E
Yeah
like
we,
we
will
keep
the
doc
sidebar
as
it
is
for
now
and
just
like
move
it
over
to
match
the
general
design
theme,
but
not
change
functions.
A
Thomas,
does
that
mean
that
you're
gonna
take
the
in
the
figma
file?
There's
a
ui
for
the
search,
that's
currently
in
place
for
like
the
infigma
file,
what
I've
been
calling
pathfinder
you're
gonna,
take
that
and
you're
going
to
use
that
for
the
traditional
or
familiar
box
explorer
view
that
you've
got
in
the
that's
going
out
for
for
v2.
E
There
is
a
also
intermediate
a
design
for,
like
it's
called,
I
think,
intermediate
stock
docs,
where
there's
there's
also
a
little
design
for
the
search,
I'm
not
sure
right
now
how
this,
if
this
differs
from
the
search
in
the
new
dock
explorer
or
not,
but
it
should
be
it's
somewhere
like
on
the
bottom.
I
think
so
like
hard
to
find.
I
would
need
to
look
it
up
now.
I
don't
think
that's.
E
Regarding
graphical
reacts,
I'd
keep
this
as
like
still
version
zero
dot
whatever,
as
it's
probably
not
gonna,
be
really
stable.
When
we
release
graphical
v2,
I
think
there
will
be
like
quite
some.
There
might
be
quite
some
changes
that
we
would
like
to
do
so
I'd
not
push
that
to
version
one
also.
We
have
like
going
ahead
a
bit.
We
don't
have
any
docs
or
whatever
yet
for
it.
So
it's
probably.
C
E
I
can
talk
about
that
in
the
next
next
thing
and
I
think
that's
pretty
much
it's
like
for
sure.
There
will
be
some
some
knits
that
that
we
need
to
do
to
really
finish
it
off
in
the
end
before
we
can
do
the
release,
but
that's
still
in
the
unknown
area,
at
least
from
my
side,
but
I
don't
expect
anything
that
would
take
up
a
week
to
to
fix.
So
it's
probably
gonna
be
lots
of
smaller
click,
fixed
issues.
D
Okay,
so
it
sounds
like
within
the
next
like
two
weeks
or
something
or
I
don't
know
how
long
it
takes
for
you
to
implement
the
search
or
whatever
is
missing.
Then
we
should
be
able
to
get
like
a
release
candidate
out
or
something
so
did
I
correctly
understand
that
we
kind
of
in
the
current
setup,
with
change
set.
We
kind
of
would
need
to
decide
that
we
moved
the
whole
ripple
and
all
over
to
release
candidate
mode
how's
that,
yes,
exactly.
E
C
E
C
Actually,
I
think
this
is
actually
possible.
I
think
I
think
it's
it's
possible
that
we
could
publish
from
the
feature
branch
similarly
to
how
we
do
with
canaries,
but
all
we
have
to
do
is
slightly
change
the
the
github
workflow
to
work
for
feature
branch.
So
we
could
say:
oh
if
it's
a
feature
branch,
then
you
can
publish
alphas
and
as
long
as
we
make
sure
to
overwrite
it
so
that
it's
a
pending
alpha
instead
of
canary
and
in
fact,
with
the
way
it
is
now.
E
I
mean
you
wouldn't
need
previews
at
all.
If
you
have
a
custom,
guitar
action
that
just
publishes
alpha
versions
yeah,
it
should
work
as
well.
C
E
I
mean
question:
is
how
much
value
will
it
bring
us?
I
mean
it
really
only
makes
sense
when
we
like,
when
emotional
changes
in,
and
it's
like,
really
usable
to
have
beta
versions
that
people
could
try
out
yeah
and
then
we
are
in
the
same
spot,
where
we
would
first
need
to
like
find
people
who
really
tried
out
to
like
get.
C
C
E
How
fast
and
how
many
people
we
can
get
if
it's
already
difficult
for
like
right
and
the
prototype,
would
probably
be
exactly
a
spot
here.
E
Two
months,
yeah
right
exactly
I'd
like
to
then
merge
the
branch
where
we
collect
all
the
changes,
rather
sooner
than
later,.
C
Yeah,
you
could
basically
use
the
better.
The
better
release
is
like
a
integration
testing,
basically
saying.
Okay,
it's
pretty
much
ready,
like
maybe
release
candidate,
is
even
a
better
fit
for
that
and
say:
can
people
try
it
and
make
sure
it
works
and
like
we
can
work
out
some
bugs,
so
we
don't
have
to
do
a
bunch
of
patch
releases
after
we
do
a
new
major
version
release,
but
yeah
or
but
yeah.
I
think
that
that
that's
overall
yeah,
we
don't
need
to
do
this
until
very
at
the
very
end
stage.
E
B
D
D
C
C
C
E
D
But
then,
which
sounds
like
the
finish
line
is
inside
that's
exciting,
a
lot
of
good
work
and
also
yeah
from
everyone.
Also
a
lot
of
work
from
thomas
here.
Actual
implementation
work
great.
So
it
sounds
like
we're
in
a
pretty
good
state,
so
we
say
we
would
like
to
get
it
out
on
alpha.
Even
if
it's
a
one-time
manual
release
that
then
can
be
spread
across
the
committee
and
again
we
could
do
a
call
out
to
the
community
twitter,
discord,
etc
and
say:
look,
that's
the
alpha.
Please
use
it.
E
D
E
Like
this
is
the
next
version
expect
like
minor,
breaking
changes
to
come
in
that's
pretty
much
stable
for
the
most
part.
C
D
And
just
quick
question
an
rc:
what
that
land
on
latest
or
how's
that
usually
with
rc's.
D
So
I
just
checked
react,
for
example,
has
an
actual
rc
tag,
which
is
also
an
option
right.
We
could
introduce
that,
but
it
seems
like
alpha.
Beta
rc
is
a
bit
overkill.
We,
I
don't
yeah.
D
E
D
A
If
that
even
helps
everybo
everybody's
going
to
have
their
own
opinion
right,
like
everyone's
history,
is
going
to
inform
whether
it's
called
a
prototype
or
a
proof
of
concept,
I
think
I
mean
technically
literally
it's
closer
to
a
prototype
than
it
is
a
proof
of
concept
because
we're
not
radically
changing
the
core
experience
right.
The
core
experience.
The
editor
is
staying
the
same.
There's
just
this
one
piece,
the
query
builder,
that's
the
focus
of
the
prototype.
I
explained
this
all
in
the
new
readme
that
I'm
written
for
that
I've
written
for
my
repository.
A
E
E
Could
help
would
be,
I
think,
it's
in
the
preview
url,
for
example,
there's
still
v2
in
there
just
to
take
the
v2
out
to
make.
A
E
C
E
There's
more
a
bit
more
things
like
we
don't
have
gas
components
anymore,
so
you
can't
pass
a
ref
and
access
like
that.
You
need
to
like
use.
C
E
Some
of
our
hooks
to
to
access
the
state-
that's
known
providers
and
stuff
like
that.
C
I've
had
some
ideas
for
prop
for
props
changes
too.
While
we
get
a
chance
at
a
breaking
version.
For
example,
we
have
like
a
lot
of
these
boolean
toggle
like
flags.
It's
a
pattern.
I
continued
for
a
while,
but
it's
just
it's
just
messy
and
there's
a
lot
of
those
I
think
could
be
like
enum
type
values
or
like
represented
in
a
new,
a
different
type
of
configuration
so
like,
like
editor,
editor
toggle,
editor
header,
editor
visible
by
default
like
these
kinds
of
like
laborious
booleans
that
we
could
maybe
imagine
in
different
props.
C
C
D
Awesome
anything
else
on
this
seems
like
we
have
a
good
path
now,
which
was
the
point
path
to
graphical
two
by
the
way
in
graphql
rockcliff
in
poland,
a
attendee
of
the
conference
was
very
meetup
was
very,
let's
say,
eager
to
get
a
monaco
release
as
well
or,
let's
say,
a
new
graphical
which
uses
more
necro,
and
I
was
saying
yeah
easy
we're
getting
the
new
design
out.
That's
a
lot
of
work
and
then
step
by
step.
D
But
the
community
is
already
excited
about
that
as
well.
A
Yeah
there's
also
a
person
thomas
and
ricky,
on
the
on
that
that
issue
that
blind
developer.
This
is
such
a
wonderful
opportunity
to
get
somebody
who
requires
accessibility
to
the
extreme
involved
early
in
the
process.
It
was
really
nice
to
hear
that
monaco
just
kind
of
worked
for
them
right.
It's
just
a.
C
Yeah
we
have
that
we've
had
that
validation
before
with
the
early
monaco
demos.
A
couple
years
ago
we
had
the
same
thing
where
a
user
came
into
discord
and
or
a
github
issue
and
asked
like.
Oh,
I
can't
use
this
and
I
just
shared
what
what
about
this
monaco
graphql
demo-
and
they
said.
Oh
my
gosh.
This
is
so
much
better
yeah
and,
as
we
know,
vs
code
accessibility
is
premium
and
they
they
really
put
a
lot
into
that.
C
But
I
also
want
to
give
a
shout
out
to
komir6
and
the
maintainer
code
mirror,
whoever
it
is
who
I
was
rude
to
once
I
apologize
he's
doing
a
great
job
and
well.
The
code
mirror
community
is
doing
a
great
job
of
of
taking
accessibility
into
account
for
six,
but
we'll
we'll
get
there
when
we
get
there.
It's
like,
I
think
it
can
be.
Our
main
target
for
3.0
is
a
fully
accessible
graphical.
I
think
that's
a
yeah,
do
we?
C
A
A
C
Yeah
yeah,
to
be
clear,
I
I
yeah.
I
do
see
that
it's
a
lot
more
than
just
monaco's
required
to
make
graphical
accessible,
but
but
you
can't,
we
can't
really
make
with
with
code
mirror
5.
Even
if
we
do
everything
we
can
to
make
the
rest
of
the
the
crow
accessible
code
mirror.
5
won't
be
just
just
for
I'm
again
speaking
to
the
audience
again.
Sorry
but
yeah.
D
Yeah,
that's
great,
I
mean
it
seems
like
with
monaco.
A
lot
would
be
covered
out
of
the
box
and
we
wouldn't
need
to
worry
about
it.
That's
great
awesome!
So
then
let's
wrap
that
point
up
and
go
to
the
last
point
that
we
have
today
here.
Let's
see
if
we.
A
Will
indeed,
oh
yeah
one
thing
really
quick
thomas,
I
haven't
done
any
work
on
the
current
repository,
but
I'm
available
if
you
need
help,
as
things
are
beginning
to
ramp
up
towards
a
release,
and
you
can
chunk
small
bits
of
work
up
that
that
that
you
can
pass
off
to
me
without
needing
to
have
the
full
history
I'm
happy
to
to
jump
in
and
help
out.
I
just
right
now.
A
E
C
C
Get
so
eddie
yeah
the
more
help
of
here
yeah,
it's
yeah,
it's
like
trying
to
learn
a
different
language.
You
know.
A
I'm
not
happy
to
happy
to
look
at
them
and
it's
probably
the
best
way
for
me
to
get
familiar
with
the
existing
code
base,
which.
E
I
haven't
I
mean
so
we'll
definitely
have
some
some
things
to
fix
then,
like
I
have
a
lot
of
things
still
piled
up
that
I
just
need
to
like
pick
out
into
reviewable
prs,
but
then,
in
the
end,
I
think
we'll
have
for
sure
some
things
that
we
can
fix
in
parallel,
where
we
can
split
things
up.
C
I
have
a
question.
I
have
a
question
for
time
for
a
question.
Yes,
it
has
to
do
with
graphical.
Do
we
plan
on
having
no
css
files
at
one
point
and
using
css
and
js
entirely?
What's
the
plan
there,
just
scary?
Oh
just
a.
D
Little
question.
E
I
think
for
me,
I'd
keep
css
files
around
for
version
two,
just
because
it's
the
most
straightforward
way
to
get
things
done.
Also
for
people
who
want
to
contribute
it's,
they
don't
have
to
learn
any
css
and
js
framework
that
we
decide
to
use.
E
But,
like
I'm
open
for
discussing
this
further
in
the
future,
okay,
I
mean
that's
great,
since
we,
since
we
now
got
theming
the
like
the
use
case
of
overriding
classes,
is,
should
officially
be
discouraged.
E
That
I'd
include
in
the
like
migration
guide,
release,
notes
whatever
for
version
two,
that
if
people
want
to
change
the
appearance,
then
theming
is
the
way
to
go
and
not
over
or
like
overwriting
class
names
can
still
do
it.
If
you
want,
but
we
won't,
we
won't
consider
changes
to
class
names
or
striking
anymore.
A
Yeah
there
should
be
a
easily
understandable,
documented
way
for
theming
like
there
should
be
almost
zero
ambiguity
for
how
to
change
the
branding,
icon
or
color
of
this
thing
or
the
color
of
that
thing.
It's
20
22..
This
should
be.
This
is
absolutely
low-hanging.
A
I
do
want
to
say
that
while
css
and
js
can
help
us
here
immensely
in
the
react
world,
this
is
changing.
This
seems
like
it's
changing
every
week
right.
The
core
react
team
is
now
pushing
back
pretty
heavily
on
css
and
javascript
solutions.
It
doesn't
look
pretty
for
anyone
who's
using
a
library
and
wants
to
you
know
it's
not
happening
now.
A
Yeah,
there's
a
it's
not
happening
right
now,
but
it's
going
to
happen
soon
right
this
whole
vanilla,
js,
streaming
css
react
is
the
react.
Team
isn't
super
into
it,
so
it
it's
good
that
we're
not
making
this
decision
right
now
about
whether
to
move
away
from
css
and
into
css
and
js.
But
it's
something
to
keep
our
eye
on
and
to
you
know,
front-end
engineers
keep
their
eyes
on
these
things
anyway.
So
it's
gonna
have
a
big
impact
on
this
decision
in
the
future.
C
Agreed
and
let's
just
make
really
exciting,
clever
use
of
css
variables,
even
maybe
for
like
spacing
and
typography
as
well
as
kind
of
so
we
yeah,
okay,
great
yeah.
It
looks
I
I'm
sorry,
that's
how
you
can
tell
how
little
I've
reviewed,
but
I'm
just
like.
Let's
make
sure
to
do
this
and
you're
like
I
did
that
already
but
yeah
like
because
we're
doing
that
we
unlock
all
those
features
so
that
they
don't
like
you're
saying
they
don't
have
to
worry
about
class
anymore
yeah.
D
D
Okay,
incredibly
short
discussion
on
such
a
big
topic.
It
is
incredible
how
much
we
are
aligned
great
then,
let's
use
the
rest
of
the
time
we
have
here
to
talk
about
dox
photo
site.
C
This
is
a
very
fast
meeting,
yeah,
so
doc,
site
yeah,
I'm
just
looking
into
a
few
options.
I
plan
on
introducing
a
doc
site
for
other
parts
of
the
ecosystem,
especially
because
right
now,
obviously,
with
graphical
things
are
in
flux.
Maybe
we
could
just
do
just
like
kind
of
move.
What's
there
into
a
static,
1.0
doc
site
that
we
can
eventually
like
add
to
so
that
we
have
like
the
two
points.
C
I
don't
know
how
we'll
do
the
versioning
of
docs,
but
I
see
like
docusaurus
other
tools
have
have
some
stuff
for
this.
C
I'm
using
like
the
guild
and
some
other
open
source
folks
is
like
a
model
of
how
they
do
doc
sites
and
whatnot
with
you
know,
type
doc
to
markdown,
and
you
know
you
know
like
the
kind
of
structure
and
things
and
getting
inspiration
from
them
and
others
so
I'd
like
to
have,
especially
for
you
know,
it's
been
a
long
time
missing,
like
you
notice,
just
the
readmes
just
keep
getting
longer
and
longer
with
graphical
tool
kit
I
finally
daringly
broke
away
and
said
you
know
what
there
will
be
a
markdown.
C
That's
not
a
read
with
docs.
In
it,
and-
and
I
think
we
can
start
doing
more-
as
you
can
tell
I've
like
kind
of
tried
to
weirdly-
stick
with
very
early
conventions
in
the
merged
borrowed
repos-
that,
like
the
resources,
folder
and
all
this
stuff
that
are
all
holdovers
from
the
old
facebook
repo
and
just
please
just
feel
free
to
just
introduce
any
new
structures
and
conventions
and
whatever
anything
that
you
ever
want
to
do
just
for
the
record
yeah
but
yeah
for
the
doc
site
yeah.
C
I
think
it
would
help
users
a
lot
to
have
just
a
simple
doc
site
that
that
we
can
use
and
brian
warner
formerly
of
graphql
foundation
linux
foundation.
I
think
he
moved
on
to
an
another
path,
but
he
had
set
up
a
graphical.dev
domain.
For
us,
I
think,
is
what
it
was
and
we
have
other
options
as
well.
We
could
do
like
graphical.graphql.org
there's
a
lot
of
options
for
that.
C
I
feel
like
the
domain
is
one
of
the
last
things
you
figure
out
with
something
like
this,
but
like
yeah,
we'll
figure
out
something,
but
something
other
than
graphical
test.net
tests.netlify.app,
which
is
what
we
currently
use,
because
someone
has
reserved
the
graphical
namespace
on
netlify
and
I
haven't
bothered
to
just
like
email
netlify
which,
because
it's
probably
a
totally
deactivated
site
from
like
10
years
ago,
yeah
just
waiting
someone's
got
it
camped
out.
Someone
took
the
gra,
the
graphql
namespace
and
docker
hub
and,
if
you're
watching
I'm
coming
for
you.
A
So
ricky,
how
do
you,
how
do
you
see?
How
do
you
see
maintenance
of
these
doc
sites
going
forward?
How
do
you
see
that
maintenance
fitting
into.
C
A
Of
chaotic
right,
I
understand
that
the
dark
side
would
help.
It
would
potentially
help
it
get
less
chaotic
because,
instead
of
editing,
readmes
we're
editing
something
else,
but
also
I
love
what
the
guild
does
is
wonderful,
but
I
use
their
tools
and
I
use
a
lot
of
their
ecosystem
pretty
frequently
and.
A
I'm
just
gonna
introducing
a
dark
site.
C
If
it
makes
any
if
it
makes
yeah,
I
think
the
way
I
see
it
is
it
would
be
basically
taking
what
we
have
in
mark
in
readme's
already
and
putting
it
in
into
a
doc
site
format
and
then
leaning
as
much
as
possible
on
type
doc
right.
We
shouldn't
be
that's
why
I
eventually
moved
in
the
graphical
readme.
If
you
notice,
I
eventually
just
linked
to
the
type
doc
for
the
props
and
what
I'd
rather
do
is
link
to
the
type
doc
in
markdown
generators,
part
of
a
docusource
site.
C
C
Does
this
too,
where
you
can
link
to
specific
types
in
the
type
dock
so
like
in
like
a
kind
of
magic,
ish
way
from
your
regular
readme
or
from
your
markdown
or
or
mdx
so
yeah,
and
I
think
that
would
be
good
and
we
could
just
add
a
couple
of
quick
examples
and
whatnot
for
people
look
at
and
what
I'd
like
to
do
is
just
have
like
an
examples
gallery
in
line
with
mdx.
You
know
that
I
that
I
can
look
at
when
doing
deploys
too.
C
You
know,
so
we
can
look
at
our
deploy
preview.
Look
at
the
docs
site.
Look
at
the
examples
gallery
that
runs
through
different
cases
of
different
configurations
of
props,
and
then
you
can
kind
of
do
a
quick
scan
for
a
manual
regression
check.
You
know
beyond,
like
the
end
to
end
suite.
You
know,
you're
talking
about.
C
Well,
I'm
just
saying
it's
an
example
but
yeah
for
the
associated
packages
we
needed
as
well
yeah,
so
there
would
be
different
sections
of
it.
I'm
trying
to
figure
out
whether
they
would
each
have
like
their
own,
like.
I
think
we
could
do
a
separate
one
for
a
graphql
lsp
and
the
vs
code
at
one
and
then
another
one
just
for
graphical.
Maybe
it's
separate
domains
and
still
deploy
it
from
the
same
repo
and.
C
A
Yeah,
I
was
thinking
that,
given
the
you
know,
given
the
sort
of
more,
I
don't
want
to
sound
like
an
but
given
like
the
sort
of
the
more
modern
approach
to
designing
the
the
new
version
of
graphical
like
it's
designed.
First,
that
has
you
know
some
kind
of
a
design
library,
that's
powering
it.
It
lends
itself
really
well
to
what
you
think
of
as
like
a
classic
kind
of
like
design
system
library
like
go,
look
at
what
purcell
does
for
their
component
library
right
where
you
can
explain,
it's
not
storybook.
A
It's
a
lot
simpler
than
storybook.
It
explains
the
props
that
you
need.
You
can
have
examples.
You
can
have
running
code
bits
all
that
stuff.
It's
really
well
supported
just
by
the
process
that
is
being
followed
right
now,
where
we're
starting
in
figla
and
moving
over
to
to
code.
We
just,
I
think
we
just
I.
I
always
saw
it
being
more
simple
to
document
how
to
use
graphical
what
the
prop
types
are.
What
the
shape
of
the
data
that
can
go
in
is
what
the
options
are
like.
A
A
It
can
live
in
the
same
dock
world,
but
just
the
representation
is
a
little
bit
different
like
it's
less
again.
I
don't
want
to
sound
like
an,
but
it
gets
less
dry.
It's
more
interactive
right,
like
the
graphical.
The
graphical
documentation
could
be
more
interactive.
Like
change
this
prop
here's
what
it's.
C
B
C
A
C
A
playground
where
you
could
do
like
the
I.
I
call
that,
like
the
old,
the
old
jquery
gallery,
the
old
jquery
plug-in
gallery,
where
you
have
you
can
toggle
the
features
and
look.
The
funny
thing
is
as
advanced
as
we
are
now.
There's
like
some
basic
service
level,
concepts
that
apply
to
the
old
monolithic,
jquery
plug-ins,
that
kind
of
apply
to
graphical,
because
it's
like
a
whole
end
user
interface,
that
you're
rendering
with
one
kind
of
component.
So-
and
I
don't
know
anyways.
This
is
just
a
funny
joke.
D
The
way
demo
please
ricky.
C
Yeah,
all
I
was
going
to
say
is
so
finding
from
this
is.
I
will
treat
the
lsp
server
and
maybe
also
code
mirror
graphical
monaco
graphical
api
docs,
as
maybe
a
separate
thing
than
the
graphical
docs,
and
just
be
like
here's,
the
the
editor
ecosystem.
These
are
the
more
advanced
stocks
or
docs
that
are
for
ide
users,
and
then
a
separate
site
would
be
for
graphical
in
the
graphical
sdk
and
things
under
the
graphical
name,
space
yeah,
so
cool,
perfect.
D
And
by
the
way
we
could
also
or
in
terms
of
domains,
just
use
github
pages
right.
It
would
also
be
fine.
C
D
There's
one
thing
I
want
to
show
you:
you
probably
don't
know
about
this,
because
this
is
still
quite
new,
but
johannes
chickling
actually
is
building
a
tool
called
packer.
That
auto
generates
a
pretty
nice
docs
page
based
on
typescript
types,
and
this
is
pretty
nice.
Actually,
it
does
not
contain
what
we
talked
about
in
terms
of
like
examples
for
components,
but
if
you
have
a
more
like
typescript
api,
this
is
already
pretty
strong,
there's
nothing
to
do
for
us
here.
So
if
we
say
that
graphical
react,
we
want
docs
for
that.
D
D
I
need
to
ask
johannes
if
we
can
even
use
this
he's
still
working
on
it.
D
The
the
thing
is,
it's
completely
auto
generated
and
there's
nothing
because
we
talked
about
okay,
how
to
keep
things
up
to
date.
We
if
we
stick
to
less
documentation
but
having
all
the
documentation
based
on
code.
This
could
be
a
pretty
good
starting
point.
I
will
ask
johannes
how
much
he's
comfortable
with
us
using
this.
It's
not
yet
doesn't
even
have
a
landing
page
yeah.
E
Can
you
add
custom
like
learning
resources
on
the
tab
on
the
top
left,
yeah.
D
Yeah,
just
as
a
quick
thing
I
will
I
will
talk
to
johannes
about
that-
might
be
a
good
start,
at
least
for
the
graphical
react
package
to
have
something.
D
Okay,
cool,
then,
if
we
don't
have
anything
else,
we
can
leave
it
at
that
for
this
time,
and
I
will
follow
up
with
the
next
call
then
with
julian,
so
we
can
actually
talk
about
design.
John
and
yeah
awesome
anything
else.