►
From YouTube: GraphiQL Working Group - 2022-11-08
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/
C
B
A
B
A
But
yeah,
so
what
did
you
have
to
talk
about
today
because
I'm
sure
you
have
a
lot
of
interesting
things
to
talk
about
I.
B
Have
a
I
have
a
couple
of
things:
the
something
I've
been
working
on
this
whole
week,
just
kind
of
off
and
on
is
replacing
context
with
the
framework.
Agnostics
store
right
with
the
stand,
which
is
something
that
we
talked
about
for
yeah
I,
can't.
B
A
It's
yeah
I
had
it
earlier:
okay,
yeah,
so
I've
been
that
doesn't
matter,
yeah
yeah.
B
So
I
picked
that
Library,
because
it
just
has
a
vanilla
store
that
that
you
can
light
up
with
State
inactions
and
then
in
anything,
that's
react-based.
You
can
either
use
that
vanilla
store
and
the
methods
that
come
off
of
it
for
getting
and
setting
and
subscribing
to
state
or
you
can.
You
can
wrap
it
in
a
stand,
react
hook
and
you
can
and
it's
just
easier
to
access
it.
B
I,
don't
know
a
whole
lot
about
its
use
in
View
and
angular
like
there's,
not
a
lot
of
there's,
not
a
lot
of
examples
of
people
actually
using
it
in
View
and
angular,
but
I
don't
know,
I,
don't
know
why
it
couldn't
work
like
it's
meant
to
work.
It
should
work
it.
Just
someone
needs
to
figure
it
out.
I
really
couldn't
find
any
good
examples
of
it.
But
that's
probably
because
just
like
in
the
react,
World
there's
alternative
State
Management
libraries
that
are
specific
to
those
Frameworks
and
no
one's
building.
B
Anything
that's
meant
to
to
work
across
these
Frameworks.
So
this
this
could
be
one
of
like
the
first
public
projects
that
is
actually
doing
that.
So
all
that
said
the
way
that
I
was
thinking
about
it
was
there's
a
store
package.
Graphical
store
that
just
gets
exported
I
think
the
most
logical
place
for
that
to
live
is
inside
toolkit.
But
it's
a
lot
of
code
like
it's
a
huge
bit
of
code,
it's
the
biggest,
maybe
its.
B
A
The
toolkit
is
the
the
framework
agnostic
part
for,
for
all
intents
and
purposes,
and
we
don't
have
to
create
separate
module
boundaries
for
front-end
packages
that
much
because
code
splitting
takes
care
of
that
these
days.
So
breaking
things
into
Lots
is
a
little
like
for
plugins.
It
works
great
but
like
yeah
I,
don't
know
but
yeah.
Well,
that's
another
question.
Speaking
of
plugins,
though
I
think
even
just
with
our
current
wait
stand
even
with
our
current
plug-in
implementation
would
be
super
powerful,
not
to
mention
other
plug-in
apis.
A
We
plan
to
introduce
and
here's
how
right
you
let's
say
you
want
to
make
so
have
we
made
doc
the
classic
doc
Explorer?
Is
that
its
own
plug-in
yet
or
no.
B
A
B
Yeah
yeah,
so
so
that's
that's
okay.
So
the
the
plan
was
get
rid
of
context
and
just
make
a
store
so
step.
One
was
just
like
take
all
of
the
context:
State
all
the
values
that
you
pass
down
through
the
layers,
the
react
providers
and
just
move
them
into
that
store
that
guy
I
don't
know
the
last
time
you
looked
at
graphical
react,
but
it
is
because
it's
because
we're
it's
using
context
in
a
way
that
context
really
isn't
meant
to
be
used.
B
A
Yeah
right
and
so
not
not
like
to
fault,
but
you
know
it's
just.
It
was
the
way
to
get
it
implemented.
B
B
What's
going
on,
so
it's
it's
layered
in
it's
just
there's
a
lot
of
layers
and
it's
very
complex
so
simply
like
taking
the
the
context,
values,
the
state
and
putting
them
into
some
kind
of
store
was
simple,
but
then
you
have
all
of
the
props
that
are
coming
in
and
the
way
that
those
context
layers
are
currently
built
and
the
complexity
is
primarily
because
of
all
those
props
that
are
coming
in
and
some
of
those
props.
A
B
It
makes
it
makes
sense
until
you
have
to
get
other
people
involved,
because
the
complexity
just
grows
like
you
can't
even
it's
well
commented
Thomas
did
a
great
job
like
putting
notes
all
over
the
code.
It's
still
very
difficult
to
follow
so
I,
so
I
realized
after
just
refactoring
a
lot
of
that
code.
B
That
I
wasn't
gonna,
get
it
right,
and
that
barrier
was
the
props
that
come
in,
because
you
have
to
pick
an
entry
point
to
set
the
initial
state
in
this
sorry,
the
secondary
state
in
the
store,
because,
with
the
stand
you
just
give
it
an
initial
state
right.
So
all
the
props
that
are
currently
coming
in
need
to
need
to
effectively
be
nullable
because
a
lot
of
those
State
values
you
can't
initially
populate
them,
it's
impossible
to
do
it
without
because
it
relies
on
other
props.
So
that
was
the
barrier.
B
It
was
really
the
props
and
so
what
I?
What
I
was
doing
and
I
have
a
branch
that
has
all
of
it
and
it
doesn't
work
really
well
I
mean
it
doesn't
have
all
those
prop
values
getting
set
correctly.
Basically,
in
the
graphical
component,
right
in
the
graphical
package
in
the
graphical
interface
component
is
where
I'm
listening
for
you
know,
there's
just
a
there's:
a
use
effect
that
listens
for
prop
changes
and
then
uses
the
set
State
value
in
the
stand.
B
Villa
store,
vanilla
store
to
set
all
that
to
set
all
that
up.
The
problem
is
that
the
problem
that
I
ran
into
was
was
again
just
related
to
the
complexity
in
the
in
the
context
and
just
trying
to
like
build
a
mental
model
around
how
it
was
all
working
so
I.
So
so
then
I
thought
you
know
what
let's?
A
C
A
I
think
it'll
be
easier
to
do
without
context,
but
in
this
strategies
like
we
ran
into
module
boundary
layer
issues
right
where
it
was
like
trying
to
it
was
it
didn't
like
the
context.
References
didn't
match
up
across
plug-in
boundaries,
but
that's
how
we
that's
just
because
of
how
we
wrote
it.
Basically,
I.
B
Think
they're,
it's
it's
layered
in
so
the
the
graphical
provider
component
has
those
context
layers
in
it.
By
default,
like
you
can't
choose
to
not
have
them
right,
so
they're
effectively
default
plugins,
which
so
it
makes
sense
that
they're
inside
graphical
react
because
they're
just
default
plugins,
but
pulling
them
out,
gives
us
an
opportunity
just
to
experiment
with
it
being
a
totally
separate
thing,
and
that
way
you
can
also
set
it
up.
So
that,
like
there's
graphical
react,
which
would
be
just
basically
would
just
be
like
editor
components
and
then
any
of
the
Chrome.
C
A
A
B
Yeah
yeah,
so
the
Prototype
sort
of
starts
with
an
editor
graphical
editor,
which
is
just
the
the
three
editor,
the
three
or
four
editors,
and
that's
it,
and
it's
just
a
component
that
you
just
takes
a
schema
and
it
runs
away
with
it
and
you
just
run
it
by
itself.
That's
why
I
built
it
that
way,
because
it
starts
with
that
core
and
then
you
can
slowly
like
you,
can
slowly
build
it
up.
B
You
can
build
the
UI
pieces
around
it
if
you
want
to,
but
in
in
graphical,
so
I
started
pulling
the
dock
Explorer
out
into
its
own
plug-in
and
using
store
inside
of
that
plug-in.
Just
so
that
it's
self-contained,
and
you
don't
have
to
you,
don't
have
to
think
about
you
could
just
use
it
and
that's
kind
of.
C
B
I
stopped
was
I
mean
not
stopped
I.
Just
that's
where
I,
that's,
where
I'm
at
right
now
is
is
is
doing
that
so
I
think
what
I'm
gonna
do
is
unwind
everything
in
that
Branch
right
and
just
because
remember,
I
I
have
both
everything
pulled
out
into
a
store
and
I
have
that
doc,
Explorer
plug-in
as
a
as
a
separate
package.
I
think
I'm
just
going
to
unwind
all
that
and
just
start
with
pulling
out
the
dock
Explorer
and
pulling
up
history
just
to
like
get
a
solid
foundation
and
then
slowly
move.
B
Try
attempt
to
move
the
rest
of
those
context
layers
into
another
store.
It's.
A
Or
like
some
like
bits
that
compose
the
full
store.
However,
you
look
at
it,
but
like
yeah,
what
you're?
What
you're
talking
about
is
a
great
idea
and
like
the
so
and
it
will
make
react
graphical
so
much
leaner,
right
and
more
focused
right,
because
the
state
will
be
maybe
extracted
out
into
toolkit
right
and
then
the
the
logic
the
business
details
of
those
plugins
will
be
moved
into
a
separate.
You
know
plug-in
workspace
right.
A
So
then,
all
of
a
sudden
graphical
react
becomes
50
percent
simpler
in
its
aims
and
its
API
surface
I
mean
it
will
spread
out
to
other
parts
of
the
repo,
but
it
starts
it
brings
some
like,
maybe
like
at
least
from
a
structure
design
standpoint
help
might
improve
just
like
understanding
the
code
base,
because.
C
B
I
think
I
think
I
think
the
green
abilities.
Readability
is
super
important
here.
Let
can
you
see
my
screen
yep,
so
let
me
jump
in
here
a
little
bit.
Okay,.
A
So
we
were
I
wanted
to
have
a
separate,
plug-in
workspace
that
would
just
be
for
graphical
plugins
instead
of
packages,
so
there'd
be
like
a
whole
like
next
to
packages
directory,
it
would
be
a
plug-in.
A
B
A
Where
yeah
I
have
an
issue
related
to
this?
Actually,
where
I
proposed
this,
where
we
would
basically,
there
would
be
like
a
language,
Service
and
then
vs
code
directory,
and
you
know
like
basically
or
or
just
kind
of
like
it.
Yeah
like
there
would
be
like
a
web
directory,
for
example,
or
something
I,
don't
know,
there's
all
kinds
of
ways
that
we
could
reorganize
it,
but
yeah
using
the
big
packages.
A
Directories
I,
guess
the
the
easy
way
to
solve
a
hard
problem
for
now
so
like
just
put
it
all
in
one
place,
everyone
knows
where
to
find
it
yeah,
because
we
put
in
plugins
and
people
might
not
know
where
to
look
because
they're
like
oh,
it's
a
mono
repo
that
has
something
has
workspaces
outside
of
just
packages
right.
If.
B
B
I'll
just
show
you
the
code:
I,
don't
need
to
show
you
the
UI!
So
there's
a
here,
so
here's
Right
Packages,
here's
that
store
package.
Here's
the
doc
Explorer
package
inside
the
store
it
just
exports,
graphical
store.
C
A
It's
letting
you
think
about
schema
and
structure
of
the
store
and
logic
and
things.
B
C
B
Because
a
lot
of
the
incoming
props
are
shared
between
the
context
layers,
it's
not
exactly
one-to-one
and
that's
where
the
that's
where
I'm
having
difficulties
is,
is
pulling
all
that
apart
and
again,
like
I
said,
I
tried
to
do
it
all
in
one:
go
which
was
kind
of
dumb
I
should
have
I
should
have
done
it
in
smaller
bits,
but
I
really
was
just
trying
to
understand.
What's
going
on
with
those
context
later
so
I
think
I
have
a
good
idea
of
how
that's
working
exactly
how
to
get
those
propped
in
and
how
to.
B
Right
now,
all
all
that
component
does
is
like
pull
out
the
react.
The
graphical
react
bits
and
pass
all
those
props
through
into
the
provider,
and
then
the
graphical
react
package
is
currently
doing
all
of
that
work.
If
we're
gonna
do
this,
all
of
that
State
setting
initially
all
of
all
of
the
initulations,
the
initialization
code,
needs
to
move
up
into
graphical
and
that
that's
a
radical
change
like
you.
A
Can
I
have
I
have
an
idea
for
you.
It
is
a
big
radical
change
and
yeah
and
but
I
think
it's
going
to
be
an
important
change
because
I
think
it.
This
will
solve
the
plug-in
module
boundary
layer
right.
So
if
we
make
it
so
that
the
store
is
instantiated
by
the
users,
somehow
even
programmatically
right,
but
it
becomes
an
instance
in
their
environment
in
their
bundler.
You
know
when,
in
in
their
code,
they're
creating
something.
That's
that's
like
a
store
that
they
pass
to
a
thing
or
it's
like
they
call
them
like.
A
Something
happens
where
that
store
is
instantiated,
but
it's
instantiated
by
their
implementation.
So
any
graphical,
SDK
or
graphical
react
or
graphical
store.
Implementation
should
be.
You
know
like
instantiating
that
store
and
then
maybe
like.
We
have
to
do
something
where
we
pass
plug-ins
to
the
store,
instantiation
or
something
yeah
or
like.
C
A
A
B
A
A
And
it's
zero
X
like
no
one's
expecting
like
people
are
probably
hoping
it
for
it
to
break
and
evolve.
You
know
like
it'd,
be
great
if
it
works
backwards
compatible,
but
there's
no
reason
that
it
has
to,
and
if
you
need
to
break
the
the
API
of
graphical
react
yeah
to
get
where
you
want
to
go.
That's
totally.
Fine
and
graphical
doesn't
even
need
more
than
a
patch
version
for
that.
If,
if
it
doesn't
change
how
props
are
used,
yeah
right,
yep.
A
A
B
B
Package
sure
anything
of
how
close
I
think
I
can
get
to
that.
Let
me
show
you:
let
me
show
you
graphical.
A
At
least,
or
maybe
there's
just
like
one
minor
breaking
change
like
maybe
a
a
minor
chain
like
Improvement,
if
anything
to
the
plugin
API
for
sidebars
or
something
like
I
I,
still
want
to
make
the
the
the
the
prop
either
go
from
plug-in,
rename
from
plug-in
to
sidebar,
Plugin
or
plugins
expands
as
an
gains,
a
type
parameter.
Yeah.
A
We'll
get
that,
though,
we'll
get
there.
We
can
grow
into
that.
That
doesn't
even
require
breaking
change,
though,
but
if
you
have
to
break
graphical
SDK
to
do
that
graphical
react
rather
then
yeah
that
should
be
fine
and
the
improvements
will
only
help
any
nascent
plug-in
maintainers
out
there
right.
So
it's
like
well
yeah.
It
breaks
the
change,
but
now
I
can
like
add
an
entire.
Like
you
know,
additional
like
state.
For
my
you
know,
Plug-In
or
I
could
even
like
interact
with
the
state
from
graphical
Explorer,
Pathfinder
plug-in.
A
B
It
certain
it
certainly
unlocks
it
certainly
unlocks
a
lot
of
possibility
for
people
to
actually
use
graphical
as
like
a
reference
and
to
build
new
stuff
I've
I
when
I
first
got
involved
earlier
this
year
and
as
I
thought
like
that's
the
way
to
go,
people
are
going
to
love
it.
The
truth
is
I,
don't
think.
A
B
A
That
those
folks
are
people
we
want
to
empower
but
they're,
like
five
percent
or
two
percent
of
our
users
right.
The
vast
amount
of
our
users
want
to
enable
a
few
plugins
for
their
internal
company
graphical
you
know,
and
or
they
want
to,
they
want
to
ship
with
their
open
source
framework
or
their
commercial
framework
version
of
graphical.
A
With
these
plugins
you
know,
and
they
they're
fine,
using
the
theme
and
like
Gatsby,
for
example,
is
happy
doing
this
they're,
not
they
don't
forked,
it
they're
not
at
a
lower
level,
they're,
happily
consuming
at
the
top
graphical
layer,
there's
another
one
that
mentioned
this
too,
on
on
Twitter.
Other
companies
I
think
graph
CMS
is
also
using
graphical
2
directly.
A
So
yeah,
like
you,
know
like
we
we
have
like,
and
we
can
continue
improving
on
this,
but
we
have
gotten
graphical
to
the
point
right
now,
where
it's
sufficient
for
a
custom
product
design
on
an
Enterprise
level
like
we're
there.
We
just
can
make
it
better.
You
know,
and
the
plugins
are
part
of
that
recipe.
Yeah
some
people
are
building
entirely
customized
clients,
but
that's
only
that's
usually
when
they
have
like.
Incredibly,
custom
use
cases
and
you
run
across
a
lot.
A
As
new
things
happen,
people
will
keep
having
more
and
more
new
cool
ideas
and
they'll
be
outside
of
the
scope
of
what
we
aim
to
do
anytime
soon
and
that's
what
this
stuff
is
here
for,
but
you're
gonna
get
like
a
kind
of
anecdotal
bias
that
this
is
a
huge
that
there's
so
many
people
with
these
extremely
out
there
use
case
needs,
but
remember:
98
of
them
are
happily
using
it
right
now
right,
the
others
are
the
ones
that
need
to
push
it
Beyond
right
and
that's
where
we
help
them
with
the
SDK
level
stuff.
B
B
The
question
is,
how
do
we,
how
do
we
trim
a
lot
of
the
fat,
and
there
was
there's
one
example
that
that
I
had
top
of
mind
when
I
was
working
on
this
early
late
last
week
it
was
a
prop.
Let
me
let
me
see
if
I
can
find
it
here.
It
was.
A
B
Trying
to
figure
out
which
one
it
was
it
was
one
Lauren
from
The
Guild
added
it
in
earlier
this
year.
I
believe
for.
B
When
they
launched
yoga,
two
I
think
it
was
for
the
the
that
forked
version
of
graphical
that
they
initially
included
with
yoga
too,
which
they've
since
replaced
with
graphical
2.0
I,
don't
remember
which
one
it
was
but
I
couldn't
find
any
like.
It
was
so
obviously
just
super
helpful
for
that.
One
use
case
for
that
forked
for
that
forked
version
like
for
that
thing,
it
was
useful,
but
I
couldn't
I
couldn't
find
like
anybody
else
using
it.
B
Just
gonna
go
to
the
issues
and
and
look
for
it
because
I
don't
remember
which
one
it
was.
A
A
B
A
B
This
and
that's
a
you
know:
when
do
we?
When
do
we
decide
to
do
that,
like
it's,
certainly
not
necessary,
but
again
for
readability
for
discoverability,
just
for
understanding
the
code
base,
anybody
else
that
wants
to
get
involved.
It
would
be
really
nice
to
name
space.
A
lot
of
these
props
right,
like
the
the
context
slices
and
the
the
store
State
slices
like
those
are
appropriately
named.
If
we
could
do
the
same
for
these,
these
props.
That
would
be
great.
B
Let
me
show
you
what
I
my
thoughts
on
that
that
I
started
building
in.
So
if
you
go
to
the
store
here,
oh
and
maybe
I
kicked
oh
there.
It
is
options
so
there's
the
store
has
an
options,
slice
and
the
options.
B
Slice
is
effectively
like
a
sidecar
just
for
all
the
incoming
props,
so
that,
when
you
need
to
in
the
graphical
component
or
anywhere
that
you
need
to
set
and
get
that
State,
that's
derived
from
the
props,
because
it's
different
right,
like
there's
in
the
context
layer,
there's
context
state
that
is
derived
from
prop
values.
Right,
it's
not
a
one
to
one.
You
know.
Sometimes
it's
like
the
fetcher
is
one
to
one
give
me
the
fetcher.
B
It's
got
to
come
in
as
a
prop
I'm
going
to
store
it
somewhere
and
I'm
just
going
to
use
it
when
I
need
it,
and
if
a
new
one
comes
in
I'm,
going
to
store
that
and
I'm
gonna
use
that
there
are
other
context,
values
that
are
derived
from
incoming
incoming
props
and
that's
why
I've
I've
been
I've
been
thinking
about
this
like
there's
just
in
options
right,
so
that
all
the
props
that
come
in
are
graphical
options
right.
It's
it's
namespaced
as
an
option.
B
It's
sort
of
no
longer
a
prop
because
we're
outside
of
react
in
the
store
in
react.
It's
a
problem
outside
of
react.
It's
just
an
option.
It's
just
an
option
for
the
for
the
state.
So
this
is
the
direction
that
I'm
heading-
or
this
is
again.
This
is
sort
of
where
I,
just
kind
of
like
ended
up
over
the
last
couple
of
days.
B
I
I've
been
doing
other
things
for
the
last
couple
days
and
haven't
touched
it,
but
the
idea
is
just
the
store
should
be
able
to
operate
without
the
props
other
than
the
fetcher.
It
needs
the
fetchum.
You
know
it
needs
to
fetcher,
and
so
the
Sidecar
sidecar
options
make
sense
I'm
going
to
just
keep
going
down
this
road,
because
otherwise,
if
you're
looking
at
my
screen
now.
A
You
want
to
know
what
I
did
once
with
the
with
so
real
quick
going
back
to
your
store
and
where
you're
at
because
right
now,
you're
you're,
just
like
composing
these
slices
right
in
a
kind
of
like
a
design
level,
almost
right,
where
you're
just
kind
of
Imagining
the
data
design.
So
I
did
this.
Similarly
with
recoil
and
something
that
was
like
one
additional
step
that
helped
me
so
much
it
was
years
ago,
and
you
know
I'm
glad
we're
I
like
recoil,
is
great
I'm
glad
we're.
Choosing
this
either
way.
A
I
made
like
a
little
sandbox
for
just
the
state
implementation
and
what
I
had
to
do
was
I.
Had
it
just
like
write
like
read
out
the
Json
stringified
values
of
everything,
there's
probably
like
an
inspector
or
something,
but
you
could
just
like
it
works
without
even
react.
So
you
could
just
like
create
a
code
sandbox
with
your
state
implementation
and
just
start
saying
like
like
abstractly
doing
the
things
you
imagine
it
doing
and
like
thinking
about
the
side,
effects
and
I
know
it's
like.
A
Maybe
that
won't
be
helpful
for
you,
but
it
was
helpful
for
me
where
I
could
be
like
here,
I'm
going
to
create
just
a
simple
little
input
and
button
and
click
this
and
that's
the
query
and
the
submit.
But
then
here's
the
the
serialized
Json
State
as
this
is
changing
like
kind
of
building
your
own
little
kind
of
like
Redux,
Explorer
or
or
Explorer.
A
B
Know
I
had
a
very
similar
thought,
but
I
took
a
different
tactic
and
instead
of
building
a
Sandbox
where
you
can
just
sort
of
play
with
state
with
random
UI
components,
because
I
don't
I,
really
don't
think
that
that's
the
issue.
The
issue
is
more
in,
like
a
the
issue,
really
is
more
integration
right
so.
C
B
More
like
how
does
the
when
you
pass
in
a
new
I
mean
passionate
passing
an
initial
query:
I,
don't
really
care
that
it
I
know
that
it's
going
to
end
up
in
the
state
store.
What
I
need
to
know
is.
Does
the
dock
Explorer
or
you
know,
do
all
the
components
like?
Are
they
responding
separately?
So
the
tactic
that
I
took
it's
not
in
here,
it's
actually
in
this
type
of
Branch
was
just
to
put
storybook
back
or
Ladle
I
was
using
Ladle
I.
B
A
I
I,
don't
think
pulling
pulling
State
and
maybe
even
some
business
logic
out
of
there
might
allow
you
to
to
do
that
in
a
way
that
where
it
could
be
like,
maybe
graphical,
react
and
then
some
of
like
or
like
graphical
com,
slash
react,
Dash
components,
the
graphical
react
implements
or
It's
Tricky,
but
yeah
like
it's.
B
A
It
would
contain
a
UI
library
and
the
hooks
and
business
logic
implementation
for
react,
just
like
one
would
for
view
or
something
where
it
takes
the
store
and
the
toolkit,
and
you
know,
Implement
and
Implement
its
own
design
Library
against
those
right.
These
are
compatibility.
Libraries
like
for
let's
say,
react
are
Dual
Purpose
right,
their
UI
library
and
a
business
logic
or
like
implementation
of
the
state
right
so
as
much
business
logic
is
this.
A
This
store
package
can
have
the
better
right
because
then
it
can
be
shared
by
both,
but
at
some
point
we're
going
to
have
to
say
well
now
we
have
to
implement
this
in
a
hook
with
a
use
effect
thing.
You
know
to
react,
you
know
and
so
that
it
can
just
be
the
the
glue
code
right,
because,
ideally
as
much
business
logic
lives
in
a
in
somewhere,
something
that's
react,
agnostic
right
and
then
it
will
feel
more
like
a
design.
Library.
B
B
Is
you
know
not
when
I
say
this
I
mean
again
the
one-to-one
replacement
of
context
with
the
with
the
with
the
store
package?
I
I
haven't
finished
it
so
I
can't
be
totally
sure
and
I
and
I'm
just
now
becoming
familiar
more
familiar
with
graphical
react,
but
I'm
I
can't
be
totally
sure,
but
I'm
pretty
sure
that
that's
correct
that
it
will
remove
all
of
the
critical
business
Logic.
The
only
the
only
business
logic
that
I
can
think
of
that
would
be
left.
B
C
B
A
A
And
actually,
if
you
want
to
see
a
great
example
of
that
sandpack
does
that
with
code
mirror
6
where
they're
just
like
we
have
one,
they
have
one
core:
JavaScript
library,
that's
used
by
all
the
react,
libraries
or
the
react
and
View
and
other
modes
that
that
like
takes
care
of
like
some
of
the
core
logic
of
mounting
the
components
at
the
end
of
the
day,
you're
still
going
to
have
to
do
something
with
a
ref
or
an
actual
ID
or
whatever
the
heck
that
the
other
I
don't
even
know.
A
If
anyone
who
does
viewers
felt
is
watching
they're
like
this
is
why
Ricky
said
svelt
was
working
and
it
wasn't
even
working
yet,
okay,
so
but
yeah,
there's
yeah,
there's
something
really
cool.
We
can.
B
A
That
are
in
here's
a
way
to
think
about
this.
Iteration
I
know
how
you're
thinking
like
just
replace
context
one
to
one
right.
But
what
if
your
only
goal
in
this
is
get
graphical,
react
to
use
a
shared,
State
and
business
logic
without
making
any
breaking
changes
to
the
graphic
goal
component
right,
you
can
break
changes
to
any
other
Library.
A
You
need
to
even
code
me
or
graphql,
whatever
you
know
if
we
have
to,
but
you
know
minor
release
or
new,
even
new
major
version,
because
we
have
to
make
an
API
change
or
whatever,
or
you
know,
maybe
we
actually,
we
should
be
I
mean
well
in
this.
Well
depends
where
you
there's
a
lot
we
can
do
without
making
breaking
changes
to
to
graphical,
but
obviously
we
have
to
stay
on.
Code
mirror
5
to
not
make
breaking
changes,
but
yes,
like
yeah,
but
yeah.
B
B
A
Well,
one
thing
to
think
about
the
plugins:
at
the
graphical
level:
do
you
use
the
SDK
but
again
plug-in
authors
won't
mind
breaking
changes
in
SDK.
You
know
and
there's
already
the
past
versions
that
can
use
if
they
don't
have
time
to
upgrade
right
away
right.
Those.
B
A
B
I
I
do
want
to
make
I
want
to
make
it
palatable
right,
because
it's
just
a
proof
of
concept
right
I'm,
not
proposing
that,
like
that.
This
is
something
that
we
tackle
right
away,
but
it
it
does
help
unlock,
there's
a
lot
of
power
in
doing
this,
there's
a
lot
of
like
Freedom
around
the
library
there's
two
issues
that
have
come
up
the
last
couple
of
weeks.
B
One
is
this:
you
know
next
try
next
JS
Dynamic
import
thing
where
code
mirror
is
when
you
use
the
Explorer
plugin
and
it
because
you're
feeding
in
the
query
and
you're
feeding
in
the
on
edit
function
right
when
it
changes
code
mirror
is
re-rendering.
That's
all
related
to
context.
If
that.
B
I,
wouldn't
that
wouldn't
happen,
but
it's
I
mean
again
it's
it's
look.
It's
like
it's
very
low
priority,
but
that's
one
issue
that
I
that
I
believe
would
be
solved
by
removing
context
and
the
other
one
is
this
person
on
on
Discord
that
is
trying
to
build
the
diff
editor
thing?
B
Yes,
where,
because
of
the
because
of
the
I
I've
got
a
of
my
own
repository
where
I
try
to
do
that
where
I.
B
Just
wanted
to
I
wanted
to
to
see
because
I
believe
the
person's
name
is
R
I'm,
not
looking
at
Discord
rrva.
They,
they
shared
some
screenshots
of
of
like
how
it
was
rendering
and
part
of.
B
B
A
A
A
Query,
editor,
plug-in
type,
sidebar,
plugin
type
results,
you
know,
and
then
you
could
configure
which
one
is
chosen
by
default.
So.
B
I
was
thinking
about
it
at
the
lower
level
than
that,
so
all
I
did
was
try
to
add
in
a
a
set
just
just
there's
two
there's
not
two
response:
editors,
one
on
the
left
and
one
on
the
right
right
that
just
takes
up
the
same
amount
of
space
that
the
other
one
did,
but
because
there's
context
layers
you
can't
simply
like
render
the
the
new
provider
inside
you
can't
let
you
can't
Nest
them
like
that,
because
it
just
uses
the
same
context.
If.
A
B
A
Here's
the
thing:
John
I'm
glad
that
you're
worrying
so
much
about
this
really
specific
case,
but
we
don't
have
to
like
figure
out
how
to
First
Class
support
like
this
90.
This,
like
.001
use
case,
but
I'm,
so
glad
that
you're
thinking
about
it,
because
maybe
maybe
we'll
think
of
some
cool
ways
to
structure
things,
to
make
this
easier
to
do
in
the.
A
Yeah
but
think
about
some
other
ones,
think
about
like
look
at
some
others
out
there,
like
one
thing,
I
like
to
do
sometimes
like,
so
let
me
double
check
I
think
it
was
Gatsby
and
grac
CMS.
Whoever
else
is
listening,
I
think
graphco
for
oh
I'm,
trying
to
see
who
else
is
using
2.0
Bruno
yeah,
wait!
No!
No!
No!
No!
No!
That's!
That's
someone
just
tweeting
on
something
else!
Gatsby
graph
base,
yes
graph
bases,
yep
yeah.
A
B
They
were
using
graphical
1.0
when
they
launched,
which
was
like
earlier,
which
was
like
early,
maybe
spring
or
early
summer,
when
they
like
first
went
out
and
their
customization
of
actually
they
weren't
using
1X
because
it
had
tabs
I,
don't
know
what
they
were
doing.
They
must
have
been
using
playground
initially,
that
might
have
been
One.
B
And
and
they're
their
integration
was
was
pretty
good.
It
felt
like
they
had
it
customized
to
sort
of
fit
into
their
UI.
It
was
a
little
wonky,
because
you
know
the
docs
would
like
slide
out
from
the
right,
and
that
was
a
little
bit
weird
because
they
had
it
embedded.
A
B
They
upgraded
to
to
2.0
and
as
far
as
I'm
concerned,
it's
not
that
it's
not
as
great
of
an
experience,
because
it's
they're
not
let
they're
not
currently
leveraging
it
just
doesn't
really
fit
inside
of
their
UI,
because
they
were
heavily
customizing,
the
last
version
and
with
the
new
version,
because
it
looks
so
much
better
by
default.
They're
doing
less
customization,
so
I
I
know
that
there
are.
A
Keep
in
mind,
though,
they
might
be
not
able
to
do
as
much
customization
right
now
for
whatever
reasons,
and
it
just
might
be
limitations
for
their
purposes
but
like,
for
example,
at
my
company,
we're
already
like
planning
on
our
like
to
write
our
own
plugins
like,
but
it's
internal
right,
so
keep
in
mind.
A
lot
of
people
are
using
graphical
and
you
can't
see
it,
and
this
is
going
to
be
the
the
wonderful
and
sad
part
of
being
an
open
source.
A
Maintainer
of
developer
tooling
is
some
of
the
coolest
versions
of
What,
like
implementations
out
there,
like.
Maybe
you
can't
see
open
source,
but
if
you're
out
there,
and
you
have
a
really
cool
implementation,
you
want
to
show
John
and
you're
listening,
because
you
want
to
you
have
ideas
for
how
he
can
help
improve
it
or
any
of
us
can.
Let
us
know
in
the
Discord.
A
Follow
us
follow
my
channel,
don't
actually
it's
not
very
it's
bad!
It's.
A
Yeah
hit
the
Subscribe
button,
that's
what
they
say:
yeah
I,
don't
I'm
I'm
too
old
to
watch
YouTube,
no
I'm,
just
joking,
so
anyways
yeah,
there's,
there's
yeah.
There's
a
there's!
A
lot
going
on
out
there
and
I
would
just
assume
plug-ins
are
being
more
used
more
than
you
realize.
Just
look
at,
for
example,
just
look
at
the
npm
stats
on
individual
plugins,
for
example.
That
might
tell
you
things
like
well,
let's
look
at
npm,
let's
see
if
there's
a
graphical
plug-in
something
maybe
someone's
already
shipped
to
an
open
source.
A
Yep
wait,
that's
a
plug-in
for
something
that
uses
graphical
yeah.
This
is
like
a
lot
there's
Frameworks
that
have
a
plug-in
for
graphical.
So
this
is
your
search.
Query
doesn't
is
confusing
plug-in
for
the
graphical
IDE
that
displays
graphql?
Oh
no,
it's
old!
No!
Maybe
they
have
to
be
recent
too.
A
Well,
we'll
get
to
it
in
a
bit
graphical
plugin,
Explorer.
Speaking
of
we
need
to
fix
I
plan
on
adding
a
graphql
or
graphical
npm
bot
user.
So
it
doesn't
look
like
everything
is
being
published
by
me
anymore.
Just
so
yeah,
let's
see
backstage,
has
a
graphical
plug-in.
Thank
you.
Yeah.
Let's
see
render.
C
A
Plugins,
like
have
you,
play
yeah
sorry,
I'm,
wasting
time
on
the
call
here:
I
apologize,
no.
B
No
I'm
looking
at
it
too.
It
is
interesting.
There's
some
like
big
companies
that
are
building
stuff.
Ibm
has
a
plug-in
for
previous
versions
of
graphical.
A
You
can
go
search
any
of
the
package
repositories
for
graphical
for
any
language,
any
package
manager
and
you
will
find
something
whether
it's
Pi
Pi,
whether
it's
cargo,
whether
it's
ruby,
gems
everyone's
got
graphical
going
on
it's
cool
and
you
can
just
see
there's
just
a
whole
universe
and
across
other
languages
you'll
see
the
you'll
sometimes
see
the
non-jsx
version,
which
is
teaches
you
a
little
bit
about
like
how
to
support
Advanced
implementations
outside
of
that
which
help
people,
because
it's
not
fun
running
a
webpack
bundler,
that's
orchestrated
by
another
language,
I
think
Ruby
on
Rails
added
and
removed
that
just
as
an
example.
A
So
that
means
that
the
people
who
maintain
the
graphical
plugins
for
various
Ruby
on
Rails,
related
framework
or
libraries
had
to
like
go
from
the
jsx
version
and
being
able
to
use
the
npm
package
too
back
to
the
CDN
version
and
using
react,
create
element,
so
I,
think
or
maybe
I
I
could
be
wrong
there,
but
either
way
it's
just
interesting.
They
all
have
all
these
other
constraints.
We
don't
often
think
of
in
the
JavaScript
world,
but
you'll
see
some
really
interesting.
A
Implementations
out
in
the
wild
there,
too,
just
yeah,
just
you
could
say,
check
I,
can't
remember
what
golang's
repository
is
called
Java.
It's
just
yeah
pH,
of
course.
Obviously,
in
the
your
neck
of
the
woods
composer,
whatever
the
packages
you
know,
sir
yeah
see,
if
I
can
remember
the
names
of
the
package
managers
and
repositories
for
all
the
major
languages,
but
yeah
it's
out
there
there's
yeah
a
lot
of
it's
still
1.0
implementations
but
yeah.
A
If
anyone's
out
there
has
a
cool
2.0
implementation,
it's
open
source
in
the
wild,
especially
useful
for
us.
Please
share
it
with
us
like.
B
A
Need
a
big
feedback
discussion
ticket
just
be
like
here
just
dump
all
your
feedback
from
2.0
from
Custom
implementations
here
and
if
it's
like
a
big
enough
issue,
we'll
create
new
issues
or
bugs
from
it
or
we,
because
you
can
do
that.
You
can
take
a
discussion
response
and
then
convert
it
to
an
issue.
Yeah.
B
I
think
that's
a
great
idea,
pinning
pinning
a
2.0
feedback
discussion
in
the
repo
is
a
great
idea.
I
still
haven't
received
any
feedback
on
the
discussion.
I
started
about
the
visual
operation,
Builder,
which
it
would
be
nice
to
hear
what
what
folks
think
about
that
I.
Just
you.
A
B
The
truth
is
like,
as
as
sort
of
wonky
as
the
one
graph
Explorer
plug-in
is
like
it
works
like
it.
Just
it's
fine
works,
you
know
it
works
and
there
doesn't
seem
to
be
a
lot
of
currently
a
lot
of
drive
to
you
know
doing
the
Pathfinder
thing,
which
is
you
know
something
which
was
in
the
original
2.0
design
proposal
or
is
also
very
similar
to
something
that
to
like
to
what
Apollo
was
doing
right.
That
I
don't
even
know
directionally.
If,
if
that's
yeah,
hey.
A
So
I
think
Pathfinder
is
important,
but
I
think
we
can.
We
can
cover
a
lot
of
it
with
the
Explorer
plug-in
like
you're,
saying
and
here's
what
I
think
we
should
do.
We
should
consider
doing
I
think
we
could
do
a
2.1
where
we
enable
Explorer
by
default
and
then
show
people
how
to
instantiate
a
more
lean
one,
and
this
is
when
this
is
maybe
for
zestan,
because
then
we
can
say:
okay
now
you
can
have
only
Explorer
and
like
and
not
query
history.
A
If
you
want
right,
if
you
like,
so
we
kind
of
say
here
now,
you
have
more
options
and
you
can
actually
disable
these
things
as
plugins.
Now,
because
we
have
a
new
state
implementation,
it's
more
stable
and
because
the
state
implementation
is
better
and
you
get
a
you
get.
What
was
what
was
that?
B
I
think
that
Ricky
I
think
we
could
do
we
could
do.
We
could
do
the
select
your
default
plugins
thing
right.
We
could
do
that
right
now.
We
could
add
another
prop
to
what.
A
A
I
bet
you
almost
no
one
will
complain
like
that,
it's
on
by
default
and
like
and
if
we
especially
if
we
document
how
to
how
to
like
enable
graphical
with
only
the
plugins,
you
want
to
override
that
behavior,
for
example
right,
but
we
can
just
say
Explorer,
doc,
Explorer
and
like
like
everything's
like
even
code
exporter,
right
I
mean
code
Explorer
I'd
like
to,
but
that's
it's
a
little
bit
language
specific,
but
I'd
like
to
I,
don't
know
if
we
had
a
way
to
for
people
to
extend
code
exporter
to
like
export
a
python
script
or
something
if
they
want
just
by
using
a
template
thing
that
would
be
sweet
but
yeah
like
let's
get
there
somehow
I,
don't
know
either
way
like
yeah
I'm.
A
Just
thinking
that
there
could
be
a
breaking
change
where
the
default
plug-in
composition.
When
you
just
enable
graphical,
you
don't
specify
any
plugins
that
like
yeah,
then
we
wouldn't
have
to
go
through
the
step
of
telling
people.
This
is
how
you
enable
graphical
Explorer,
because
no
matter
how
much
we
say
like
here
in
the
docs
or
on
the
even
if
we
had
a
DOC
website
like
here's,
how
you
enable
graphical
Explorer
people
aren't
even
going
to
look
at
that
they're
just
going
to
look
at
graphical
two
and
be
like.
A
Oh,
this
is
cool
no
Explorer
yet,
and
they
don't
realize
that
they're
just
a
few
steps
away
from
enabling
it.
So
if
you
make
it
enabled
by
default,
it
could
just
be
like
hey
gr.
This
works
in
graphical
now,
that's
how
most
people
would
see
it,
even
though
they
don't
realize
it's
right
like
not
far
away.
You
know
what
I
mean
It's
Tricky,
though,
because
then
it's
it's
gonna.
A
You
know
it's
gonna,
add
additional
code
in
a
little
like
runtime
complexity
by
default
to
all
implementations,
but
we
could
just
try
it
and
if
it's
a
huge
issue
and
people
are
like.
Oh
my
God,
this
breaks
everything
and
it
like.
We
have
like
four
bugs
related
to
it,
then
maybe
we
should
reconsider
it,
but
I
think
it
would
be
fine
to
enable
Explorer
by
default
I'm
making
my
my
RSC
right
now.
B
Would
it
would
it
be
see
what
you're
saying
by
making
it
the
fault,
but
let's
say
we
did
that
within
a
very
strong
equivalent,
be
just
changing,
adding
a
prop
to
graphical
that
allows
you
to
select.
B
You
know
just
it's
a
Boolean
history,
true
or
false
doc,
Explorer,
true
or
false,
One,
graphics,.
B
A
That's
what
I'm
thinking
it's
yeah
right
but
then
like
what's
the
default
when
there's
no
plug-in
prop
specified
right.
So
if
you
want
to
disable
or
enable
something
you
we
can
make
it
so
that
when
you
add
plugins,
it
disables
everything
else
and
only
enables
the
plugins
you
provide.
But
then
we
have
our
happily
path
set
of
defaults
for
all
users
right.
B
Yeah
I
think
you're,
right
and
I
mean
we
all
kind
of
know
this
already
that
the
vast
majority
of
these
installs,
if
they're
aware
of
the
Explorer
plugin
they're
using
it.
A
For
a
lot
of
installs,
the
more
the
better
like
I've
even
had
people
tell
me,
don't
worry
about
bundle
size.
This
is
for
developers,
you
know
like
if
it's
a
big
fat
load
initially
and
we
get
everything
we
need,
and
it's
all
these
bells
and
whistles.
That's
all
we
want,
like
you
know,.
A
A
It
probably
matters
to
maybe
like
less
than
10
of
our
users
somewhere
in
there
right,
because
they're
doing
something,
that's
public,
maybe
it's
customer
facing
they
want
the
first
load
to
be
Snappy,
because
they're
competitors
for
slow
to
Snappy,
but
you're
only
going
to
remember
the
first
load,
the
first
time
all
that
shit's
gonna
be
cast
in
your
browser
and
you're,
not
gonna,
see
a
heavy
load
very
often
and
plus
I
wouldn't
be
surprised
if
there's
user
research
that
shows
like
a
cool
little
like
like
loading
screen,
gives
people
like
an
impression.
B
It's
delay
they
have
like
their
implementation
is
really
great
and
it
may
be
different,
but
initially
like
it
would
sort
of
like
pop
up
from
the
bottom
of
the
screen,
and
it
would.
It
was
sort
of
this
like
modal,
slash,
dialogue,
slash
takeover
thing,
and
so
the
loading
was
was
sort
of
obfuscated
by
that
initial
transition
into
like
it
displaying
it
was
really
cool,
and
probably
just
as
effective
at
disguising
the
fact
that
that
you
know
that
this
was
a
whole
different
set
of
code.
A
Like
especially
when
we
get
into
Monaco,
they'll
definitely
need
to
do
this,
like
no
offense
to
Monaco
editor,
but
it's
there's
a
price
for
the
all
the
bells
and
whistles.
It
comes
with
right,
but
it
it
like.
A
You
know
they
need
to
be
able
to
load,
be
loaded
as
lazy
components,
code
split
even
from
the
route
chunk
right,
because
you
don't
you
don't
want
to
block
the
initial
render
of
a
whole
developer
tool,
page
for
your
I,
guess,
I'm
thinking
an
SSR.
Oh
sorry,
I
I
forget
a
lot
of
the
most
developers.
Don't
have
to
deal
with
server
under
but
yeah
yeah.
It's.
B
A
About,
but
we
need
to
give
them
a
nice
client
render
path.
Maybe
we
could
even
have
a
special
export
of
graphical
component.
It's
like
graphical
slash
next
in
the
graphical
repository
and
so
you'd
import,
a
graphical
component
called
graphical
next
that
just
wraps
it
in
like
a
react,
lazy
loader
that
that
is
somehow
protects
for
this.
You
know,
I
I,
think
that
or
graphical
server
or
something
well.
B
There's
a
there's
an
issue
right
now
that
came
up
I,
think
last
week,
where
somebody
was
having
this.
This
cursor
jumping
thing
right
in
next,
when
you're
dynamically
importing
and
you're
using
the
Explorer.
We
talked
about
this
right.
B
The
simple
you
know:
we've
already:
we
already
have
a
like
a
a
fix
in
another
issue.
That's
that's
published
right.
You
just
see
if
it's
rendering
in
the
client
or
not-
and
you
don't
you-
don't
use
the
component
so
that
the
issue
that
we're
talking
about
here
is
actually
related
to
Explorer,
the
Explorer,
the
one
graph
Explorer
plug-in
and.
C
B
Or
re-rendering
internally
because
of
context,
but
in
the
larger
picture
of
of
using
graphical
and
xjs,
there's
already
a
fix
like
it's
a
fine
fix,
you
just
don't
render
it
if
you're
on
the
server
right,
there's
I,
don't
see
any
other
reason
to
to
to
make
a
change
to
accommodate
for
that.
But
this
is
this
issue,
though,
with
the
with
the
cursor
changing
is
related
to
context.
B
Context
not.
A
Really
really:
okay,
right,
right,
okay,
yeah,
so
yeah,
I,
agree.
I!
Think
we
could
do
this
all
at
once,
like
we
would
make
everything
in
the
sidebar
plug-in
Implement
zest
stand,
make
whatever
breaking
changes
beneath
graphical.
We
need
to
make
and
then
do
a
minor
version
to
graphical
and
say
now:
graphical
Explorer
is
enabled
by
default.
But
if
you
provide
plugins
breaking
change,
you
override
everything.
So
if
you're,
providing
just
the
Explorer
plug-in,
then
suddenly
schema
Explorer
and
everything's
going
to
be
gone.
A
B
Know
I
think:
let's
let
me
let
me
get.
Let
me
get
the
history
and
the
classic
dock
Explorer.
Let
me
get
them
pulled
out
into
separate
plug-in
packages
with
their
own,
just
saying,
stores,
removing
contacts
from
the
equation
and
then
and
I'll
I'll
PL,
I'll,
I'll
PR.
That
and
everybody
can
have
a
look
and
it'll
be
a
do,
not
merge
and
then
slowly
start
plugging
away
again
on,
like
removing
individual
context
layers,
because,
like
I
said
I
just
did
it
all
kind
of
whole
I
got
it
got
kind
of
messy
and.
A
C
B
Know
that
it
is
exciting,
it's
also
just
about
other
people
who
I
don't
know
how
many
other
people
are
following
along
with
this
I.
Don't
know
how
many
other
people
care
about
doing
this
with
graphical
like
who,
who
knows,
but
it
is
important,
I
think
for
the
future
of
graphical
to
have
this
store
separation
and
to
kind
of
get
out
of
the
context
world.
So
if
there's
anybody
following
along
doing
it
in
easily
explainable
chunks,
it's
probably
a
lot.
It's
just
a
lot
more
palatable.
B
So
my
initial
approach
was
just
to
like
shotgun
it
and
I'll
I'll
wind
that
back
and
I
should
have
code
in
on
probably
tomorrow,
because
it's
it's
not
that
difficult
to
extract
the
the
history
plugin.
A
A
A
B
What
the
the
storage
context
layer
provides.
It
has
those
methods
for
doing
that
which.
A
B
You
go
actually
I,
don't
know
if
you
could
do.
Let
me
look.
A
A
B
Think
the
re
I
think
now
that
I'm
looking
at
it
again,
I
think
the
reason
that
there's
a
context
layer
for
storage
is
because
you
can
pass
in
your
own
storage.
A
A
Yeah,
okay,
so
this
is
a
change
that
we
wanted
to
make.
Harith
was
going
to
work
on
this
as
part
of
making
it
part
of
toolkit
was
to
make
the
all
the
methods
async
so
that
you
could
say
use
local
forage
or
index
DB
or
idb
Library,
idb,
key
Val
I
think
would
even
work.
That
could
be
our
default.
If
we
wanted
to
switch
to
indexeddb
from
local
storage,
which
might
make
things
faster,
a
while
and
then
also
index
DB,
then
we
can
share
state
with
web
workers
for
I.
A
Don't
know
what
maybe
something
called
Monaco
you
know,
and
then
we
don't
have
to
spit
the
the
graph
graphql
schema
and
language
service
results
over
the
boundaries,
the
main
to
to
worker
boundary
layer
anymore,
because
then
they
can
just
both
write
and
read
from
the
indexeddb
State.
That's
the
change
that
me
and
Christine
wanted
to
make.
So
just
a
thing
to
think
about
when
you're
thinking
about
the
the
persistence,
if
they
close
the
faster,
we
jump
to
index
DB
the
better,
but
we
don't
have
to
do
that
now.
B
A
Does
this
all
of
it,
what
we
wanted
to
do
was
make
the
storage
API
be
compatible
with
Redux
persist,
style
apis
which
that
one's
probably
compatible
where
you
could
just
take
any
Redux
persist,
plug-in
and
use
it
right.
It's
like
a
spec
kind
of
almost
because,
like
Redux
persists,
sets
an
API.
Is
it
like
that,
because
I
know
it
was
like
that
for
a
couple
where
there's
just
like
any
State
Library
Dash
persist.
Libraries
we're
all
reusing
that,
but
it's
not
a
big
deal.
B
A
B
I'm
going
to
dig,
but
just
because
of
hydrating
the
store
initially
looks.
C
A
B
To
make
it
well,
it
is,
it
is
because
of
the
the
the
number
of
touches
across
graphical
react
for
for
the
storage
context.
Layer
is
high.
There
is
a
lot
of
there's
a
I
mean
it's
there's
a
lot
of
usage
there,
so
I
think
what
I'm
gonna,
what
I'm
gonna
do
is
I'm
just
I'm,
just
gonna
pretend
like
I'm,
not
gonna
make
it.
This
is
I'm
not
gonna.
B
Make
any
changes
here,
I'm
just
going
to
pull
this
out
into
a
storage
slice
and
treat
it
like
the
rest
of
the
contact
slices
without
making
any
changes
to
it.
Just
to
see
I
honestly
at
this
point,
I'm
not
even
sure
that
I'm
not
positive
that
this
single
store
idea
is
like
simple
to
do,
because
I
don't
understand
the
initial,
the
initialization
of
all
the
slices
with
the
props
and
everything
and
like
right.
A
B
Graphical
component,
so
I'm
just
going
to
try
to
I'm
again
I'm
just
going
to
one
for
one
and
just
just
to
see
like
it
may
not
even
be
worth
worth
working
on,
but
I
mean
I
mean
it
may
not
be
worth
seeing
through
to
the
end.
I
need
to
get
to
the
point
where
I
can
be.
A
A
A
B
Great
yeah,
so
I'll
I'll
have
some
of
this
in
probably
by
tomorrow,
by
the
end
of
the
week.
I'll
probably
have
this
in
and
and
we
can
begin
to
slowly
have
conversations
about
its
viability,
but
this
direction
has
been
great.
A
For
adding
slices,
real
quick
for
adding
slices,
add
them
to
the
plug-in
array
instead
of
using
a
hook
to
add
or
retrieve
the
slice.
Does
that
make
sense?
So,
when
you're
defining
your
plugins
in
the
array
right
and
you
you
have
this
little
Json
object
for
each
plugin
right,
we
add
a
new
key
for
like
store
or
state
or
something
and
there
you
can
pass
your
slices
alongside
the
component.
Do
you
see
what
I
mean
no
I.
C
A
I
want
to
understand,
I
I
think
this.
Will
this
will
help?
Let
me
look
at
how
our
current
I
have
to
I'd
have
I
have
to
look
at
how
graphical
works
first,
because
it
keeps
changing
but
I'm
pretty
sure
the
the
plug-in
prop.
So
so
this
is
him.
Graphical
react
now.
A
A
What
if
we
also
had
something
here
that
was
like
store
or
state
or
even
to
be
super
descriptive,
slices
and
slices
could
be
the
array
of
slices
that
need
to
be
passed
to
that,
so
your
plugin
would
export,
like.
Maybe
just
a
thing
called
slices
that
you,
like
you'd,
have
like
Explorer
slices
and
whatever.
B
B
Store,
oh
yeah,
because
because
react
plug-in
react,
doc,
Explorer
is
a
react
plug-in.
You
can
export
a
hook
just
like
what
happens
now
right
now,
it's
inside
of
graphical
react
but
explore
that
Explorer
effectively
explores
two
things.
One
is
the
component.
The
second
is
use
Explorer
context
right
so
that
you
can
access
those
values
so
that
you
can
excuse
me
so
that
you
can
click
inside
of
code
mirror
and
it'll
it'll.
Make
change
that
surface
doesn't
need,
does
does
not
need
to
change
the
package.
Graphical
Dash,
plug-in
Dash
react
Dash,
doc.
B
A
A
B
Doc,
Explorer
State
machine
is
simple
enough.
All
it's
all
that
is
doing
is,
is,
if
not
all
is
doing
most
of
what
it's
doing
is
just
keeping
an
array
of
this.
It's
the
nav
stack
like
the
it's
a.
B
Just
using
it
as
like
a
generic
term
like
the
the
context,
values,
the
Setters
and
Getters,
it's
really
just
focused
around
that.
So
there's
no
there's
no
slices
there.
There's
no
like
there's,
really
no
need
to
to
Mush
that
up.
That's
actually
I'm
the
since
I've
I've
tried
to
do
the
dock.
Explorer
already
I'll
tell
you
the
bit.
That
is
the
most.
That
was
the
most
Troublesome.
It
was
not
the
state,
it
was
not
the
the
hook
for
the
state.
B
A
A
For
you
yeah
also,
so
you
can
move
these
tests
out.
You
test
it
at
a
lower
level
where
it
like
mounts
the
component
inside
something.
That's
maybe
like
a
test
like
basically
like
this
is
where
I
feel
like
plugins
are
more
and
more
need
to
have
their
own
like
little
CLI
like
our
own,
create
graphical
Plug-In
or
something
whether
that
be
cgp.
A
So
if
you
use
cgp
cgp
test
would
give
you
like
would
allow
you
to
like
have
like
test
utilities
or
something
maybe
or
allow
you
to
like
kind
of
like
run
unit
tests
in
a
way
that
they're
like
they
assume,
like
the
the
frame
like
a
like,
not
actual
graphical,
but
like
a
skeleton,
maybe
or
just
better
unit
tests
like
like
the
test
more
at
the
unit
level,
because
I
think
I
bet.
A
You
could
like,
like
just
write
a
unit
level
test
for
the
content
component,
so
to
speak
and
write
unit
tests
even
for
smaller
components
within
the
plug-in
right.
So
like
well,
because
a
lot
of
this
stuff
that
we're
testing
it's
like
an
integration
test
right.
This,
like
you,
know
exactly
what
it
is,
but
like
graphical
dot,
dot.
Spec.Tsx
is
definitely
like
a
giant
integration
test
service
so
yeah.
Maybe
it
needs
to
be
broken
up
into
more
files,
so
it
can
be
run
concurrently.
But
that's
just
me
thinking
in
a
well.
B
One
atomization
one
thing:
that's
again:
I'm,
there's
a
there's,
a
there's,
a
really
simple
one-to-one
way
to
replace,
like
all
those
context,
layered
tests
with
these
store-based
tests.
In
fact,
it's
it's
actually
easier
to
do
it
with
the
store,
centers
and
Getters,
because
you
can
set
up
before
you
can
before
all
I.
A
B
It
actually
makes
it
it's
all,
there's
already
like
typed
fixtures
that
are
there.
So
it's
it's.
The
testing
is
the
reason
I
said
I
was
like
it
was.
The
most
difficult
part
was
not
because
I
can't
figure
it
out.
It's
just
because
it's
very
time
consuming
to
go
in
and
replace
all
of
that,
but
it's
all
there.
It's
just
I
just
need
to
like
get
in
and
do
it
so.
The
the
dock
Explorer
and
the
history
pulling
them
out
as
as
plug-ins
should
be
relatively
painless.
B
I've
got
most
of
that
work
already
complete
one
thing
that
I
do
that.
I
am
thinking
about
as
a
part
of
that
work
is
again
I,
don't
know
the
full
react.
Graphical
react
code
base,
I
know
that
there's
tests
inside
of
it
and
so
I
I
just
need
to
make
sure
that
when
I
pull
out
the
Explorer
and
the
history
from
graphical
react
into
separate
packages
that
all
of
those
tests
that
might
be
spread
that
are
using
that
somewhere.
Because
again,
it's
all
layered
context.
It's
in.
B
A
One
of
the
one
of
the
advantages
of
having
a
mono
repo
here
is
that
you
can
actually
use
integration
tests
to
provide
unit
test
level
like
coverage
in
the
sense
that
even
in
a
coverage
report,
it
would
work
so,
for
example,
we
do
this
with
graphql
language
service,
so
a
lot
of
the
tests.
So
if
you
look,
the
underlying
graphql
language
service
doesn't
even
have
that
many
of
its
own
tests,
most
of
it,
is
tested
by
our
implementation,
like
our
integration
tests.
A
A
B
The
majority
of
the
code
to
test
the
plug-in
is
is
going
to
be
integration,
because
this
is
other
than
that.
Like
Doc
Explorer,
it's
just
UI
like
it's,
the
only
the
only
you
know,
unit
tests
that
you
need
to
do
is
when
I
click
a
button
that
tells
it
to
cycle
through
the
stack,
because
it's
like,
through
the
stack
when
I
click
exactly
to
remove
something
from
the
stack.
Does
it
remove
like
it's
relatively
these
unit
tests
aren't
even
probably
necessary,
because
it's
it's
hard
to
break
that,
but
so.
A
Here's
here's
here's
a
way,
a
test,
driven
way
that
you
could
do
this
refactor
right
is
that
you
move
you
move
it
to
a
plug-in,
and
then
you
can
still
run
those
integration
level
graphical
spec.tsx
tests
against
it
right.
Then
you
don't
even
have
to
move
the
tests
or
rewrite
that
part
of
the
test,
because
those
tests
are
already
testing
what
you're
going
to
be
doing.
They.
B
A
B
B
This
is
a
lot
easier
and
that
I
Can
Begin
by
by
starting
with
these
two
plugins
and
pulling
them
out
can
begin
to
actually
remove
the
code,
the
context
layer
and
and
like
actually
replace
it
rather
than
just
having
like
a
this
other
store
that
the
component
is
using
and
the
tests
are
using
the
context.
So
this.
C
B
A
Great
yeah,
remember
you
it's
just
yeah
minimal
breakages,
so
the
graphical
workspace
but
break
any
other
apis.
You
need,
let's
get
there
because
any
other
user.
That's
watching.
This
will
say
the
same
thing
like
yeah:
we're
not
worried
about
the
plugin
API.
You
know
the
graphical
react
or
toolkit
apis
having
breaking
changes
right
now,
because
whatever's
on
the
other
side
is
gonna
help
all
of
us.
You
know
and
yeah,
and
it's
zero
zero
XY.
So
it's
you
know
like
yeah.
A
It's
expected
you
could
do
breaking
changes
as
minor
versions
of
the
zero
X
versions
until
you're
ready
to
call
it
stable
as
one
that's
a
that's
a
rule.
That's
in
the
center
spec!
Not
everyone
knows
about
it,
but
it's
something
to
keep
in
mind.
I
don't
know.
Have
we
already
gone
to
a
1x
version
for
graphical
react,
I.
A
See
then,
you
have
infinite
infinite
breaking
capability
within
the
minor
versions
until
you're
ready
to
call
it
one.
It
could
be
two
years
until
it's
1.0
who
cares,
you
know,
just
keep
breaking
it
till
till
you
get
till
you
can
give
the
90
of
our
users
The
Experience
they
want
is.
B
So
the
only
other
thing
that
I
had
to
talk
about
today
was
I
was
hoping
that
Tim
Hall
from
Postman
was
going
to
be
on
the
call,
because
last
month
he
bravely
took
on
the
task
of
attempting
to
replace
code
mirror
with
Monaco
in
graphical
react
bravely
I
say
that,
because
I
tried
it
and
I
I
couldn't
figure
it
out.
Part
of
it
was
because
of
the
the
layering
and
the
context,
and
just
all
the
complexity
there.
It's
also
just
it's
a
it's
a
big
job.
C
B
Know
we
we
don't
have
we
don't
have
like
Monaco
specific.
There
aren't
designs,
yeah.
B
A
That's
why
I
felt
like
it
was
I
regret,
saying
that
we
should
we
can
do
one
or
the
other,
because
it
was
easier
to
do
Monaco
at
the
same
time
as
context
for
us
at
least
like
from
the
much
simpler
original
graphical
one
components
to
just
switch,
those
to
add
Monaco
and
then
and
we
did
tight
scripts.
Then
we
added
we
first,
we
converted
the
components
to
typescript
released
that
and
then
we
converted
them
to
Monaco
and
then
yeah,
because.
B
A
B
A
A
Yes,
Thomas
introduced
a
PR
to
remove
it.
We
can
reintroduce
it
I
guess
we
could
undo
it
remove
rst
example
yeah.
B
A
B
I'm
sure
it's
on
one
of
these
older
branches
it'd
be
easy
enough
to
just
go,
get
it
from
there.
I,
don't
know
that
it
needs
to
live
in
the
repository.
I
was
just
curious
because
when
I
was
early
last
week
when
I
started
this
replacing
the
context
thing
I
wanted
to
reference
that
code,
because
I
was
curious.
B
If,
when
Thomas
was
doing
the
conversion,
if
he
added
optimizations,
if
he
had
added
a
bunch
of
rendering
optimizations
or
if
they
were
already
there,
I
was
curious
because
it
seems
like
it's,
it
seems
like
it's
just
for
me.
It
seems
like
it's
over
optimizing,
it's
optimized
early,
but
that's
because
I
don't
know
what
dear
I
couldn't
I
needed
to
see
what
the
original
code
was.
I
wanted
to
run
it
and
see
if
it
was
like.
Indeed,.
A
The
original
code
was
yeah.
It
was
like
thrashing
context
State
everywhere.
If
you
watched
with
like
the
react
inspector
with
highlighting
on
update,
it
was
like
you.
A
It's
why
it's
part
of
the
reason
I
kind
of
like,
and
just
as
that
was
happening.
It
was
like.
Oh
so
this
is
like
gonna,
be
an
actual
pandemic
and
we
have
to
shelter
its
place.
It
was
all
at
the
same
time,
so
I
was
like
I,
give
up
I'll
come
back
to
this,
and
then
we
found
some
easier
wins
for
the
community
and
for
what
I
was
working
on
at
the
time,
but
it
was
like
yeah
but
yeah,
but
yeah.
A
We
can
yeah
I
just
linked
you
to
the
PR
where
we
removed
it.
So
if
you
want,
we
can
revert
it
and
re-add
it
to
examples.
A
Yeah,
it's
I
I
had
to
search
in
pull
requests
with,
is
open
room
like
filter
removed,
so.
B
Brings
up
another
thing,
and
maybe
this
since
we're
getting
kind
of
late,
we're
getting
kind
of
along
here.
Maybe
this
could
be
the
last
thing
we
talk
about
and
that's
just
cleaning
up
old
stuff
right
in
the
in
the
issues,
particularly
oh
gosh,
yeah,
because
there's
there's
still
pruning
that
can
be
done,
I'm
happy
to
go
through
and.
C
A
I've
been
meaning
to
do
that.
I
have
some
other
projects
that
I'm
side
projects
outside
of
work,
I'm
wrapping
up
right
now,
so
I'm
gonna
be
doing
jumping
in
a
little
bit
more
and
and
also
there's
the
the
graphco
Foundation
grants
we
might
we're
both
applying
for
it.
So
that
can
help
us
too,
but
either
way
soon.
A
I
can
just
jump
in
and
start
doing
more
of
that
I'm
trying
to
just
like
do
a
little
bit
here,
and
there
at
least
have
some
comments
to
help
move
things
along,
but
I
can
like
start
setting
like
a
weekly
limit
like
I
should
be
able
to
close
I
should
be
able
to
close
10
issues
a
week.
There's
no
reason.
I
can't
I
have
so
much
in
context.
B
A
B
A
And
I
think
and
the
guild
is
a
big
fan
of
projects
and
it
works
well
for
them.
What
do
you
think
we
could
use
GitHub
projects?
We
could
use
GitHub
projects
beta.
A
We
can
use
just
a
roadmap
issue,
I'm
fine
with
any
of
those,
but.
B
B
But
for
me
personally,
if
you
from
what
I've
learned
in
the
last
like
nine
months,
there's
a
lot
of
people
that
are
very
passionate
about
graphical,
but
they
come
and
they
go
they
pop
in
and
they
pop
out
as
as
time
allows,
and
some
of
them
are
maintainers
and
some
of
them
are
most
of
them
are
not
maintainers
or
just
people
coming
in
and
going
with
important
critical
opinions
that
we
need
to
pay
attention
to.
B
But
when
we
use
issues
and
discussions
for
for
any
like
for
decision
making
and
sourcing
feedback
that
is
going
to
contribute
to
a
roadmap,
it
seems
like
everybody
is
able
to
people
contribute
like
their
initial
thoughts
and
their
opinions,
but
because
they're
popping
in
and
out
it's
less
of
a
discussion
and
more
of
just
kind
of
like
somebody
trying
to
put
a
pin
in
something
without
actually
putting
a
pin
in
something
so
yeah.
I.
Think
I.
Think
the
structure.
B
So
I
think
I
think
the
projects
is
great.
It
gives
a
lot
of
structure,
but
we're
we
just
we'll
we'll
need
to
kind
of
redo
it
right
reset
the
Baseline
and
then
ensure
that
everybody
knows
that,
like
that's
the
place
where,
where,
where
roadmap
stuff
is
happening
and
and
I'm
guilty
of
of
doing
roadmap,.
B
A
A
You
can
set
up
a
really
nice
automated
bug,
triage
system
with
GitHub
projects,
and
it
make
you
get
like
a
nice
little
dashboard
where
you
can
explore
all
the
bugs
and
whether
they're
confirmed
or
not
like
a
little,
not
not
even
a
dashboard
like
a
kanban
style
board,
almost
at
least
that's
how
the
GitHub
Project's
original
works
and
then
GitHub
projects
beta
is
more
like
jira
like,
but
is
also
kind
of
neat
I.
A
Think
in
fact,
I
think
the
guild
said
that
they
prefer
GitHub
projects
and
not
projects
beta,
which
is
interesting.
A
That's
why
they
I
think
they
still
have
both
enabled
because
I
think
there's
camps
but
I,
like
both
I'd,
be
fine
with
either
I'd,
be
fine
with
road
map
issues,
but
just
consolidating
around
one
for
all
front
end,
such
or
just
all
graphical
and
then
like
cross-linked
to
one
for
Monaco
or
something,
but
like
road
map
issues
with
our
mono
repo
is
really
tricky
to
wrap
your
head
around
for
any
project
management,
because
it's
like
there's
so
many
threads,
there's
like
it's
essentially
One
Mono
repo
with
like
two
dozen
different
products
in
it,
if
you
think
of
it
from
product
technical
product
management
standpoint.
A
To
help
yeah,
if
there's
any
like
they
want
to
help
vote
us
and
folks
from
State
and
everyone
else,
trying
to
wrap
their
head
around
that
if
anyone's
just
like
I'm
the
best
TPM
for
this
job
and
you
are
willing
to
work
for
volunteer
or
grant
money,
come
join
us
yeah.
B
B
Spending
more
time
on
the
actual
graphical
repository,
not
in
my
own,
like
prototype
future
world
space
as
I've
been
getting
more
involved,
I've
had
to
spend
a
lot
more
time,
searching
and
finding
and
looking
inside
of
the
issues
in
the
discussions
and
I
just
noticed
that
there's
a
lot
of
like
there's
a
lot
of
stuff
that
just
needs
to
get
closed
like
it's
ready
to
it.
Just
it's
done
it
just
never.
A
Got
uploaded
you're
right,
it's
done
so
much
of
it
is
stuff.
That's
because
2.0
is
happening.
We
have
these
basic
plugins.
We
can
already
just
say
this
is
done.
This
is
this
feature.
Request
is
supported.
This
bug
has
been
fixed
since
2018.,
so
yeah,
there's
no
reason.
I
can't
do
at
least
10
a
day
or
10
a
week,
rather
with
my
schedule,
and
maybe
just
put
like
a
kind
of
a
reminder
for
that.
A
Maybe
we
could
I
know
Tomas
had
this
great
idea
of
doing
this,
like
a
graphical
triage,
he
had
scheduled
and
we
were
supposed
to
like
go
through
together
and
squash,
bugs
and
I
feel
terrible.
I
didn't
help
with
that,
because
I
probably
would
have
been
able
to
help.
So
much
with
that
and
when.
A
It
was
a
while
ago
he
was
the
idea.
Is
he
wanted
to
do
it
weekly
to
just
kind
of
clear
things
out,
and
it
was
a
it
was
him
and
oh,
my
gosh,
our
colleague
our
other
colleague
from
State
Tim,
of
course,
yeah
Tim,
yeah,
so
yeah.
A
So
we
can
that
we
could
pick
that
back
up
again.
We
could
just
do
it
async
or
just
do
it
in
chat.
You
know
where
we
just
kind
of
had
like
hey
vaguely
like
Thursday
nights
or
something
we
all
try
to
show
up
and
squash
some
bugs
together
or
something
and
Discord
or
something
we
just
pick
a
night.
You
know
or
yeah
yeah
something,
and
then
that
way
you
can
know
that
I'll
be
around
to
answer
some
questions
right
away
or
whatnot,
because
it's
just
a
long
tail.
C
B
Okay
does
would
it
be
helpful
for
so
I
don't
I'm,
just
fearful
of
like
going
through
and
closing
stuff
that
I
think
is
closed.
But
actually,
like
might
not
really
be
like
looks
closed.
But
not
look
not
closed
would
be
helpful
to
go
through
like
like
Pages
at
a
time
and
just
grab
like
issue
numbers
and
paste
them
into
Discord
and
say:
hey
I.
Think
that
I
think
that
these
are
ready
to
be
closed.
Does
every.
A
Hours
remembering
something
someone
did
for
us,
that
was
super
helpful
with
playground
once
and
we
cleared
a
bunch
of
bugs
right
away.
We
made
a
freaking
spreadsheet
with
like
statuses,
for,
but
you
know
what
graphical
prod
or
I'm
sorry
GitHub
projects.
Beta
is
basically
that
if
we
were.
B
The
reason
Ricky
that
projects
might
be
a
little
like
I,
think
people
the
reason
why
people
might
be
apprehensive
about
it
is
because
there's
a
lot
of
stuff
there
already.
That
is
like.
B
A
We
don't
need
a
separate
graphical,
two
tab
anymore,
because
guess
what
it's
launched,
so
it's
all
graphical
too
now
or
maybe
three,
but
we
don't
even
need
to
think
about
that
yet
yeah
yeah.
So
we
could
totally
if
you
want
to
redo
any
of
that,
if
you
want
to
just
say,
hey
I
decided
to
redo
the
GitHub
project,
speed
to
dashboard
like
this,
and
now
we
have
like
a
a
bug
project,
and
then
we
have
the
graphical
develop.
A
B
Wanna
I.
A
Don't
want
to
if
you
have
to
delete
my
well,
actually
save
it,
but
you
could
just
put
it
as
like
archived,
you
know,
or
something
like
put
a
label
on
it
and
then
create
a
new
one
right.
You
could
do
that
if
you
want,
but
because
I
think
I
had
yeah.
There
is
like
a
kind
of
like
little
road
map.
I
captured
there
like
these
are
the
ones
I
want
to
put
in
the
roadmap
but
yeah,
but
I
think
it.
A
It
works
in
a
way
where
you
have
to
you
apply
the
bulk
labels
to
do.
I
can't
remember
even
how
I
set
all
that
up,
but
yeah
feel
free
to
create
a
new
one
that
you
want
to
propose
like
a
new
approach
and
maybe
there's
multiple.
Maybe
you
want
to
have
a
bug,
tracker
project,
that's
separate,
or
maybe
you
just
have
a
tab
for
bugs
whatever
you
want
to
do
yeah,
but
the
tabs
are
cool.
It's
just
like
that's.
You
just
set
a
couple
filters
for
each
Tab
and
yeah.
A
B
A
C
A
Need
to
give
you
from
I
need
to
give
you
permissions
real,
quick.
Let
me
find
my
death.
A
A
I
think
it's
required
yeah
yeah
I'm,
just
remembering
the
old
days.
Okay,
great
I
figured
you
would
cool
yeah,
yeah,
so
yeah
now
you
have
I
think
you
have
permission
for
that.
Maybe
we
can
create
a
different
role
or
add
permissions
to
maintainer
role,
but
maybe
maybe
it's
just
like
create
and
delete
project,
and
you,
you
probably
can
do
everything
else
as
a
maintainer
but
I'll.
B
Make
it
yeah
I'm
going
to
spend
some
time
going
through
that
current,
like
roadmap
project?
That's
in
there
just
to
reset
my
bearings
there
yeah.
A
Yeah
there's.
B
A
lot
of
stuff:
that's
in
there,
that's
good!
It's
just
it's
out
of
context
now,
because
it's
the
it's
a
graphical
two
road
map,
but
like
we're
there,
and
so
we
need
to
sort
of
shift
you've
got
it
filtered
in
such
a
way
that
all
we
need
to
do
is
rename
it
and
like
just
change
the
way
that
we
just
need.
C
B
A
Yeah
because
I
don't
think
I've
updated
this
list
and
like
over
a
year
or
almost
two
I
think
it's.
This
could
be
two
years
old,
so,
like
yeah,
there's,
probably
a
lot
of
new
stuff,
but
this
is
a
good
way
to
start
organizing
the
old
the
ancient
feature
requests
that
we've
probably
already
resolved
or
can
easily
just
pick
off
now
that
we've
made
these
changes
so
well.
B
I
think
also,
there
really
can't
be
outside
of
like
the
three
or
four
big
feature
initiatives
that
we
talked
about
last
month
that
are
sort
of
currently
like
all
in
various
stages
of
being
in
progress.
I
there
isn't
I,
don't
think,
there's
graphical
two
and
the
work
that
Thomas
did
basically
like
closed
a
whole
swath
of
feature.
B
C
A
A
A
B
C
A
I
want
those
like
when
they're
on
the
road
map,
and
you
see
something
like
render
results.
As
you
know,
markdown
render
results
as
Json
5
render
results
as
what,
as
you
know,
whatever
remember,
those
are
all
meant
to
be
closed
for
an
RFC
for
a
results
plug-in
right.
So,
like
you,
don't
like
some
of
those
that
are
on
there
don't
interpret
them
literally,
maybe
read
the
comments
and
maybe
I
suggested
how
it
could
be
plug-inified.
A
But
if
it's
something
that's
like
really
specific
you're
like
that
seems
really
specific
for
the
road
map
like
that,
you
see
maybe
ask
or
even
just
ask
in
the
comments,
if
I,
don't
meant,
say
how
or
I
someone
else
doesn't
it's
like?
How
did
you
imagine
this
as
being
part
of
the
plug-in
API
or
how
did
you
like?
Do
you?
Do
you
see
a
wider
use
case
that
we
can
honor?
That
would
include
this.
A
You
know
or
something
you
know
so
yeah
like,
but
also
if
it's
just
like
this
is
obviously
not
important
or
or
there's
a
better
way
to
do
this
also
yeah,
because
we
didn't
really
know
what
the
plugin
API
was
going
to
look
like
we
just
said
hey,
this
could
probably
be
a
plug-in.
A
potential
plug-in
label
is
the
one
we
use
for
that.
A
So
there's
a
bunch
of
old
ones
that
people
used
to
go
through,
and
that
was
like
one
of
the
triage
things
we
used
to
do
was
just
Mark
things
as
enhancement
or
potential
plug-in,
because
we
were
deciding.
Is
this
something
that
we
can
enhance
in
One
X
or
that
we
can
do
in
a
2X
Plugin
or
back?
Then
it
was
in
1X,
because
one
at
1.0
was
going
to
be
Monaco
and
everything
is
it's
all?
A
B
There's
so
much
there's
so
much
thinking
and
and
there's.
A
C
B
Much
effort
to
you
know
into
the
projects
here
and
at
least
to
label
things
and
to
get
them
sort
of
into
you
know
buckets
that
are
so
close
to
where
they
need
to
go.
I
just
think
that
in
the
last
five
or
six
months
right
since
two
was
was
launched
and
we
all
in
some
ways
began
to
coalesce
around
you
know,
bigger
kind
of
tent
pole
features
that
that
were
just
bigger
than
little
tiny
kind
of
very
specific
things.
B
This
is
all
super
helpful
that
this
is
already
done,
because
it
fits
really
well
into
those
existing
buckets
right
so
like
plug
in
plug-in,
API
or
plug
in
something
you
can
just
say
like
that's
one,
that's
a
project,
and
then
inside
of
that
you
can
filter
for
our
the
plug-in
namespaces
our
plug-in
types
right,
pane,
plug-in
sidebar
plugins.
C
B
C
C
A
B
I'm
gonna
focus
on
because
I
just
got
a
lot
of
momentum
on
this
replacing
context
with
store
thing.
I'm
gonna
hit
that
over
the
next
couple
of
days
and
have
and
get
some
code
in
and
then
I
think,
like
you
suggested,
we
really
need
to
Round
Table
how
we're
gonna
road
map
and
talk
about
feedback
I
do
think
that
an
open
feedback
for
2.0
sticky
discussion
is
a
really
really
good
idea,
because
there
hasn't
been
a
lot
of
feedback
in
Discord
right.
A
A
Smashed
some
right
away
or
squash
them
right
away.
Some
were
there
a
couple
bugs
that
came
in
from
the
initial
flat
2.0
release,
but
then
we
solved
it.
It
was
solid
because
there
was
a
lot
of
people
even
on,
like
CDN
bundle
versions
that
just
you
know
at
next
or
at
latest
or
whatever
yeah
on
the
CDN
so
like.
If
they
were
on
at
latest,
then
they
would
have
gotten.
Then
they
would
have
been
all
of
a
sudden
2.0
overnight.
A
They
wouldn't
have
gotten
any
canaries
right
so,
and
that
was
the
like
in
our
docs
and
everything.
The
CDN
example
uses
unpackage
at
latest
so
yeah
they
would
they
just
they
literally
got
the
2.0
upgrade
like
you
could
see
the
stats
for
it.
It's
there
I
think
some
people
probably
had
to
pin
the
version
because
of
something
or
whatever
they
did
broke,
but
most
people
it
seems
to
be
yeah,
it's
very
smooth,
but
yeah.
A
We
should
get
more
feedback
and
I
think
we'll
get
a
lot
of
feedback
on
plug-in
apis
and
now
I
want
to
do
this
cool
feature
and
that's
fine.
If
that
all
gets
dumped
into
one
discussion,
because
then
we
can
just
each
thread.
That
starts
where
it's
like.
Oh
I,
really
want
this
Plug-In
or
whatever.
We
just
create
a
new
issue
for
it,
or
we
just
tell
them
to
that.
There's
already
an
issue
for
it
and
close
that
discussion.
A
You
know
or
that
little
thread
in
the
bigger
discussion,
but
yeah
yeah
it'll
be
good
yeah
one
place,
and
then
that
way
it
also
creates
a
like
I
think
you
were
just
saying
this
like
it
creates
a
a
place
where
you
can
see
someone
saying
something
that
you're
already
saying
without
having
to
search
in
GitHub
issues.
There's.
B
C
B
A
Neither
do
I,
honestly,
it's
kind
of
Cloudy
like
who
actually
created
graphical
I'm,
pretty
sure
it
was
Lee,
but
it
could
have
been
like
a
colleague
first
created
the
repo
and
copied
his
code
into
it.
I
had
no
idea
or
like
they
were
all
like.
I
saw
a
little
documentary
about
them
and
they
were
all
working
on
it
together.
It's
like
there's,
there's
yeah,
that's
interesting
to
think
about
and
I've
accidentally
kind
of
booked,
some
of
the
git
history,
Through
The
Years
too.
So
some
of
it's
not
even
yeah.
B
A
B
B
A
A
If
you
have
questions,
ask
in
the
Discord,
that's
another
one
that
we
can
say
like
if
it
seems
like
it's
a
custom
case
where
it's
like
they're
doing
something
highly
customized
or
use
case
specific
they're
requesting
then
just
say
this
can
be
done
with
the
SDK.
This
can
be
done
with
the
Plug-In
or
we
plan
on
doing
this
with
a
plugin
you
know
or
whatever
you
know,
yeah
it's
like
even
just
labeling
it
in
a
quick
sentence
is
all
you
need.
A
Okay,
cool
well
I
need
to
go
Ricky.
C
A
C
B
Really
glad
that
that
you
popped
in
gave
me
some
direction
here.
I'm
I'm
I
feel
a
lot.
A
Of
nice
to
see
you
too
yeah
I'm
glad
to
help,
and
always
just
amazed
by
what
the
community
is
able
to
do
to
make
this
project
Thrive,
no
matter
how
much
I
am
involved,
but
I
do
plan
on
yeah
helping
out
more
once
I
wrap
up
some
things,
but.
B
Cool
well
I
hope
to
get
some
code
in
the
next
couple
of
days
and
start
to
give
some
folks
some
ideas
of
of
what
replacing
context
might
look
like:
okay
and
I'm
gonna
I'm
gonna
get
to
it.
I
appreciate
everything.
I
really
do.
Okay,.
A
B
C
B
Sure,
though,
well
I
mean
you're,
I
I
know
you're
correct
and
that
there's
a
lot
of
people
in
the
community
this
community,
specifically
that
are
paying
attention
to
it,
and
it's
a
it's
a
great
it's
a
great
place
for
it.
I
think
I've,
I've,
I've
pulled
back
from
the
Prototype
over
the
last
like
couple
of
weeks
and
I
haven't
really
spent
a
lot
of
time
in
it.
I
think
we
discussed
I
I
want
to
get
the
figma
file,
just
cleaned,
which
I
haven't
done
yet
there's
a
lot
of
stuff.
B
That's
in
there
that
just
needs
to
get
it
just
needs
to
get
explained
better
inside
the
figma
file
and
then
I
think
in
terms
of
like
the
Prototype.
The
live
prototype.
That's
running
right
now,
like
I,
think
it's
in
a
good
place,
but
before
we
start
really
hitting
the
community
for
feedback
like
you've
been
suggesting,
I
really
want
to
get
the
figma
file
ready
to
go
so
that
I
that
there
can
be
a
lot
of
explaining
done
there.
So
it's
it's
on
my
list.
I
just
I
just
haven't
had
time
to
get
to
it.
B
So
I
do
appreciate
you
suggesting
that
we
use
Twitter
and
I
want
to
do
that.
I
just
also
want
to
make
sure
that
I'm
not
I'm,
not
preempting,
discussion
that
that
isn't
ready
that
isn't
ready
to
be
had
yet
I
mean
there's
already
enough
to
talk
about
the
visual
query,
Builder
and
there's
enough
out
there
for
people
to
have
opinions
and
I
need
to
get
the
figma
file
done.
First,
basically,.
A
Okay,
yeah!
Well,
thanks
for
doing
all
this,
this
is
also
important
and
I'm
kind
of
useless
when
it
comes
to
figma
files.
So
I'm
glad
you're
here
and.
B
Well,
this
one,
the
one
that
the
pigment
file
that
I've
been
working
on
is
I'll,
have
notes
and
stuff
in
it
for
how
to
use
it.
It's
actually
the
way
that
I've
got
it
set
up
with
like
a
light
theme
and
a
dark
theme.
It's
really
simple
to
just
kind
of
go
between
the
two.
It's
I
think
it's
I
think
it's
easier
to
use,
because
it's
super
well
organized
than
a
lot
of
other
figure
files,
so
I
hope
that
it
encourages
everybody
to
like
get
in
and
actually
okay
play
with
it.
A
Great
awesome
and
what
else
was
I
going
to
say:
okay
yeah
for
anyone
who's
who's
listening?
A
If
we
would
love
it
and
you
don't
have
to
well
I
mean
yeah
what
was
I
gonna
say:
oh
for
anyone
who's
listening
little,
we
I
think
we
both
have
GitHub
sponsors
set
up.
I
do
do
you
John
I,.
A
Good
okay,
well
safety
at
safety
John
here
our
friend
Jonathan
awesome,
we'll
we'll
be
setting
up
his
GitHub
sponsors
soon
and
will
also
gladly
accept
donations,
but
also,
if
you're,
representing
a
company
and
you're,
not
a
member
of
the
graphql
foundation.
Yet.
But
you
would
like
to
sponsor
this
work
in
a
in
a
like
a
big
way.
A
A
big
officious
way
then
contact
the
graphql
foundation,
because,
where
there's
going
to
be
a
grant
program
for
us
and
for
anyone
in
the
community
who
wants
to
submit
a
grant
proposal
to
improve
our
tooling
or
other
Tooling
in
a
graphicality
business,
so
yeah
consider
having
your
your
company
become
a
member,
but
also,
if
you
want
to
give
me
a
few
G's
and
John
a
few
G's
to
keep
keep
the
fires
burning.
Let
us
know
yeah.
B
A
Community
and
the
hundreds
of
people
that
will
watch
this
video
reach
out
and
Discord
just
discussions
issues
whatever
yeah
love.
You
all
bye.