►
From YouTube: JupyterLab Team Meeting - March 30, 2022
Description
A meeting to share and discuss features, ideas, issues, and pull requests in JupyterLab and other Jupyter frontends. This meeting is open to anyone and everyone.
Join future calls via the Jupyter community calendar: https://docs.jupyter.org/en/latest/community/content-community.html#jupyter-community-meetings
Notes for upcoming meetings can be found on the agenda: https://hackmd.io/Y7fBMQPSQ1C08SDGI-fwtg
Past notes can be found on the JupyterLab team compass: https://github.com/jupyterlab/team-compass/issues?q=is%3Aissue+label%3A%22Dev+Meeting+Minutes%22+
A
Okay,
great
hi,
everyone
welcome
to
the
march
30
jupiter
lab
weekly
call.
If
you
haven't
been
here
before,
looks
like
we
have
20
people,
so
maybe
you
all
have,
but
just
in
case
in
the
chat
I
pasted
the
meeting
minutes.
A
Please
add
your
name
to
the
attendees
if
you
would
like
and
also
if
there
is
something
you'd
like
to
talk
about
for
roughly
five
minutes
or
so
add
a
bullet
point
to
the
agenda
and
if
there's
something
that
might
take
a
bit
longer
add
a
bullet
point
to
the
additional
discussion
section
at
the
bottom
and
as
always
at
the
end
of
the
call,
there
will
be
some
time
to
talk
about
things
without
recording
so
I'll
give
a
few
minutes
for
that.
But
for
now
let
us
get
started.
B
B
D
B
C
Yeah,
frederick
himself
has
given
feedback
on
the
topmost
one
12028,
the
cell
toolbar,
which
I
would
say
is
is
close
to
being
done.
There's
still
some
additional
changes.
I
need
to
make
to
address
all
of
his
stuff,
but
I'm
optimistic
that
I
can
get
into
like
a
releasable
state
by
the
end
of
this
week.
B
And
for
the
others
they
are
almost
done.
I
mean
this
one
for
the
terminal
in
the
launcher.
It
was
waiting
for
a
release
of
jupiter
server,
and
now
it's
done
so
it
should
be
okay,
I
reviewed
the
one
of
eric
of
forexing,
the
virtual
notebook,
and
you
should
be
good
to
go.
There
is
still
one
one
thing
that
need
to
be
looked
for,
but
I
think
should
be
easy
to
figure
out.
Hopefully
that
one
is
ready
and
the
one
from
marta
is
ready.
I
think
I
even
yeah
we
are.
B
We
have
both
approved
it
mike
and
me.
I
did.
I
just
waited
like
I'm
waiting
a
couple
of
days
before
in
case
somebody
wants
to
to
say
something
and
I'll
review
the
latest.
One
that's
from
eric,
too,
is
look.
Fine,
maybe
like
isabella
may
have
a
say
about
the
design.
So
the
idea
of
this
one
is
to
have
a
link
to
jump
directly
to
the
preferred
directory,
and
so
the
pr
by
eric
is
to
propose
having
a
button
instead
of
just
a
display,
slash
next
to
the
folder
icon.
B
So
yeah,
the
question
is
mainly:
are
we
target
trying
to
target
something
close
or
not?
I
I
like
the
proposal
of
alex
to
say:
okay,
let's
try
to
to
say
we
freeze
that
list
and
or
maybe
wait
for
the
end
of
this
week
for
to
make
the
final
list
and
then
say:
okay,
when
that
list
is
finished,
we
do
the
release.
D
I
like
the
idea
of
that
picking
a
freeze
date
to
the
list
and
then
just
releasing
when
the
list
is
done.
Are
people
comfortable
with
friday
being
the
freeze
date
for
that?
Do
we
want
to
make
it
this
meeting
next
week,
I'm
good
with
it
there.
I
don't
have
anything.
I
don't
have
any
skin
in
the
game.
The
ones
I
have
skin
in
the
game
are
already
merged.
B
So
yeah
then
so
I
already
mentioned
that
a
couple
of
pr's
are
in
need
of
review.
Two
of
them
are
listed
above
and
are
waiting
for
review
for
like
three
weeks
now,
so
it
would
be
nice
if
somebody
can
have
a
look
at
that.
I
get
a
question
too.
It's
about
icons
in
recent
pr.
Some
people
have
added
icons,
in
extension,
like
a
json
edit
icons
in
the
cell,
toolbar
extension.
C
D
D
A
Well,
two
things
one
is:
it
would
be
good
to
add
to
the
core
jupiter
lab
docs
for
the
contributing
section,
but
not
for
extension,
authoring
docs,
because
we're
not
talking
about
external
extensions,
but
the
the
question
I
was
going
to
ask
is
it's
immaterial
to
whether
it's
in
one
package
or
another,
whether
it
respects
theming
right
or
is
that
actually
one
of
the
rationales
for
having
all
of
them
in
one
place.
C
Yeah
theming
is
more
a
matter
of
using
the
proper
classes
like
if
you,
if
you
apply
the
appropriate
classes,
to
path
elements
which
is
something
I
had
to
do
manually
then,
when
you
switch
to
dark
themes,
you
will
have
higher
contrast,
icons.
A
And
yeah
cool,
I
think
that's
that
that's
a
compelling
reason
to
ship
them
that
way,
because
then
it
gives
other
extension
offers
the
ability
to
write
extensions
that
use
the
same
ui
language
and
I
think
that's
really
good.
B
Cool
regret
anything
else
just
to
mention.
There
has
been
a
report
like
for
a
couple
of
weeks
or
even
month
or
of
some
asynchro
bug.
So
it
seems
that
the
bug
is
happening
only
for
when
users
start
jupiter
up
with
some
notebook
already
open
that
open
at
startup,
and
it
seems
that
it's
more
concerning
for
a
windows
user
than
another
operating
system,
and
so
the
bug
seems
very
tricky
to
figure
out,
and
so,
if
somebody
have
high
knowledge
about
us,
thank
you
and
is
willing
to
have
a
look
and
provide
some
guidance.
B
The
pr
in
which
there
are
the
there
is
the
letter
of
intent
so
feel
free
to
to
comment
on
that,
and
also
the
issue
on
the
jupiter
governance
repository
asking
for
an
approval
from
the
steering
committee,
and
that's
all
for
me.
Thanks.
D
G
So
I
recently
submitted
a
pr
to
add
check
boxes
to
the
file
browser
and
this
work
got
me
well.
It
raised
questions
about
the
kernel
status
indicator,
which
is
this
little
green
dot
that
appears
next
to
notebook
files,
as
I
understand
it
in
the
file
browser.
So
last
week
I
was
like
brainstorming
with
isabella
on
what
to
do
about
this,
and
in
order
not
to
hold
up
my
pr
forever,
I
sort
of
just
moved.
You
know,
shifted
margins
and
paddings
around
and
column
widths
and
arrived
at
something.
G
But
isabella
had
raised
this
issue
that,
like
that
status
indicator,
doesn't
apply
to
every
file
or
folder
in
the
file
browser.
So
maybe
it
shouldn't
have
its
own
column,
like
it's
sort
of
like
dedicated
column,
just
to
that
kernel
status
indicator.
So
she
had
like
an
idea
of,
like
maybe
overlaying
it
on
top
of
the
file
icon.
I
put
a
screenshot
of
that
in
the
hackmd
and
then
you
can
also
see
below
that
a
screenshot
showing
how
it
looks
with
the
check
boxes,
pr
that
I
currently
have
up.
G
So
I
just
basically
two
questions
for
this
group
is
like
how
did
like
what
do
people
like
or
like
you
know
somebody
some
thoughts
on
these
proposals,
and
then
you
know
if
we
do
like
isabella's
idea.
I
was
wondering
about
like
if
we
should
put
pause
on
the
pr
that
I
currently
have
up
in
order
to
reduce
any
potential
like
ui
churn
doesn't
matter.
So
those
are
the
two
questions
for
the
group.
G
D
D
Can
be
on
the
same
page
share
with
screen
the
the
screenshots
are
in
the
hacking
yeah
yeah.
So
I
actually
had
no
idea
that
those
were
kernel
status
indicators,
the
little
green
buttons.
I
just
thought
they
were
like
open
or
active
files,
or
something
so
I
mean.
Maybe
you
could
put
like
a
play
button
there.
That
shows
like
it's
being
played,
because
the
semantics
weren't
clear
to
me-
and
I
use
it
every
single
day.
C
Yeah
when
I
see
a
dot
next
to
a
file
name,
I
usually
think
that
it's
been
modified
in
some
way,
because
that's
the
mac
os
convention
and
that,
if
you
save
that
dot
will
disappear.
G
D
You
are,
you
are
correct
in
what
it
does
that.
That
is
what
the
indicator
is
for,
but
I
also
did
not
know
what
it
was
for
the
first
nine
months.
I
worked
on
this
project,
so
you
are
also
not
crazy
for
not
knowing
that.
F
I
don't
actually
know
like
we
were
just
going
back
and
forth
like
there
could
be
some
more
polish
here.
My
thought
was
just
we
have
like
a
really
wide
space
on
any
file
that
isn't
a
notebook
essentially
like
to
the
left
of
it
so
and
now
with
the
then,
it
becomes
really
large
with
these
check
boxes,
so
that
was
basically
the
logic,
but
I
don't
think
this
is
like
the
only
right
solution.
I
just
want
to
say.
D
E
F
I
To
the
file,
but
it
just
displays
the
file
name
and
the
full
path
is
on
the
hover
mouse
over
hover
of
it.
But
it
is
probably
more
useful
if
most
people
just
use
the
file
browser
for
most
of
the
time
right,
and
so
it
is.
There
is
value
to
having
the
little
dot
in
the
file
browser,
but
there
might
be
an
argue
of
how
much
quality
it.
E
Offers
of
course
it
is
of
value
to
myself
when
I
have
say
20,
notebooks
open
and
this
happens.
I
do
not
see
them
all
and
having
an
extension
file
browser
which
allows
me
to
see
notebooks
in
separate
directories.
E
So
the
three
photo,
I
don't
remember,
which
extension
unfold.
This
is
very
handy
to
see
which
notebooks
are
up
and
running
and
taking
space
in
the
memory,
because
kernel
is
running.
I
Yeah
right,
that's
the
the
running
terminals
and
kernel
thing.
Of
course
it
doesn't
have
the
file
tree
right,
so
it's
that
at
least
definitely
shows
everything
that
is
open
on
the
server
as
a
resource,
well,
yeah
that
that
that
one
also
could
probably
have
a
better
ui
and
if
we
can
combine
them
more
reliably
in
one
view,
maybe
you
could
get
rid
of
that
separate
thing
as
a
tab.
I
don't
know.
A
G
I
think
I
have
some
ideas
I'll
bounce
around
with
and
maybe
bring
some
of
this
chat
to
the
getter
channel
or
something
the
the
one
thing
I
don't
have
direction
or
clarity
on
just
this
is
just
like
a
newbie
question
is,
since
I
do
have
a
pr
up
right
now,
like
is,
should
I
try
to
get
that
through
or
should
I
hold
off
on,
like
figuring
this
kernel
status,
stuff
out.
A
Well,
I
don't.
I
don't
know
what
the
I
guess.
The
question
is:
are
we
going
to
release
a
version
that
has
a
ui
thing
that
disappears
in
the
next
version
or
not
because
if
not
cool
great
like
you
can
have
different,
you
can
have
the
the
actual
branch
contain
an
iteration
upon
a
ui
element
pre-shipping.
We
just
have
to
make
sure
that
it's
targeting
a
milestone
that
doesn't
ship
until
we
have
consensus
on
what
ui
element
does,
but,
like
your
branch,
just
straying
farther
and
farther
is
probably
not
great
and
on
net.
A
This
is
a
positive,
so
I
I
wouldn't
think
you
should
wait
forever.
I
think
it
might
be
better
to
go
in
once
in
once,
like
the
core
functionality
is
ready
and
we
can
treat
this
as
a
piece
that
still
needs
some
ux
iteration,
but,
like
the
core
functionality
of
your
pr,
presumably
is
ready
for
review
right.
Yes,
okay,
I
don't
know.
Does
anyone
disagree
with
that.
D
J
Flagged,
whether
you
have
this
checkbox
disappear,
no
flag,
currently,
all
right.
It
might
come
up
in
review
to
add
it
right,
because
if
it
was
flagged
so
especially
if
it
was.
K
Over
in
the
in
the
chat
we've
been
talking
about
the
the
the
columns
themselves,
that
could
be
data
driven,
and
so
you
know
the
status
quo
released
for
a
3.3
or
whatever
would
be
the
existing
ones
so
status
whatever
it
is
today,
file
name
file
modified.
K
There
would
be
another
column
available
which
was
check
box
or
something
like
that
and
you'd
be
able
to
set
the
ranks
so
that
you
could
order
how
big
they
were
and
all
that
kind
of
stuff.
K
L
But
I
mean
if
it
just
gets
in
it's
still
in
the
4.0
like
slash
master
branch,
so
they
get
the
the
bar
is
pretty,
is
pretty
low
right
so
and
they
use.
It
would
be
like
really
useful.
Also
because
the
the
top
comment
of
this
pull
request
links
to
an
issue
in
retro
lab,
and
if
we
have
it
already
in
in
a
folder
or
priorities,
it
would
make
it
much
easier
to
to
integrate
it
in
notebook
seven
and
we
could
already
already
try
there.
I
Yeah,
so
if
it's,
if
it
also
has
so
if
this
is
very
useful
for
retro
lab,
you
could
also
have
this
as
a
option
for
the
class
that
you
don't
have
necessarily
a
configure
for,
but
that
is
this
yeah.
I
don't
know
enough
enough
about
how
fracture
lab
works,
but
you
can
change
the
config
of
the
different
extensions.
L
I
think
it
could
be.
It
should
eventually
be
configurable
for
sure,
but
it's
probably
not
a
blocker
right
now
right
now,
but
before
the
the
release,
the
final
release-
yes.
F
I
would
also
say
from
my
understanding
what
gabriel
has
done
is
just
added
a
ui
like
visually.
It's
just
adding
the
element
of
the
check
box
and
like
adjusting
the
padding.
I
don't
think
is
that
drastic,
like?
Maybe
I'm
I'm
wrong
here,
but
I
don't
think
this
would
be
a
horrible
thing
to
jump
from,
because
I
think
what
it
looks
like
right
now
to
me
is
we've
added
something
and
then,
if
we
like
end
up
releasing
it
again
and
it's
like
oh,
we
combined
two
things.
F
A
E
E
Also
very
specifically
to
ask
whether
you
think
that
the
minimum
width
of
the
tap
should
be
stored
as
a
setting,
as
well
as
some
parameters
of
the
screwing,
smoothness
and
screwing
animation
speed,
and
how
to
make
that
happen,
because
I
don't
think
that
we
stole
things
which
in
settings
which
currently
are
could
be
defined
as
css.
So
then,
if
we
do
that,
as
a
setting
user
can
change
those
things,
but
then
teams
cannot
override
them
and
really
much
look
for
feedback
here.
D
I
was
taking
a
look
while
we've
been
in
the
meeting,
because
I
saw
that
ahead
of
us
and
it's
a
very
interesting
feature,
I'm
on
the
fence
about
whether
or
not
we
want
the
scrolling,
because
it
is.
D
I've
seen
some
ids
and
web
browsers
just
keep
squishing
the
tabs
until
they
all
fit,
and
then
others
do
this
scrolling
and
I'm
not
sure
which
one's
the
better
experience
to
the
point,
where
I'd
almost
say
that
it
might
be
better
to
make
this
ui
configurable
from
the
offset
make
it
a
configurable
setting
whether
or
not
you
even
have
it
enabled,
in
which
case
we,
if
you
do
have
that
enabled
you
could
toy
with
how
to
handle
the
minimum
width
as
a
setting
for
people
who
do
enable
it.
D
I
think
there
are
ways
to
set
minimum
width,
but
not
let
them
touch
the
css
directly.
So
you
don't
you
don't
let
them
set
like
the
minimum
css
width,
but
you
like,
for
instance,
you
could
give
them
an
enum
like
medium
small,
wide
or
something
and
they
choose
one
and
that
gets
set
to
some
preset
one
that
could
still
be
overridden
by
a
by
a
third-party
css
if
they
really
want
to.
A
I
think
I
agree
with
alex,
because
this
is
lumino
and
not
jupiter
lab
that
ought
to
be
configurable
to
be
a
generic
component.
You
can
use
a
lot
of
different
kinds
of
applications,
so
I
don't
know
what
the
like
nick
just
suggested,
something
in
the
chat
about
the
placement
of
the
buttons,
and
so
I
think
there
are
might
be
questions
on
whether
you
want
them
straddling
like
book
bookending
or
whether
you
want
them
grouped
at
the
end.
A
I
think
there
might
be
use
cases
where
you
just
want
them
to
get
small
the
way
they
currently
do.
I
think
we
should
probably
leave
a
default,
which
is
the
current
behavior,
but
and
and
hide
this
not
hide
this
add
this
behind
an
option
and
then
jupiter
lab
can
make
opinionated
choices
about
what
the
default
for
jupyter
lab
is.
The
default
for
luminal
probably
should
be
the
former
behavior,
and
this
new
behavior
there's
an
option.
A
The
default
for
lab
can
be
to
enable
this
option
like
we
can
make
that
decision
on
the
lab
front-end,
but
not
impose
it
on
other
consumers
of
the
same
components
in
other
applications
so
yeah.
I
I
echo
what
alex
is
saying.
D
And,
and
to
kind
of
expand
on
that,
that's
how
the
plus
button
works.
The
reason
I
was
able
to
add
the
plus
button
to
jupiter
so
easy
is
somebody
had
already
added
that
to
lumino
and
they
made
it.
So
you
could
just
you
enable
it.
Then
you
tie
it
to
a
signal
and
then
you
can
tinker
with
the
css
to
like
display
where
it
is
how
it
behaves,
and
that
was
how
I
enabled
it
in
jupiter
lab.
D
So
if
we
add
this
functionality
in
lumino
the
same
with
the
same
methodology
yeah
even
with
like,
where
the
left
right
buttons
go,
if
you
make
that
config
make
that
in
a
way
that
you
can
change
it
via
css,
like
how
I
was
able
to
float
the
plus
button
to
where
I
wanted
it
to
be,
with
some
tinkering.
E
You
can
tweak
the
settings
and
css
and
you
can
disable
that
all
together
in
drupal
app
as
well,
I
I
was
just
rather
asking
whether
on
the
jupiter
lab
level,
we
want
to
have
those
us
as
settings
which
are
exposed
to
the
user
or
just
make
of
innovative
choices.
H
Hey
everyone.
I
just
saw
a
quick
update
on
the
desktop
application.
I
released
a
new
version
and
this
version
comes
with
some
major
changes
to
the
installation
process,
and
now
the
bundled
python
environment
is
installed
on
demand
instead
of
automatically
installing
it
every
during
every
installation.
H
H
This
is
a
much
smoother
drag-and-drop
installation
process
and
these
changes
are
also
paving
the
path
for
automatic
updates,
which
I
would
be
working
on
soon
and
also
another
change
here-
is
the
mac
application
is
now
code
signed
and
authorized.
H
A
That
is
awesome
the
package
signing
and
stuff.
Does
that
mean
that
it's
on
the
mac
app
store,
or
is
it
still
you
download
it
from
github.
H
You
still
download
it
from
github
yeah,
but
previously
it
would
show
you
like
a
warning
message
that
this
app
could
have
security
issues,
but
now
it
will
just
say
that
this
is
downloaded
from
internet,
but
apple
checked
and
verified
the
application
yeah
that
that
would
be
a
different
message
shown
this
time.
But
the
other
thing
is:
if
you
don't
code
sign
the
application,
you
cannot
provide
automatic
updates.
D
But
quick
question
about
that,
while
it's
there,
I
can't
remember
if
this
was
answered
previously
are,
is
the
standalone
app
able
to
install
front-end
and
or
back-end
extensions
like
third-party
ones,
yet
or
because
I
want
to
say
we
were
able
to
do
front-end,
but
can
we
install
like
server
extensions
as
part
of
that
or
is
it
not
support
those?
As
of
now.
H
I
am
not
hundreds
of
percent
sure
about
server
extensions,
but
you
can
definitely
install
pre-built
from.
D
H
D
Yeah,
just
a
quick
update
for
everybody
since
there's
a
bunch
of
us
who
submitted
talks
and
tutorials
to
sci-fi,
we
did
hear
back.
Our
tutorial
is
on
standby,
so
if
another
tutorial
is
dropped
due
to
one
reason
or
another,
we
could
be
picked
up
as
a
tutorial
for
sci-fi
this
summer,
and
that
will
be
a
developing
extensions
tutorial.
D
So
no
news
yet
on
our
what's
new
jupiter
lab
talk,
which
I'm
taking
as
good
news
since
we've
now
gotten
a
tentative
tutorial.
That
means
we
at
least
haven't
been
rejected.
Yet.
A
We
don't
have
any
other
agenda
items.
We
have
additional
discussion
continuing
from
where
we
left
off
last
week.
Talking
about
web
components.
A
A
Frederick,
why
don't
you
reacquaint
us
with
this?
Please.
B
Yeah
sure-
and
so
I
started
a
repository
with
a
toolkit
of
components
that
are
based
on
a
custom
component,
that's
a
standard,
but
they
are
not
used
directly.
So
it's
not
vanilla
gs.
They
are
using
a
library
that
is
called
fast
from
microsoft.
B
B
System
for
designing
the
design
like
coloring
and
to
enforce
rules
for
contrast
and
stuff
like
that.
So
that's
why
I
I
went
for
that
one.
B
B
We
have
been
doing
also
some
trial
because
one
of
the
trouble
with
custom
elements
in
the
standard
artist
that
you
can
only
define
once
the
same
new
tag
and
that
may
bring
trouble
in
the
sense
that
if
you
have
two
different
library
or
the
same
library
but
two
different
version
trying
to
to
bind
to
the
same
name
you
you
can
have
weird
issues.
B
So
far.
We
didn't
encounter
that
in
jupiter
lab
and
mainly
it
was
due
to
webpack,
because
the
the
library
where,
due
to
federation,
where
just
one
version
was
picked,
the
only
thing
is
like.
If
you
don't
specify
like
which
one
you
want
to
pick,
you
don't
control
exactly
which
one
is
going
to
be
picked.
B
It's
going
to
be
the
first
one
phone,
let's
say,
but
I
mean
if
it
makes
it
to
the
core,
it
will
be
controlled
by
us
a
bit
like
react
or
good
mirror,
so
that
we'll
be
safe
on
that
front,
and
I
think
the
the
main
question
darian
asked
last
last
week
was
about
is
microsoft,
library,
droppable
or
not,
so
the
answer
is
is
not
droppable.
B
So
if
we
go
that
path,
we
are
like
committing
to
use
the
the
library
we
could
probably
change
part
of
the
like
the
the
templates
that
are
providing
for
creating
the
the
components
stuff
like
that.
It's
probably
easy
to
switch
to
another
library
like
lit
from
google,
but
then
there
is
the
other
part
like
the
design
system.
A
So
someone
who
thinks
about
these
things
is
brian,
and
I
see
him
on
the
call.
I
wonder
what
your
thoughts
are.
We
were
talking
in
a
recent
governance.
Call
about
creating
a
group
that
handles
design
language,
because
that
isn't
like
an
official
effort
in
jupiter
and
different
projects
do
different
things,
but
what
you're
describing
takes
us
part
of
the
way
there
and
yeah?
I
don't
know
brian,
do
you
have
any
thoughts.
A
A
F
A
This
seems
like
it's
one
of
the
building
blocks
that
would
help
make
a
richer
extension
playground
as
well
for
you
to
drag
and
drop
and
compose
extensions
without
writing
javascript,
and
I
wonder
what
your
thoughts
are
on
that
like.
If
we
does
this
get
us
any
closer
to
the
dream
of
an
extension
builder
studio
inside
jupiter
lab
itself.
B
Depends
what
you
expect
from
that
builder,
because
at
some
point
somebody
is
still
to
bring
the
the
the
intelligence
in
it.
So
that
means
we
need
to
provide
higher
component
than
just
checkbox
and
button.
Probably
what
you
will
get
for
free
is
that
the
the
element
will
style
according
to
jupiter
lap
team,
but
you
could
have
that
with
other
kind
of
technology
too.
B
I
mean
it's
the
main
difference
maybe
compared
to
something
like
react,
although
no
there
are
solutions,
it's
like
you.
We
could
imagine
that
that
in
the
long
term,
we'll
create
higher
component
like
a
cell
or
something
like
that,
and
it
will
be
really
easy
in
your
ginger
template
to
just
say:
okay
here
I
want
a
gp
cell
and
that's
all,
and
you
will
have
just
a
line
saying
I'm
importing
that
library
of
component
and
that
will
instantiate
where
you
put
the
tag.
The
cell.
B
A
I
That's
it
just
since
nobody
else
seems
to
be
fighting
for
the
mic.
This.
I
I
B
A
Okay,
I
don't
feel
like
we
are
arriving
at
something
approaching
consensus
on
this,
but
it
also
feels
like
lots
of
people
haven't,
spent
that
much
time
thinking
about
the
issue,
so
I
don't
think
there's
disagreement
either.
I
just
think
there
aren't
that
many
well-formed
opinions
right
now.
What
would
you
like
to
gain
out
of
this
conversation?
A
B
Is
do
we
think
it's
the
right
path
forward
so
that
we
move,
or
I
move
slowly
by
proposing
pr
of
replacing
core
components
by
component
from
that
library?
That's
basically
the
path
I
hope
to
take.
A
Okay,
so
I
think
that
this
should
be
one
of
the
first
questions
we
put
to
the
jupiter
lab
council,
because
this
is
a
serious
decision.
We
don't
appear
to
be
approaching
consensus,
but
I
also
don't
think
we
have
enough
people
in
the
jupiter
lab
council
yet
for
that
vote
to
feel
particularly
binding.
So
what
I
might
propose
is
we
bring
this
up
again
in
another
couple
discussions
we
get
more
people
involved
and
try
to
get
either
a
sort
of
consensus
within
the
one
of
these
conversations
where
everyone's
like
cool.
A
Let's
do
it
or
somebody
presents
a
slam
dunk
case
for
why
we
don't
and
barring
that
we
try
to
we,
we
use
the
governance
mechanisms
and
we
put
it
to
a
vote
except
we
need
a
larger
council.
I
think
our
current
council
is
like
so,
if
I
I
don't
know
this
draft
pr
and
team
compass,
I
don't
remember
what
the
last
count
is,
but
whatever
it
is,
it's
not
enough
right.
So
we
grow
that
a
bit
and
if
we
have
by
the
time
we've
grown
that
enough
that
we
feel
like
it's
representative.
D
But
in
parallel
to
that,
federique
is
this
something
that
you
were
hoping
to
get
into
4.0,
because
if
this
is
a
post
4.0
pitch,
I
feel
like
taking
that
time.
That
darien
was
talking
about
is
good.
D
Wait,
wait
for
the
governance,
wait
for
because
I'm
on
the
fence
about
this,
like
I
don't
have
a
strong
enough
opinion
to
decide
how
I
feel
about
it
yet,
and
so,
if
this
isn't
something
that
we're
like
really
pushing
for
4.0
I'd,
say
if
you're
at
a
point
where
it's
ready
to
go
or
not
to
go
to
table
it
until
yeah
until
we
can
make
that
decision
and
aim
for
it
for
five
or
four
four
dot
x.
B
Yeah,
actually,
it's
not
going
to
be
in
a
breaking
change
because,
basically,
is
instead
of
saying
I'm
taking
a
stock
button
or
a
button
from
from
some
react
library.
I
will
use
the
button
from
that
component.
So,
basically
the
interface,
the
look
of
it
can
change
a
bit,
but
everything
that
is
about
the
api
and
what
you
are
doing
with
the
component
will
be
the
same,
because
I'm
definitely
not
speaking
about
replacing
the
component
that,
like
I'm,
creating
a
cell
component
here,
I
just
want
to
go
small
step
like
so.
A
Okay,
so
that's
really
good.
That
means
that
that
actually
helps
yeah.
So
there's
no
immediate
pressure
to
like
arrive
at
a
quick
decision.
On
the
other
hand,
you
don't
want
a
bunch
of
work,
you've
done
to
sit
around
forever
either.
So
it's
just
more
that
we
now
understand
the
sort
of
deadline.
Type
constraints
are
less
pressing,
so
yeah
I
mean,
I
think,
maybe
oh
martha's,
on
the
call.
Maybe
you
can
talk
a
bit
with
martha
about
how
this
would
work
in
the
setting
editor.
A
I
think
maybe
tagging
isabella
and
brian
to
talk
about
design
language
like
basically
have
a
few
sort
of
smaller
group
conversations
about
this
to
see
what
the
shape
of
moving
in
this
direction
would
actually
look
like
and
whether
we
have
enough
buying
to
do
it
and
let's
revisit
this
in
the
next
few
calls,
I
would
say.
I
One
question
today,
I
think,
could
be
useful
is
the
api
that
would
be
exposed
to
users
so
not
their
api
again.
You
said
that
stable,
that's
good,
but
if
somebody
wanted
to
start
using
these
components,
how
much
sure
or
how
final
do
you
think
those
apis
are
how
stable
do
you
think
they
are
if
we
wanted
to
make
changes
like,
for
example,
I
don't
know
now
if
there
is
a
way
to
express
kind
of
how
dense
you
want
thing
to
be.
I
So
I
don't
that's,
probably
not
there
if
immediately,
but
if
somebody
wanted
to
add
feature
like
that
like
how
much
would
the
apis
change
or
would
they
be
stable,
and
that
would
be
something
you
added
behind
the
scenes.
So
if,
if
you
believe
that
the
apis
are
stable,
it's
a
lot
less
problematic
to
to
to
go
early
right.
So
those
those
kind
of
questions
are
can
be
if
things
are
stable.
That's
a
good
argument
for
to
make
to
ensure
people
are
comfortable
with
this
change.