►
From YouTube: JupyterLab Team Meeting - 12 April 2023
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/170
A
Hello
and
welcome
to
the
April
12th
Jupiter
lab
weekly
call.
Today
we
have
14
people
on
the
call
and
we
have
a
few
items
on
the
agenda
already.
At
the
same
time,
we
are
also
testing
the
real-time
collaboration
of
Jupiter
lab
for
beta.
One
and
Jason
has
shared
a
link
in
the
chat,
just
in
case
you
just
joined.
A
Maybe
we
should
paste
that
in
again
so
that
you
can
try
out
that
session
with
everyone
else
and
after
the
agenda
we
can
talk
through
what
things
we
noticed
if
there
are
things
and
get
some
feedback
on
the
RTC
experience,
and
also
there
appears
to
be
a
section
at
the
bottom
where
people
are
writing
notes.
So
that's
really
helpful
in
the
additional
discussion
section,
but
for
now
why
don't
we
get
started?
The
first
person
on
today's
agenda
is
Fred.
C
Yeah
I'm
really
interested
in
the
RTC
stuff
I
think
it's
really
cool.
It's
a
great
work
that
you
guys
are
doing.
My
question
is
where,
where
can
I,
when
I'm
running
Jupiter
lab
locally
with
the
collaborative
slag
on?
C
Is
there
a
database
or
a
cloud
file
where,
like
the,
why
certainty
State
vectors
are
being
written?
If
I
wanted
to
try
and
replay
all
the
activity?
Is
there
a
way?
I
can
do
that.
D
So
I
think
the
collaborative
flag
is
not
needed
anymore.
You
just
have
to
install
Jupiter
collaboration
and
yes,
there
is
what
we
call
a
white
star
so
by
default.
I
think
the
name
is
that
Jupiter.
D
Db,
so
it's
sqlite
database
and
you
can
you
can
replay
it
that's
what
we
do.
Actually,
when
we
debug
I
can
show
you
how
to
do
that.
C
That'd
be
great
if
you
could
get
a
demo
and
and
which
directory
is
the
the
Y
store
file.
D
F
G
Yeah
I
was
I,
said:
I
was
ready
to
go
and
then
I
was
muted,
so
Jupiter
lab
issue
14
300.
We
discussed
this
at
triage
yesterday.
This
was
something
that
I
opened
based
on
some
feedback.
I
got
internally
from
my
boss,
Frank
Ranger,
who
I
don't
think
is
on
this
call.
G
G
So
Brian
wants
us
to
disable
that
scrolling
toggle
if
the
output
is
too
small,
Mike
I
believe
is
on
the
call
pointed
out
that
there
are
some
use
cases
for
wanting
to
enable
scrolling
and
even
if
it
doesn't
seem
like
it
would
be
necessary,
so
I'm
curious
about
what
the
broader
group
has
to
say
about
it.
Vidar
has
his
his
hand
up.
Did
you
have
a
question
about
this
or
comment
about
this?
G
E
I
know
that
if
a
cell
is
producing
large
outputs,
you
would
want
a
ux
to
be
able
to
do
it
right.
So
also,
it's
not
clear
to
me
like
if
I
have
a
salvos,
large
outputs
and
I
turn
on
scrolling,
behavior
and
I
clear
all
the
outputs.
G
That's
that's
a
good
observation
there
about
clearing
the
output.
Other
comments,
questions
about
this
one
also
feel
free
to
leave
comments
on
14
300
as
well.
I
just
figured
based
on
the
conversation
we
had
yesterday
at
triage,
which
is
typically
a
smaller
meeting
that
we
might
benefit
from
having
a
broader
conversation
about
this.
G
A
A
Scroll
bars
within
scrollable
areas
are
indeed
not
nice
experiences.
That
is
true,
I,
don't
know,
I
really
think
this
discussion
is.
It
involves
thinking
through
a
bunch
of
angles.
Gabriella.
Thank
you
for
volunteering
to
think
about
this.
Some.
If
anyone
has
anything
to
contribute
on
this
issue,
but
aren't
sure
yet
what
their
fully
formed
thoughts
are
Please
add
on
the
actual
woven
issue,
but
if
anyone
wants
to
chime
in
now,
I'll
wait
a
second
before
moving
on.
A
B
B
G
B
B
There
is
one
one
issue
that
Jeremy
is
on
Notebook
V7
due
to
it
something
weird
about
the
settings
that
are
not
loaded
the
same
way
and
so
the
toolbar,
like
some
some
items
in
the
in
the
notebook
7
UI,
is
not
loaded
properly.
We
are
looking
at
it,
otherwise
nothing
has
been
seen
as
far
as
I
know:
I,
don't
think
German,
oh
yeah
German,
don't
hesitate
to
jump
in.
H
H
Didn't
change
so
I
suspect
this
is
could
be
considered
as
breaking
between
beta0
and
better
one.
In
that
case,
we
should
definitely
have
a
look
before
RC.
B
Thanks
for
telling
me
then,
last
week
we
spoke
about
translation,
so
after
looking
at
the
the
condition
of
the
crowding
platform
that
we
are
using.
Actually,
as
we
were
having
a
free
plan,
recruiting
was
already
allowed
to
use
everything
that
was
produced
by
people
translating
strings
so
I
contacted
Steve,
because
it's
Steve
that
created
the
project
on
the
platform
he
applied
for
the
premium
icon
for
open
source
project,
and
we
got
it
so
yeah.
B
That's
just
so
that
everybody
is
aware
of
that,
and
just
like,
before
this
meeting,
I
have
pushed
the
string
for
Jupiter
4
and
notebook
7
for
Notebook
7
I'm,
using
the
the
main
branch.
So
the
add
for
Jupiter
4
is
the
priorities
candidate
and
they've
been
about
uploaded
to
crowding,
so
they
are
ready
for
translation,
for
if
some
people
wants
to
contribute
yeah,
that's
about
translation
about
documentation.
I
have
been
opening
a
couple
of
PR
on
documentation,
mainly
I
have
tried
to
generate
dynamically.
B
The
list
of
plugins
and
tokens
that
are
provided
by
all
core
packages
directly
in
the
documentation
and
also
kind
of
related
to
documentation.
I've
created
a
draft
PR
to
use
our
notification
system
that
we
have
now
and
the
ID
would
be
to
notify
old
user
about
RC
I,
don't
know
if
it's
a
good
idea
if
we
should
not
notify
VRC,
but
only
wait
for
final,
so
yeah.
Let's
let
me
know
what
you
think
with
her
right
now:
I
will
shut
up
for
10
seconds.
I
I
guess
a
lot
of
streams
would
have
different
channels
that
you
can
subscribe
to.
So
you
could
be
on
the
nightly.
If
you're
crazy,
you
could
be
on
the
pre-release
Channel
and
you
could
be
on
the
final
Channel
and
you
default
people
into
the
final
Channel.
D
I
I
mean
if
it's,
if
the
setting
in
there
is
so
every
you
know,
the
current
one
remains
unchanged:
it's
just
final
releases
and
then
you
can
opt
into
a
tool
tip
for
the
school
talk.
Then
you
can,
you
know,
opt
into
which
which
channel
you
want
I
assume
the
toast
thing
can
do
that.
B
That
makes
sense.
Okay,
I
looked
at
that
so
but
anyway,
like
the
pr
is
also
good,
so
that,
if
it's
for
the
final,
then
we'll
change
just
to
text
for
final
and.
B
Then
about
the
propagation,
so
we're
Florence
is
still
working.
On
approaching
the
extension
example,
there
is
16
done
out
of
19.
We
haven't
seen
any
major
blocker
I
also
have
ported
all
the
Jupiter
renderer
that
we
have.
There
are
all
cross-compatible
at
that
part
of
the
API,
isn't
changed,
so
that's
that's!
A
good
news.
B
I
have
migrated
also
the
the
tutorial
and
this
week
list
11
updated
documentation,
though
in
the
in
core
for
it,
but
so
yeah.
That's
that's
the
update
and
so
to
question.
When
should
we
start
release
candidates?
What
are
the
blocker
for
it?
B
So
there
is
the
book
from
Jeremy,
so
the
the
idea
mainly,
is
that
if
you
want
to
Jupiter
con
Jupiter
corn
is
in
a
month,
so
we
should
start
very
quickly
to
to
do
a
risk
and
did
it
and
from
the
release,
blockers
I,
don't
see
any
of
them
that
are
blocking
release
candidate
I,
see
them
as
blockers,
but
not
blockers
for
the
release
candidate
and
that's
my
opinion.
Yeah
I.
J
Have
a
good
question
so
like
release
candidates
for
my
general
education,
it
can
still
have
like
blocking
bugs
right
or
is
it
new
or
it
needs
to
be
one
to
one
to
the
final
release
and
we're
just
like
testing
it.
G
My
my
thoughts
and
I'm
I
welcome
other
opinions
on
my
thoughts
about
a
release
candidate
is
that
every
release
candidate
is
something
that
like.
If
we're
satisfied
with
it,
it
would
just
become
4-0
with
no
further
changes.
So
to
that
end,
I
I
would
like
to
see
all
we
currently
have
six
release.
Blockers
I
would
like
to
see
all
of
them
either
mitigated
or
we
decide
that
they're
no
longer
release
blockers
prior
to
declaring
rc0,
okay.
J
A
So
one
thing
we
might
consider,
then,
is
not
a
four-week
RC
period,
only
a
two-week
RC
period
because.
B
Mike
on
the
chat
says
it's
okay
for
shortening
RC
and
he's
agreeing
with
Jason
that
we
should
probably
more
do
better
and
shortening
the
RC
I'm
fine
with
that.
It
was
just
to
raise
the
point
I'm
totally
in
agreement
with
that
proposition.
So
do
we
Target
that
solving
the
blocker
issue
that
are
listed,
hopefully
in
not
a
new
broker
issue
will
be
seen,
but
let's
see
and
wait
for
those
for
before
starting
RC
and
then
allows
ourselves
to
reduce
the
the
size
of
the
RC
period.
A
A
I
see
oh
there's
six
release
blockers,
one
of
them
actually
blocks
notebook
seven.
So
it's
pretty
high
priority
because
it
prevents
notebook
7
from
having
a
true
beta.
Let's
see,
three
of
them
are
unassigned.
A
G
Could
always
do
the
assignment
during
the
issue
triage
once
the
recording.
C
G
A
G
F
It's
one
of
the
release:
blockers
is
rather
minor
and
it's
just
a
quick
CSS
fix
and
I
can
do
that.
But
I
also
have
my
hand
up
because
I
added
one
point
to
the
object
that
you
might
have
not
seen-
and
this
is
my
question
about
the
experimental
RSP
extension
and
I-
do
not
mean
the
LSP
package.
I
mean
the
LSP
extensions
or
the
integration
refusal
interface.
The
settings
Etc
I
I.
B
A
Well,
that
wasn't
him
behind
the
flag
that
was
hidden
behind
feature
detection
of
the
kernel,
so
it
was
easier
to
roll
it
out
without
people
seeing
it,
you
didn't
have
to
change
your
runtime
flag.
You
had
to
install
a
kernel
that
supported
it
and
it
just
magically
disappeared
without
that,
so
I'm
not
sure.
A
A
But
if
core
is
shipping
with
those
rest
apis,
then
that's
not
going
to
accomplish
the
same
thing.
Oh
you're
saying
the
API
will
just
say:
sorry:
I
have
no
language
servers,
so
the
language
servers
rest
endpoint
exists,
it's
just
that
it
will
return
an
empty
list.
Is
that
got
it?
Okay,
yeah.
F
F
My
concern
is
mostly
about
like,
if
you
go
to
settings
editor,
there
is
a
tracer
schema
and
I
believe
that
this
will
likely
change
some
of
these
things,
I
think,
once
we
started
that
ticket
foreign.
A
Schemas
changing
we
have
prior
art,
where
without
a
major
version,
but
what
we
did
was
create
a
temporary
extension
that
loads
the
old
formatted
settings
and
changes
them
to
the
new
format,
settings
and
writes
them
so
that
the
end
user,
never
notices.
Would
that
sort
of
pattern
accomplish
everything
we
need.
A
B
I
have
her
last
Point.
Sorry,
no.
B
So
it's
it's
more
of
a
suggestion,
but
I
know
it's
last
minute.
So
that's
why
I
want
to
get
opinion
about
it.
I've
been
playing
around
a
couple
of
since
a
couple
of
weeks
with
another
project
templates
that
is
called
copier,
copier
I,
don't
know
what
to
pronounce
it
exactly
in
English.
So
it's
a
kind
of
cookie
cutter,
but
much
more
powerful
and
nicer,
and
modern
and
yeah
lots
of
things
that
we
don't
get
with
Cookie
Critter
and
in
particular
it
also
can
do
update.
B
So
if
you
update
the
template,
then
you
can
more
easily
upgrade
your
extension
because
he's
doing
git
diff
to
be
able
to
know
what's
happening.
B
The
main
difference,
let's
say,
is
that
it
doesn't
like
you
need
to
create
a
first
directory,
it's
not
like
the
computer
and
you
need
to
create
a
directory
go
in
it
and
it
will
expand
all
the
tree
in
it.
B
The
reason
for
it
is
because
it's
saving
a
small
file
in
which
you
will
record
all
the
answer
from
the
for
the
template
and
which
Ash
of
the
git
repository
has
been
used,
so
yeah
I
can
make
it
so
that
we
have
a
new
extension
template
for
using
it
and
changing
the
documentation
related
to
it
yeah.
My
question
is
that
do
you
think
it's
okay?
Do
you
think
it's
too
dangerous
because
we
are
close
from
release
and
the
other
question
is
what
should
we
do
should
in
terms
of
repository
organization?
B
I
I
It
has
a
lot
of
nice
advantages,
but
I
think
the
biggest
win
would
be
actually
having
a
a
first
class
template
piece
of
equipment
that
was
based
on
Jupiter
lab
that
new,
which
underlying
Jupiter
lab
tool
you
had.
So
you.
I
I
And
Bam
it
gives
you
settings,
it
creates
a
schema
folder.
It
updates
your
package
Json
all
that
stuff
I
think
there's
really
a
lot
of
benefit
into
embracing
something
like
that.
More
than
just
publishing
repos
there
you
go
as
it
is
in
Spanish,
which
I
also
don't
know.
So
that's
great,
so
yeah
the
absolutely.
I
We
should
embrace
that
technology,
but
we
should
do
it
on
our
terms
and
not
have
it
be
weird,
just
be
like
hey,
there's
a
developer
experience,
it's
not
cobbled
together
out
of
50
different
things
that
you
go
independently,
install
it's
it's
a
thing
you
go
and
get
it's
not
part
of
the
build
system
right,
Jupiter,
lab
extension.
Build
would
not
require
this,
but
it
does.
It
does
directly
extend
the
Jupiter
lab
extension.
Api
see
you
live.
I
Because
I
mean
we
looked
at
this
at
Yeoman
years
ago,
right
yeah,
it
would
just
be
a
different
repo
yeah.
It
would
just
be
a
different
repo
entirely
but
but
Pi
Pi
packaged
because
that's
then
we
could
do
things
like
migrate
without
super
hacky
stuff.
On
top
of
it,
and
all
that.
A
So
the
promise
of
this
is
I
wrote
a
Jupiter
lab
for
extension,
and
a
year
has
passed
because
we
Faithfully
release
every
single
year
on
the
dot
and
I
just
want
to
upgrade
my
extension.
But
it's
not
really
fresh
in
my
mind
and
I.
Don't
really
feel
like
thinking
about
it.
I
run
a
command
called
copier
update
and
for.
A
I
Yes,
so
you,
it
would
remember
your
choices
and
can
apply
diffs
intelligently.
That's
currently
done
inside
the
migration
script
as
like
custom
stuff,
but
it
is
more
aware
of
that
first
party,
it's
more
like
yeoman
more
like
create
react,
app
nonsense,
200
megabytes,
whatever
I
think
the
important
thing
is
actually
treating
it
as
software
and
not
just
a
repo.
That's
hanging
out
out
there,
so
you
didn't
but
I
assume
since
you're
updating
for
Jupiter
lab
five,
you,
you
know
pip
install
Jupiter
lab
five.
You
previously
had
Jupiter
lab
Dev
CLI
installed.
I
I
What
it
has
you
know,
here's
new
tricks
that
I
can
I
can
teach
your
your
extension.
You
wanna
you
wanna
upgrade
for
a
Midstream
breaking
release.
I
got
you,
you
know
it's
got
all
that
stuff,
but
not
have
to
go
Discover
it
and
look
at
different
versions
up
on
on
GitHub
and
all
that
just
have
it
right
in
the
package.
You
know
and
actually
treat
it
as
releases.
A
A
A
Thanks
for
bringing
all
this
up,
we
can
skip
mine
because
we
basically
talked
about
it.
I
just
wanted
to
pitch
that
our
next
release
is
an
RC,
but
I
think
we
have
a
plan.
A
Next
release
is
a
beta
one
after
that
is
an
RC
and
we
shorten
the
RC
phase
so
that
we
could
still
hit
a
final
by
Jupiter
Khan.
Hopefully,
if
everything
works
out
okay,
so
we
have
some
commentary
on
RTC,
but
before
that,
I
just
want
to
double
check.
Anyone
wanted
to
talk
about
anything
they
didn't
add
to
the
agenda.
A
I
see
some
things
in
the
chat.
Let
me
check.
Oh
I
see:
okay,
okay,
so
how
about
we
do
this?
How
about
we
stop
the
recording
and
we
do
two
things
one
is
we
want
to
talk
a
bit
about
RTC,
the
other
is
at
the
very
least
we
want
these
release,
blocker
issues
assigned.
Oh,
my
sorry
for
the
background
noise,
so
I'm
going
to
stop
the
recording
right
now!
Brace
yourselves!