►
From YouTube: JupyterLab Team Meeting - 29 March 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
March
29th
weekly
Jupiter
lab
call.
Today
we
have
18
people
on
the
call,
but
we
also
have
a
very
sparse
agenda.
So
if
you
would
like
to
talk
today
is
a
good
day
because
you'll
actually
have
some
time,
but
why
don't
we
get
started?
The
first
person
on
today's
agenda
is
Fred.
B
Hey
everybody
so
before
jumping
to
the
point,
I
already
done.
Just
to
give
you
a
some
updates
on
the
foreign
beta0
is
released.
Jeremy
has
been
working
a
lot
to
use
it
in
London
project,
especially
with
use
it
for
having
a
new
beta
for
Notebook
7,
and
it
will
look
at
Jupiter
lead
and
there
is
also
people
in
constack
that
are
looking
at
porting.
B
The
extension
example
to
be
sure
that
simple
extension
can
can
be
ported
easily
with
the
new
version
and
that
there
is
no
no
problem
with
the
the
new
API
and
on
my
site.
I
was
looking
to
to
start
configuring
the
quoting
platform
that
we
are
using
for
translation
for
Jupiter
4..
The
idea
was
to
use
another
Branch
because
they
have
that
ability
to
ask
translators
to.
B
Different
version
of
the
same
application,
unfortunately,
that
feature
is
not
available
for
the
free
plan.
We
are
using
so
yeah
the
the
solution
that
I'm,
seeing
that
we
can
do
is
so.
What
we
are
doing
for
now
is
that
we
are
merging
different
version
to
provide
the
full
list
of
strings
that
need
to
be
translated
to
the
translation
platform,
and
so
the
idea
would
be
to
and
to
merge
them
the
the
Jupiter
up
for
priorities
with
the
stable
release
to
get
the
new
string
translated
if
possible
by
the
contributors.
B
To
drop
like
to
to
take
the
to
take
the
habits
of
cleaning
up
the
list
of
string
when
we
do
a
major
new
release,
because
otherwise
we
may
have
diabetes
to
just
keep
increasing
the
number
of
string,
and
there
is
another
one.
The
limitation
of
having
your
free
plan
is
that
we
have
also
limitation
on
the
number
of
word
that
we
can
submit
to
the
platform.
I,
don't
know
how
far
we
are
right
now,
but
I
know
that
there
is
also
a
limitation
there.
A
A
If
we
merge
the
branches
worth
of
language,
listen
to
One
bundle.
Does
that
is
that
a
good
thing
like?
Does
that
actually
maybe
make.
A
Language
updates
more
transparent
because
they
propagate
downward,
like
maybe
it's
actually
a
good
idea
in
general,
because
you
make
one
language
bundle
for
Jupiter
lab.
It's
always
one
language
bundle
for
Jupiter
lab,
and
it
actually
accidentally
turns
out
that
if
you
have
an
old
version
of
Jupiter
lab,
your
language
might
still
get
updated.
B
B
That
I
know
that
we
have
that
other
limitation,
that
at
some
points
the
platform
will
tell
us.
We
have
a
free
plan.
You
cannotif
have
as
many
words
for
translation,
but
I
mean
we
can
move
forward
with
merging
I'm
fine.
With
that
solution.
I
was
just
wanting
to
ask
opinions
about
it,
and,
and
if
we
in
the
future,
we
eat
that
limitation
of
the
number
of
words,
then
we
can
revise
what
to
do.
E
E
The
automatic
recognition
of
drinks
from
screenshots,
which
is
a
function
of
crowding
very
useful
for
translators,
might
not
work
as
well
us
if
there
were
two
separate
branches,
but
that's
probably
a
minor
point
and
essentially
I
think
we
are
not
asking
at
the
right
place
because
it
ideally
would
just
post
a
message
to.
Everyone
contributes
on
crowding
using
the
interface
to
see
how
much
of
a
priority
system,
translators,
think
and
speaking
that
by
countries.
B
It's
just
that
raising
the
question
of
money.
I
was
not
feeling
comfortable
to
to
write
that
question
there
as
there
are
providing
us
their
money
through
their
time.
It's
more
like
in
Disco
and
I
feel
confident,
because
we
know
that
some
of
the
people
here
are
working
for
companies
that
sometimes
are
ready
to
to
contribute
financially
so
yeah,
but
I
will
reach
out.
That's
a
good
point.
Mike
to
the
contributors
on
clothing,
one.
D
E
There
are
some
small
things
to
do
so.
If
the
location
of
the
string
changes
in
the
source
file,
it
will
be
recorded
that
it's
in
a
different
line
and
if
the
platform
cannot
fully
recognize
the
string
as
the
same
as
it
was
previously,
it
will
still
suggest
the
translator
to
copy
from
the
most
similar
string.
So
there
is
like
a
single
word
single
letter
changed.
It
will
still
be
quite
easy
for
the
translator
to
get
it
back.
B
B
Full
extent
of
features
we
should
speak
about,
I
was
more
going
for
speaking
only
of
what's
new
in
Jupiter,
app
for
and
Marta
suggested
that
we
could
probably
extend
that
list
as
a
the
last
Jupiter
con
was
in
2020,
meaning
it
was
when
we
were
released
like
just
before
really
this
official
Egypt,
Arab,
3.0.0
and
so
yeah.
That
was
also
to
bring
the
discussion
and
as
the
community.
What
would
you
like
to
us
to
present
as
a
features
like
as
new
shiny
things
that
took?
That
was
the
other
question
for
me.
For
me,.
C
C
F
Jupiter
scheduler
in
at
the
conference
yeah
we
have
a
poster
approved
for
Jupiter
scheduler,
but
not
I,
don't
believe
we
have
a
talk
on
the
agenda
about
it.
E
A
Thanks
all
right:
next
up,
we
have
Vidar.
G
We
did
it
also
in
the
hack
MD
a
little
bit
asynchronously,
so
I
just
saw
that
the
Jupiter
lab
render
mime
pulls
in
the
yjs
code
via
the
code
editor
package,
that
it
depends
on
it's
fine
and
then
in
my
just
test
in
nvidime
there.
It
fails
with
the
reference
error
to
crypto
and
I.
Don't
know
it
does
just
use
webpack
webpack
under
the
hood.
Is
that
maybe
the
issue
and
that
went
back
forward
to
webpack
5
transition
means
that
you're
losing
on
the.
G
G
A
This
is
probably
three
six
I
imagine,
but
is?
Is
it
not
possible
to
tell
just
to
use
something
when
crypto
is
mentioned
like?
Can
you
not
set
a
global.
G
I'm
sure
I'm
sure
I'm
sure
there's
a
conflict
setting
somewhere
I
just
want
to
make
sure
to
I,
know
I
understand
why
it
started
failing.
G
G
Yes,
I
said
this
is
not
the
webpack
building
step
right.
This
is
this,
isn't
just
just
and
I,
don't
know
if
that
use
this
way
back
under
the
hood.
A
G
G
Any
moment
as
well,
so
this
is
all
useful
stuff,
anyways.
B
Because
I
know
there's
been
changes
in
lumino
called
utils
about
it
and
that
brought
some
some
issue
for
a
couple
of
people.
G
That's
why
I
was
asking
here
but
yeah
I
know
that
Lumen
also
did
some
back
and
forth
on
crypto,
so
yeah
I
was
just
picking
people's
brain
so
but
yeah.
If
it
was
recently
added
in
like
three
five
or
three
six
is
dependency,
then
it
probably
makes
sense
so
I'll,
just
look
at
the
just
config
then
make
sense
thanks
a
lot.
A
H
Hi
everyone
I
had
a
technical
question
about
crypto
lab
LSP
and
wanted
to
understand
if
it's
possible
to
just
turn
on
linking
and
suppress
all
the
other
features.
Different
LSP
extension
is
installed.
H
I
saw
that
there
are
already
some
responses
here,
I'm
still
not
sure.
If
this
is,
for
example,
a
disabled
extensions
is
an
example
available
that
I
can
see
to
kind
of
know
what
values
it
can
be
used
for.
That.
H
E
H
E
Unfortunately,
the
naming
of
the
command
which
disables
individual
plugins
is
disable
lab
extension
disable.
H
Okay,
so,
for
example,
LinkedIn
is
another
extension
and
completion
is
another
one,
so
it'll
disable.
You
can
disable
them
selectively.
Yes,.
E
That's
at
least
the
state
for
the
extension
Jupiter
lab
NSP
extension
in
version
three
and
four.
It
might
change
in
the
future
as
we
integrate
that
with
the
core.
But
this
is
the
case
right
now.
H
So
is:
is
the
Assumption
correct
that
in
four
there's
native
integration,
at
least
the
interface
for
LSP
in
in
4.x,
or
is
it?
Is
it
still
that
users
need
to
install
the
LSP
extension
in
order
to
use
LSP
completions,
for
example,.
E
So
there
isn't
most
of
the
features
that
are
in
their
accession
are
not
implemented
in
the
core.
The
core
basically
put
in
some
of
the
core
infrastructure,
which
is
used
for
communicating
with
the.
F
E
And
there
is
an
implementation
of
some
completion
in
in
Jupiter
lab
core
for
Jupiter
lab
4.0,
but
it
actually
interferes
with
the
extension
as
it
is.
So
this
is
one
of
the
tasks
that
we
actually
should
be
looking
at
right
now,
as
the
beta
was
released,
how
to
get
it
back
together,
how
to
put
the
Jupiter
LSP
extension
so
that
it
still
provides
the
all
the
features
and
then
selectively
pull
the
plugins
from
the
extension
into
the
call,
as
we
have
consensus
on,
including
any
of
the
specific
features
like
you
mentioned
LinkedIn.
E
A
E
So
December
debate
there
is
Jupiter
lab
LSP
extension
version
4
and
there
is
Jupiter
lab
flow.
So
if
you
have
Jupiter
the
extension,
the
LSP
extension
would
not
work
at
all,
and
this
is
also
because,
like
for
the
linking
for
the
Diagnostics
to
show
up,
we
have
to
put
all
that
code
to
code
mirror
6.,
but
the
server
parts
of
the
extensions
of
Jupiter
LSP
server
extension.
It
is
still
kind
of
working,
but
it
doesn't
provide
user
interface
features.
E
I,
don't
have
an
answer
to
that.
I
I
would
like
okay
for
that
to
be
the
case
and
would
probably
have
time,
but
that
might
require
some
coordination.
D
A
Great
anyone
who
hasn't
added
to
the
agenda
but
might
want
to
talk
about
something
now
is
your
chance.
I'll
give
you
a
minute.
I
Yeah,
let's
see
here,
I've
been
just
talking
myself
in
the
chat
over
here
yeah.
So
regarding
the
very
last
topic.
I
I
Then
another
one
is
just
a
general
hot
tip
of.
If.
I
E
I
I
have
a
point
here:
I
I
actually
opened
an
issue
proposing
that
we
have
a
list
of
an
ability
to
disable
plugins
selectively
I'm
linking
this
issue
in
the
chat
right
now.
I
Yeah,
that's
cool,
I
mean
I,
don't
know
it's!
It's
really
easy
to
get
into
a
real
broken
state.
If
you
disable
something
that
something
else
needs
and
then
you're
UI
might
not
draw,
and
you
will
not
be
able
to
recover,
like
we've
been
talking
about
having
a
ring
zero
configurator.
That
itself
is
not
extensible
for
everything
all
the
way
down
to
server
forever,
but
you
know
having
hacked
a
fair
amount
of
hot
lab
with
chains
of
extensions.
I
At
the
same
time,
it's
it's
very
easy
to
get
yourself
in
a
you're
not
going
to
space
today
kind
of
situation,
but
yeah
that
would
be
cool
it'd
be
great.
If
we
could
do
that
in
a
way
that
you
could
get
back
from
you
know,
it's
just
the
the
broken
Coke
bottle
is
super
important
and
then
scrolling.
Unless
you
have
anything
else
to
retort
on
there,
your
doctor,
I'm
gonna,
say
no
okay
and
then
scrolling
back
up
a
little
bit
further.
I
All
this
webpack
stuff
yesterday
I
want
to
say,
Safari,
16.4
landed
and
16.5
is
already
planned
or
whatever
and
a
crucial
feature
is
coming
there.
That
has
been.
You
know
a
thorn
in
everybody's
side
for
a
long
time,
and
that
is
yeah.
If
you
make
disabled
extensions
work,
if
you
actually
don't
load
them,
even
though
they're
called
for
it,
that
would
be
great,
is
scoped
import
Maps.
I
So
now
that
you
can
put
a
you
know,
you
can
make
a
script
tag
with
type
module.
Every
browser
supports
that
you
know
95
browser
penetration
or
whatever,
with
import
Maps.
They
do
basically
what
our
module
Federation
stuff
does
where
it
says.
If
you
do
a
require,
but
now
an
import,
a
real
native
import
without
require
from
a
particular
file.
I
The
scoped
import
map
can
tell
you
which
file
that
should
resolve
to
so
it
can
handle
our
duplicate
Vegas.
It
can
handle
all
that
kind
of
stuff.
There
would
still
be
probably
an
upfront,
an
upfront
wouldn't
call
it
a
build
step.
It
would
be
more
like
extensions,
would
declare
what
their
Imports
are
just
like.
We
have
today,
and
the
server
would
be
responsible
for
generating
that
import
map,
and
that
is
just
a
Json
script.
I
But
yeah
I
mean,
if
you
rode
clean,
is
the
driven,
no
es20.
Whatever
is
supported
by
the
browsers
there
wouldn't
have
to
be
a
transpilation
step.
You
could
even
write
es
with
the
the.
I
You
wouldn't
get
the
you
know,
let's
not
start
using
typescript
decorators,
let's
just
not,
but
that's
a
pretty
exciting
thing
and
looking
out
to
five,
you
know
standards,
alignment
more
schema,
more
use
of
web
standards,
maybe
web
components
will
get
themselves
figured
out,
but
I
see
that
as
a
really
good
way
to
get
us
off
of
Jupiter
Island
and
working
better
with
with
other
tools.
E
So
the
problem
is
that
we
cannot
currently
know
which
arguments
does
an
extension
accept
without
looking
at
the
code
where
it
is
defined
and
Nick
proposed
to
create
a
schema
and
register
every
command
with
types
arguments.
There
are
some
other
possible
solutions,
but
just
so
that
it's
not
toward
us
talking.
G
So
I,
don't
I,
don't
have
any
erect
input
on
that
other
than
to
say
that
there
are
a
bunch
of
other
command
palettes
and
command
Solutions
around
in
electron
right.
So
if
anybody
has
any
data
on
how
other
plugable
extensions
or
systems
do
this,
that
could
be
very
informative,
maybe.
G
G
A
G
For
Australia
there's
others
extension
system
just
has
some
sort
of
Central
Command
thing
so
more
the
more
lessons
you
can
learn
from
other
people's
mistakes,
the
better
right.
E
A
I
So
dap
dap
is
good
at
schema.
Right,
like
it
has
great
stuff
in
there.
It
doesn't
really
Define
custom
junk
that
you
can
do.
I
The
LSP
has
repeatedly
pushed
back
against
using
a
standards-based
format
for
documenting
itself
or
working
on
the
wire
or
anything
so
they've
got
their
own
special,
several
layers,
deep
meta
model
for
things
and
you're
expected
to
go,
learn
how
to
parse
their
thing,
which
is
great
if
you're
inside
of
the
micro
Microsoft
wire,
but
otherwise
you
just
kind
of
left
your
own
devices.
G
I
They're
using
that
to
build
out
yeah
right
arguments
or
any
nice
they're
using
that
to
build
out
the
python
LSP
types
which
are
used
by
piggles,
the
toolkit
for
building
python
stuff.
To
my
knowledge,
the
python
LSP
package
still
does
its
own
thing,
so
whatever
but
yeah
I
mean
it
a
lot
of
like
so
the
the
ones
that
people
have
won
do
custom
stuff
I
want
to
say
that
latex
had
the
law.
Tech
LSP
had
some
stuff
that
it's
kind
of
broken.
I
If
you
can't
use
it
for
like
building
a
latex
document,
and
then
it
has
some
additional
interesting,
complicated
stuff
for
going
to
the
place
in
a
lot
Tech
document
from
a
PDF
where
the
log
Tech
came
from,
which
is
kind
of
incredible.
But
we
have
not
tried
to
to
bite
that
off,
but
I
wouldn't
be
surprised
if
their
commands
are
just
any,
and
that
looks
like
what
what
Mike
has
has
found
here.
I.
A
I
Yeah
I
mean
I
think
from
the
Jupiter
from
the
Jupiter's
perspective,
putting
it
in
a
we've
already
claimed
package
Json.
We
already
have
a
sub
key
in
package.json.
If
you
put
Jupiter
lab
in
your
package
Json,
we
we
got
that
you
know
and
I
see
the
the
only
problem
with
package.json.
Is
it
actually
ends
up
on
the
wire,
a
fair
amount
for
different
kind
of
things,
so
it
ends
up
in
the
the
registry
and
all
that.
I
So,
if
you
look
at
file
browser,
if
you
actually
fully
described
with
you,
know
nice
human,
readable
semantics,
all
of
the
definitions
of
your
commands,
I
pick
a
Json
is
going
to
start
getting
pretty
big.
That's
not
my
problem
and
right
now
we
don't
actually
in
in
vanilla
in
core
lab.
We
don't
fetch
the
package
jsons.
I
No,
it's
got
its
own
thing.
It
does
its
own
thing
to
list
that,
so
it
doesn't
actually
serve
up
the
whole
thing.
So
yeah,
it's
really
it
there.
There
are
certainly
some
trade-offs
for
for
how
that
would
work
and
not
everything
would
fit
in
there.
There's
definitely
places
that
dynamically
Define
commands
where,
where
I
see
the
value
from
it
is
the
documentation
side.
So
you
know
you
drop
into
you
drop
into
Sphinx
and
we
go
off
and
generate
all
those
things
and
we
know
exactly
which
plug-in
it
maps
to.
I
Oh,
that
was
the
other
thing.
It
could
go
in
the
schema.
The
schema
thing
where
we
also
have
our
own
the
the
plug-in
settings
whatever
wherever
it
goes,
if
it's
at
rest
outside
of
typescript,
suddenly
we
can
do
all
kinds
of
things
with
it
that
are
not
shackle
to
the
typescript.
A
Right
someone's.
A
Okay,
I
suggest
stopping
the
recording
and
then
we
can
maybe
look
at
some
open
bugs,
but
I'll
give
you
one
last
opportunity.
If
there's
something
you'd
like
to
say.