►
From YouTube: JupyterLab Team Meeting - July 6, 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
A
If
you
haven't
already
and
are
so
inclined,
please
find
the
link
to
the
agenda
in
the
chat
and
add
your
name
to
the
list
and
if
there's
something
you'd
like
to
discuss,
add
an
agenda
item
so
far
we
have
one.
So
I'm
going
to
put
my
name
here
because
I'd
like
to
talk
a
little
bit
follow
up
on
rtc
but
spoiler
alert,
I'm
going
to
say:
let's
talk
in
the
rtc
meeting
more
about
this,
but
for
now
why
don't
we
get
started
so
the
first
person
on
the
agenda
is
frederick.
B
You
are,
I'm
sorry
yeah.
I
praised
too
quickly.
Sorry,
thanks,
darian,
hey
everybody.
So
I've
tried
to
do
a
patch
release,
but
we
got
a
bit
hit
by
a
a
new
version
of
nb
classic.
B
We
have
discussed
about
it
in
the
previous
mythic,
and
that
was
just
before
this
one
and
eric
schwarz
is
gonna,
push
her
pr
on
nb
classic.
That
should
fix
the
ci
for
3.4.x.
B
On
jupiter
lab
on
the
master-
and
there
is
some
some
trouble
with
that
are
linked
with
learner,
so
basically,
when
you
want
to
bump
a
version
using
lorna,
it's
using
to
the
graph
dependency,
taking
into
account
also
the
depth
dependency
of
your
package,
and
so
if
a
package
is
bumped,
then
all
downstream
packets
are
bummed
too,
and
the
trouble
is
because
test
tutils
is
like
creating
a
circular
dependency
to
everything.
B
This
makes
the
bump,
like
not
very
fine
grain,
as
we
would
like
to
so
I've.
I've
done
a
pr
to
to
refactor
all
the
test
utils,
because
basically
that
means
removing
almost
everything.
That's
in
the
separate
package
to
place
them
in
the
right
package
and
link
them
together
correctly
in
core,
and
also
I
realized
that
there
are
some
packages
that
are
in
a
chicken
and
egg
configuration
like
the
render
mime
package
test.
There
requires
notebook
package
for
being
done,
so
that's
bad
for
like
the
hierarchy
of
the
package,
and
so
in
that
pr.
B
A
Thank
you,
fred
yeah,
it
is
looking
like
it
might
be.
A
short
call,
okay,
first
off
any
comments
on
what
fred
had
to
say
before
we
move
on.
A
Okay
cool
so
last
week
we
had
a
conversation
that
extended
into
overtime
on
real-time
collaboration.
A
In
the
interim,
I
have
been
looking
at
kevin's
pr
that
he
presented
somewhat
last
week,
but
we
couldn't
talk
at
length
about,
and
I've
also
been
looking
closely
at
carlos's
pr.
I
actually
have
been
working
on
sort
of
experimenting
with
some
of
the
ideas
in
carlos
pr.
A
So
if
you're
following
this
closely,
you
may
have
seen
that
I
pushed
some
code
up
there
today
that
isn't
actually
being
used,
but
it
is
meant
to
potentially
replace
some
of
the
shared
types
that
carlos
introduced,
but
to
not
have
them
in
a
special
package,
rather
have
them
be
in
the
jupiter
lab
application
package
as
just
collaborative
primitives.
A
But
this
this
for
me,
is
all
fairly
experimental
actually
still,
and
I
am
using
it
as
a
learning
experience
to
really
truly
understand
these
two
different
approaches.
A
I
don't
have
a
smart
thing
to
offer
right
now
and
I
think
I've
set
it
as
my
own
goal
to
give
more
thought
to
this
and
to
have
some
coherent
idea
before
monday's
rtc
call,
because
that's
when
it's
regularly
scheduled,
but
I
guess
I
would
invite
anyone
on
this
call
who's
interested
in
rtc
to
take
a
look
at
these
two
pr's
or
even
if
you
don't
feel
like
taking
a
look
at
these
two
pr's
at
least
look
at
their
front
matter
and
see
how
they
describe
themselves
and
if
you'd
like
to
discuss
the
issues
come
on
monday.
A
But
since
we
also
have
a
sparse
group,
our
expressed
agenda
today,
if
there's,
if
there's
some
sort
of
like
thing
that
you
would
want
to
know
that
helps
you
think
about
this
between
now
and
then
I
noticed,
I
think,
kevin's
on
the
call
too.
Perhaps
we
can
answer
some
questions
you
may
have
or
if
not,
that
is
okay
too.
E
D
Oh
okay,
so
yeah
now
I
can't
hear
anybody
else.
Sorry,
okay,
so
I
think
I
got
my
sound
figured
out
now.
Are
you?
Can
you
hear
me
now
because.
C
D
Was
no
one
was
talking
while
you
were
okay
yeah,
I
pressed
some
of
the
wrong
buttons
on
my
screen.
Okay,
so
I'll
just
I'll
just
say
this
real
quick,
sorry
about
all
that.
So
anyhow,
this
isn't
a
technical
issue
and
it's
probably
not
what
people
are
concentrating
on
right
now,
but
do
either
of
the
implementations
reserve
cell
level
undo.
A
So
this
is
a
design
decision
that
is
capable
of
being
implemented
with
both
of
these
designs.
Neither
design
has
an
implication
on
that.
That's
a
user
experience
joints
that
we
can
build
on
top
of
either
implementation
and,
as
far
as
I
know,
the
idea
is
that
we
should
have
some
level
undo,
but
that's
what
people
want
yeah,
so
I
don't
know
kevin.
Do
you
have
any
thing
else
you
want
to
add
to
that.
E
No,
that's
right
either.
Approach
will
support
that.
It
all
depends
on
the
move
feature
that
we
are
currently
shipping
and
it's
just
just
so.
You
know
how
it
why
it
takes
so
long.
It's
really
hard
to
implement
that
correctly,
and
we
have
now
two
versions
where
we
need
to
implement
that
the
python
version
or
the
rust
implementation
and
yjs
ygs
supports
move,
but
the
rust
implementation
does
not,
and
for
compatibility's
sake,
we
need
to
really
make
sure
that
we
that
everything
works
correctly.
Otherwise
we
we're
going
to
have
bad
issues.
A
Okay,
well,
I
think,
since
we
do
have
some
points
in
the
additional
discussion
section,
we
can
jump
to
those,
but
I'll
just
give
everyone
a
minute.
D
A
No
discussion
needed
for
who
is
at
science
next
week,
but
if
you
are
going
to
be
at
scipy
next
week,
add
your
name,
maybe
you
guys
can
be
done
and
the
second
item
there
is
from
lane
talking
about
the
code,
mirror
six
outstanding
pr
and
one
specific
feature
of
it
so
lane
I
will
pass
it
on
to.
F
Yeah
sure
so
it's
just
kind
of
a
question
and
if
so
jupiter
doesn't
have
indentation
guides,
like
you
see
on
the
screen
here,
which
you
know
for
an
ide
is
just
kind
of
a
little.
F
A
B
Hey
elaine,
so
the
idea
first
is
to
move
on
with
dot
pr
as
a
pure
move
of
features
that
are
existing,
because
it's
already
a
huge
vr
as
you've,
probably
seen,
and
definitely
after
that,
we
will
be
able
to
to
add,
add
features
if
they
are
available
in
code
mirror
6..
I
think
there
are
a
dollar
from
a
discussion
with
you,
and
so
it's
probably
possible
to
do
that
kind
of
indent
line
in
code
mirror
6.
F
Okay,
I'll
take
a
look
at
some
of
I'll,
have
a
look
at
the
existing
issues
and
see,
if
there's
anything
in
there
and
make
one,
if
not,.
A
F
C
Yeah,
I'm
just
going
to
add
at
the
I
don't
know
how
the
I
don't
haven't
fully
understood
how
the
converter
6
extension
system
worked,
but
I
know
it's
gotten
an
overhaul
and
it's
supposed
to
be
more,
maybe
easy
to
use
larger
deploys
like
this
a
stripper
lab.
So
it
might
also
be
good
to
think
about
how
how
to
for
people
to
add
code
mirror
extensions,
because
this
is
a
good
example
of
one
of
them.
C
C
So
it's
if
you,
if
you
by
looking
at
it,
learn
anything
please
write
it
down
somewhere
or
or
make
an
issue
or
make
a
suggestion.
If,
if
you
feel
you
learned
something,
it
will
be
useful
later
on.
B
Oh,
I
can
jump
in
here
too,
so
the
indeed
in
corn
mirror
six.
The
idea
is
that
you
get
two
main
elements
for
every
editors.
You
get
the
states
and
a
view
and
everything.
A
B
I
mean
it's,
it's
probably
better
in
the
sense
that
you
don't
need
anymore
to
to
deal
with
the
global
object
and
everything
is
per
editor,
so
it
will
be
probably
cleaner,
but
I
yeah-
probably
the
first
example
we
come
out
with,
would
be,
will
make
that
obvious
or
not.
B
So
that
would
be
a
good
example.
Another
trouble
that's
like
hit
by
johann
like
yesterday
is
about
teaming.
So
codemirror6
is
using
css
and
gs.
Teaming
there
are
no
more
using
I
mean
you
can
still
use
the
old
way
of
doing
css
pure
css
rule
for
a
teaming
code.
Mirror
6,
but
that's
not
the
recommended
way,
and
the
recommended
way
is
not
to
use
a
css
in
gs.
B
It's
not
too
complicated,
so
we're
probably
going
to
migrate
to
that
and
the
consequence
to
it
is
that
we're
probably
going
to
drop
the
default
team.
That's
our
ship
we've
got
mirror
5
and
that
for
now,
user
can
pick
up
from
from
the
menu
in
jupyter
lab
for
jupyter
4,
because
there
are
no
more
in
code,
mirror
6
and
it's
probably
better
to
to
wait
for
feedback
to
decide
which
one
are
interesting
to
bring
back
in.
B
So
so
what
yuan,
for
now
is
just
doing
is,
is
creating
our
own
jupiter
team
to
bring
back
the
theming
of
code,
mirror
using
the
jupiter
css
variable,
but
I
think,
like
there,
there
is
some
team
that
are
called
zen
burn
and
other
things
like
that,
so
those
are
are
gone
like
officially
in
code
mirror
6
and
if
we
want
to
provide
them,
we'll
need
to
create
our
own
second
mirror.
Six
team.