►
From YouTube: JupyterLab Team Meeting - August 10, 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
Today
we
have
13
people
on
the
call
and,
if
you
haven't
already,
please
check
out
the
linked
meeting
minutes
in
the
chat
and
add
your
name
to
the
attendees
list.
If
you
would
like
and
add
an
item
to
the
agenda,
if
there's
something
you'd
like
to
discuss,
if
you
have
something
to
discuss,
that's
going
to
take
a
bit
more
discussion,
please
add
it
at
the
bottom
in
the
section
called
additional
discussion,
so
we
can
allocate
a
bit
more
time
to
it.
A
But
with
that
being
said,
why
don't
we
get
started
today?
The
first
person
on
the
agenda
is
Isabella.
B
Hello,
hello:
we
actually
have
13
people
in
one
chat
on
the
call
I
guess
if
you
heard
me
start
recording
them
is
now
here.
So
with
that
I'm
going
to
announce
to
you
all,
the
Jupiter
lab
accessibility
meeting
will
be
15
minutes
after
the
scheduled
end
of
this
one.
B
It's
that
week
again
so
yeah
I
hope
to
see
some
of
y'all
there
totally
open
and
always
an
adventure.
So
thank
you
cool.
Thank.
A
You,
okay
Fred,
you
are,
you
are
number
two
on.
C
On
you
yeah
it's
just
to
please
you
so
that
you
are
not
number
two
but
I'm
not
free
already.
So
let
me
do
it
like
just
a
short
announcement.
We
have
released
Jupiter
at
3.4.5,
so
it's
a
batteries.
C
We
have
seen
that
there
is
a
currently
back
in
the
redeser,
so
for
people
using
it
yeah,
you
may
need
to
wait
a
day
or
two
to
for
fix
to
be
pushed
Upstream,
but
that's
just
for
sharing
information,
and
apart
from
that,
yeah
I
wanted
just
to
come
back
to
discussion.
We
had
two
weeks
ago,
I,
think
or
two
or
three
weeks
ago
about
the
two-factor
Water
authentication
enforcement
on
GitHub.
C
So
the
discussion
like
at
the
end
of
the
discussion
there
there
is
two
possibilities:
either
we
enforce
it
directly
and
people
will
get
a
notification
from
GitHub
saying
they
have
been
removed,
and
so
they
could
reapply
if
the
turn
on
to
fa
matches
also
brought
the
fact
that
it's
possible
to
create
a
private
discussion
and
to
Ping
all
of
the
people
that
don't
have
2fa
in
that
discussion.
C
The
drawback
is
that
everybody
that
is
in
the
organization
will
see
that
the
the
list
of
people
that
don't
f2fa
activated-
that's
the
only
difference-
let's
say,
but
maybe
it's
nicer
to
Ping
them
at
least
first
so
yeah.
My
question
is:
which
which
solution?
C
C
A
A
Okay,
so
I
think
Rick
made
the
case
for
2fa
and
I.
Think
it's
really
sound
I
think
September
is
in
20
days.
A
We
could
ping
them
to
be
to
give
them
a
heads
up
that
this
is
happening,
but
I
also
think
it's
probably
fine.
If
we
removed
access
and
people
enabled
it
and
got
re-added,
because
that's
immediately
more
secure
but
I
don't
have
a
yeah
I,
don't
have
a
strong
feeling
about
it,
because
September
is
coming
soon
regardless.
It's
like
anyone
else,
have
any
thoughts.
B
B
A
A
A
A
C
Sweet
cool
I
think
everybody
is
here,
has
it
I
did
not
download
the
last
version
like
the
up
to
date
list
but
I
think
everybody
does
here?
Has
it.
A
Okay,
all
right
so
I
mean
because
there's
not
clear
agreement
here.
Maybe
we
should
take
the
nicer
route
and
try
the
private
discussion
that
Matthias
is
proposing.
A
A
Okay,
so
next
person
is
me
on
the
agenda.
I
have
two
links
here:
one
is
to
a
project
board
that
I
made
in
the
Jupiter
lab
org
for
lumino
2.,
and
there
are
a
bunch
of
issues
that
are
in
progress.
There's
a
bunch
that
nobody's
working
on
and
there's
a
bunch
that
are
done
so
far.
Primarily
it
has
been
fairly
straightforward
things,
but
some
of
them
are
a
bit
more
complex
than
I,
initially
imagined,
which
is
why
I'm
asking
for
help.
A
Fred
has
been
helping
and
that's
really
helpful
and
really
reassuring,
because
I've
been
a
bit
overwhelmed
and
Gabriel's
been
looking,
and
that's
really
good,
because
more
is
better.
And
so,
if
you
have
any
interest
in
working
on
modernizing
luminal,
we
are
doing
things
like
using
es6
only,
which
means
we
get
a
whole
bunch
of
stuff
like
generators
and
iterators,
and
we
get
and
we've
upgraded
the
typescript
version.
A
Like
supporting
the
p
dash
CSS
Styles
and
like
the
data
store
experiment
that
we
ended
up
not
using
for
crdt,
the
piece
that
I'm
currently
working
on
is
removing
the
custom
lumino
iterators,
because
at
the
time
they
were
implemented,
iterators
were
not
available
in
all
the
browsers
that
we
intended
to
support,
but
they
are
now
I
actually
have
an
open
question
about
whether
we
need
to
do
any
browser
detection
because
there's
a
platform
namespace
that
will
tell
you
if
you're
looking
at
IE
or
whatever
and
I,
don't
know
if
we
still
need
that.
A
But
it's
not
it's
not
obvious
yet
so
anyway,
please
take
a
look
at
those
two
links.
Please
work
on
these
issues.
If
you
have
the
time
I'm
happy
to
help
anyone
with
any
aspect
of
them
and
it's
it'll
it'll
be
a
big
impact
thing
right,
because
lumino
is
the
basis
for
Jupiter
lab
and
notebook.
7.
Fred
has
his
hand
up.
What's.
C
Up
first,
thanks
for
working
on
that,
Second
I
think
the
the
code
for
detecting
the
browser.
We
will
need
to
keep
it
because
I
think
it's
used
in
the
keyboard
shortcut
thing.
A
Cool
yeah,
no
I
haven't
even
looked
at
that
piece.
Yet
so
that's
unsurprising,
but
makes
sense,
I
wanted
to
ask
the
folks
who
work
on
accessibility
if
there
are
any
sort
of
wishlist
items
that
require
API
changes
to
lumino,
to
make
your
lives
easier,
because
now
is
the
time
to
do
it.
Otherwise
we
have
to
wait
until
earliest
Jupiter
Lab
5
lumino3.
If
we're
modifying
apis
that
are
backward
incompatible.
A
Write
an
issue
in
the
luminal
repo
if
it
is
purely
an
ask,
write
an
issue
in
the
lumino
repo.
If
it's
an
ask
and
an
offer
to
do
it
and
comment
on
a
pre-existing
issue,
if
it's
relevant
to
that
veto
has
his
hand
up.
D
A
D
B
A
Sure,
but
even
that
is
predicated
on
if
there
is
a
thing
that
makes
the
lives
of
people
working
on
accessibility
easier,
because
that
is
one
of
the
that's
one
of
the
directions.
I
think
we
should
push
on
for
luminal
and
I'm,
not
sure,
there's
really
anything,
that's
backward
and
compatible.
That
needs
to
be
done.
It
could
really
be
that
every
single
widget
needs
to
be
visited
and
have
an
audit
done
on
which
components
of
it
need
to
have
Aria
labels.
A
Etc,
like
it's
unclear
that
it's
anything
backward
incompatible,
but
it's
a
good
time
to
look
but
more
than
anything,
please
help.
This
is
a
lot
and
we
can.
We
want
to
do
it
as
soon
as
possible,
because
there's
going
to
be
as
Vidar
says,
Fallout
and
big
chunks
of
Jupiter
lab
that
rely
on,
say
the
iterators
in
widget.children
or
in
data
models
for
data
grids
will
need
to
be
amended
and,
at
the
very
least,
all
the
package
jsons
will
need
to
be
updated
and
the
same
goes
for
extensions.
E
Is
this
change
could
have
good
to
be
done
in
the
Blackboard
compostable
way
and
could
be
deprecating
enough
iterators
for
future
business.
Tropical
support.
A
A
Stuff,
I
really
strongly
recommend
not
writing
in
a
Backward
Compatible
way
because
oftentimes
when
we
have
iterators
they're
at
performance,
low
points
to
begin
with,
so
writing
some
sort
of
sham
or
writing
some
sort
of
compatibility
layer.
There
seems
like
a
mistake
and
I
think
it's
just
time
to
use
the
native
functionality
now
that
we
have
the
opportunity
to
if
I
don't
know,
I
mean
I,
don't
want
to
unilaterally
make
this
decision,
but
I'm
the
person
working
on
it.
So
I
can
unilaterally
decide
what
I
want
to
type
on
the
keyboard.
A
B
I
just
wanted
to
verbally
reply
to
Vidor
and
say
no
I
think
that's
a
really
good
point
about
not
just
doing
things
to
do
them
so
we'll
be
cautious
about
it,
but
I
also
appreciate
the
flagging
that
some
things
may
be
being
changed
anyway.
So
thank
you.
A
Cool,
thank
you
so
again,
please
follow
those
links
and
at
least
take
a
look
because
you'll
see
visually
what
is
happening
there
and
with
that.
Why
don't
we
move
on
to
like.
E
I
just
wanted
to
highlight
an
issue
on
Twitter
people,
because
many
of
you
might
not
be
subscribed
here.
There
was
an
advanced
user
who.
E
So
that's
a
huge
sort
of
knowledge
for
user
experience,
understanding
from
coming
from
someone
who
is
outside
of
the
country
that
I
want
to
developers
team,
but
there
are
issues
which
are
specific
to
Jupiter
lab
desktops
such
as
expecting
and
a
package
rather
than
an
installer
on
Macos
or
problems
with
environment
selection.
E
But
there
are
also
some
rough
edges
and
crypto
are
pointed
out
which
lets
this
kind
of
comparison
to
a
culmination
of
the
or
expiration
of
their
limit
of
patients,
and
they
could
just
give
up
and
I
also
linked.
E
Another
story
on
helpful
some
novice
users
using
R
in
capital
lab,
might
be
difficult
because
it
doesn't
come
with
the
German
R
kernel
shipped
by
default
and
taking
a
little
bit
out
of
this
issue,
but
also
from
things
that
I
would
like
to
change
in
Jupiter,
lab,
tested
distribution,
I
wanted
to
talk
about
or
seek
feedback
onto
other
issues,
and
one
of
them,
maybe
taking
a
step
back.
E
So
currently,
there
are
some
problems
in
the
startup
which
are
platform
specific,
so
the
problems
which
appear
on
the
Mac,
for
example,
on
Windows,
and
if
you
are
a
Jupiter
web
developer
and
maybe
heard
about
the
example,
try
that
maybe
that
would
be
very
helpful
if
you
could
take
a
look
and
see
if
you
can
have
diagnose
these
issues,
because
and
well,
because
everyone
has
smart
enough
to
go
and
use
Windows
and
just
a
collective
damage
like
have
that.
E
So
moving
on
to
the
other
set
points,
I
am
proposing
to
expand.
The
environment
selection
file
actually
includes
Auto
detectives
environments,
choices
for
which
device
environments
for
crypto
lab
desktop
and
again
because
there
isn't
much
attraction.
I
want
to
highlight
that
feedback.
I
know
that.
So,
if
you
might
push
against
this
propose
also
that's
what
I
eventually
get.
And
similarly
another
issue
is
that
when
you
open
Jupiter
lab,
let's
go,
it
operates
as
a
global
application.
E
There
is
no
concept
of
a
project
which
can
be
annoying
because
we
often
want
to
change
the
digital
environment
that
you
are
working
with,
and
we
want
to
have
workspaces
specific
to
different
projects
that
require,
and
currently
because
of
this
Global
specification
and
also
some
issues
with
our
worst
devices
are
implemented.
It
means
that
every
time
that
you
switch
between
projects
or
open
book
projects
at
a
simple
time,
you
will
get
workspaces
mixed
up
and
one
simple
proposal
to
resolve.
E
That
would
be
to
add
a
DOT
Jupiter
for
the
which
would
store
the
default
workspace
and
settings
on
this
so-called
project,
which
is
really
in
line
with
our
DOT
by
Python
and
that
vs
code.pytron.
So
it's
it
seems
like
practice,
but
I
guess.
For
some
reason
it
wasn't
done
for
many
years,
and
maybe
there
are
good
reasons,
but
I
just
couldn't
find
any
issue
or
a
prior
discussion
about
so
I'm
hoping.
This
is
a
proposal
for
a
quick
fix
that
would
produce
a
lot
of
problems
for
using
Jupiter
lab
in
a
desktop
environment.
A
With
regard
to
the
first
bullet
point,
that
seems
like
a
really
terrible
user
experience.
Somebody
had
so
much
so
that
it
makes
me
wonder
if
we
should
label
Jupiter
lab
desktop
beta
or
something
right.
If
this
is
People's,
First
experience
in
Jupiter,
lab
I,
don't
know,
I
don't
want
to
turn
them
off.
William
was
that
a.
F
Yeah
I
was
just
reading
through
the
whole
thing
that
you
said
and
I
to
some
extent
it's
not
necessarily
Jupiter
lab
desktop,
but
just
Jupiter
lab
that
was
causing
him
like,
for
example,
he
set
it
up
and
didn't
have
any
kernels
installed,
and
then
he
complained
a
lot
about
the
easier
experience.
Not
saying
hey,
you
don't
have
any
kernels.
F
So
it's
the
sort
of
thing
you
would
have
kind
of
irregardless
of
whether
using
Jupiter
lab
desktop
or
just
installed,
Jupiter
lab
in
some
other
way,
VIA
pip
and
then
there's
complaints,
for
example,
about
how
word
line
wrapping
Works
in
code
mirror
compared
to
say
other
editors.
So
there's
like
a
whole
page
about
that
which
is
again
not
at
all
a
Jupiter
lab
desktop
thing.
F
So
just
my
little
comment
is
that
it
it's
labeled
as
Jupiter
lab
desktop,
but
really
almost
every
or
like
half
the
things
he's
complaining
about
are
General
Jupiter
lab
behaviors,
which
may
or
may
not
actually
be
a
valid
complaints.
In
all
cases,.
B
F
Is
good?
This
point
was
just
that
if
you,
if
somehow
you
managed
to
have
Jupiter
set
up
without
any
kernels,
then
Jupiter
should,
instead
of
letting
you
proceed
to
try
to
make
a
document
and
be
really
confused,
it
should
just
say:
hey.
You,
don't
have
any
kernels
for
some
suggestions
for
installing
one,
but
it's
not
really
necessarily
a
Jupiter
desktop
specific
issue
because
I'm
sure
there's
some
way
to
get
it
installed.
It
doesn't
have
a
working
kernel.
E
And
this
way
is,
is
really
complex:
interaction
between
macros
security
policies,
making
a
desktop
application
which
works
without
an
installer
and
convey
installer,
which
creates
an
environment
in
that
it's
really
at
odds
with
how
Macross
character
works,
but
I
I
have
difficulties
trying
to
fix
that.
How
fixing
that,
because
I
don't
own
a
Mac
so
yeah,
that's
that's
really
my
call
or
anyone
might
be
using
links
to
have
a
look,
because
that
might
be
a
really
simple
solution
to
the
problem.
That's
not
using
this
experience.
A
As
for
your
second,
no,
your
third
bullet
point,
the
dot
Jupiter,
so
I
think
you're
totally
right
about
the
implication
that
you
only
ever
have
one
project
which
is
kind
of
how
workspaces
work
is
not
a
good
one.
The
Saving
Grace
of
workspaces
is
supposed
to
be
the
different
URLs,
but
people
tend
to
not
use
URLs.
That
way,
no
one
goes
to
a
separate
URL
because
they
want
to
load
a
separate
project
unless
they're
on,
like
a
hosted
environment
where
somebody
links
it
to
them
already.
A
So,
if
you
have
ideas
for
how
to
improve
this
experience,
I
did
most
of
the
work
on
workspaces
and
I
totally
concede
that
that
was
a
1.0
effort.
So
it's
definitely
like
worth
taking
a
look
at
Vidor.
What
do
you
thinking.
D
A
Yeah
there's
a
few
pieces
of
work
that
we
did
to
get
to
something
like
this.
One
is
you're
right.
We
did
make
an
API
endpoint
to
list
workspaces.
For
this
reason
another
is.
We
made
workspaces
themselves
double
clickable
to
treat
as
like
a
project
file
so
yeah.
No,
we
should
revisit
this.
We
should
make
this
better.
A
E
E
Mike
problem
is
not
really
that
there
isn't
a
workspaces
really
it's
doubts.
I
want
to
open
Jupiter
lab
desktop
in
one
place
and
have
the
the
default
workspace
come
up
automatically
for
that,
so
for
the
project
and
I
would
even
want
to
have
multiple
different
work
expenses
for
different
projects,
so
it's
more
of
a
meta
issue
of
which
can
be
currently
configured
using
and
the
environment
variable.
D
So,
just
some
brainstorming
on
that.
There's
like
for
vs
code
right,
there's,
two
ways:
you
can
open
portfolio.
You
can
open
a
blank
previous
code
to
a
no
folder
and
then
you
can
say
open
folder
and
you
find
a
folder
somehow
or
you
can
create
a
new
one
or
whatever
right
and
then
at
least
the
windows.
For
me,
there's
like
a
shortcut
previous
code.
You
can
just
right,
click
it
and
then
select
all
your
last
used.
Folders
are
there
as
a
thing?
D
Those
kind
of
things
are,
you
know,
platform
specific
and
might
not
be
that
easy
to
support
via
electron,
so
well,
that
is
probably
the
best
user
experience.
I
would
probably
go
for
the
first
experience
as
a
stepping
stone,
at
least
on
the
way
that
you
can
having
a
UI
to
switch
once
you've
started.
It
up
will
resolve
some
of
that
right.
Even
if
you,
you
know,
it's
not
the
same
experience
as
starting
up
with
a
specific
workspace.
E
Yes,
so
there
are
also
other
programs
desktop.
You
don't
have
an
access
to
the
year
round.
You
cannot
great
country
workspaces.
E
D
E
Yes,
and
and
in
computer
lab
desktop,
even
if
you
you
would
have
to
have
a
way
to
create
the
new
new
workspace
example.
So
that's
I
can
understand,
and
currently
we
can
start
script
and
lab
desktop
using
jlab
commands
in
any
directory
from
column
line
or
create
a
shortcuts
to
do
that.
So
that's
why
I
think
that's
making
just
a
top
Jupiter
folder
as
an
implementation,
detail
and
a
way
to
Mark
any
directory
or
a
new
route
as
a
project
in
in
the
Jupiter
lab
interface
would
solve
a
lot
of
these
problems.
E
A
So
this
I
don't
this
seems
somewhat.
A
A
D
Just
that
the
the
less
you
store
in
these
folders
generally,
the
better
right
so
I
think
for
like
for
me,
it's
called
if
you
just
open
the
folder,
if
you
don't
change
any
settings,
I
don't
know
how
much
stuff
gets
stored
right,
so
the
fewer
files
that
gets
put
in
it
and
last
kind
of
coded
are
in.
It
is
better
right
because
some
people
like
to
put
all
of
this
and
like
Global,
get
ignore
things
or
local
ones
or
whatever
some
people
like
to
commit
these
things
into
their
repositories
right.
D
So
it's
a
having
something
that
is
not
too
noisy.
It
could
also
be
useful.
A
D
A
Yeah
yeah,
okay,
so
let's,
let's
probably
discuss
this
some
more
I
think.
Nonetheless,
the
problem
you're
trying
to
solve
is
a
real
one
and
we
should
solve
it
whatever
the
mechanism
is
I
think
this
probably
is
a
sound
mechanism
just
given
that
other
Ides
will
have
to
do
it
but
yeah.
It's
probably
I'd
like
to
hear
some
more
people's
ideas.
A
So
if
you
don't
have
anything
to
say
right
now,
but
this
is
something
you
care
about,
please
chime
in
on
that
issue:
let's
bring
it
up
again
next
week
too,
and
but
yeah,
let's,
let's,
let's
head
toward,
actually
fixing
this
right,
because
I
think
it's
a
real
use
case
that
real
users
have.