►
From YouTube: JupyterLab Team Meeting - 1 February 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
Just
hello
and
welcome
to
the
February
1st
weekly
Jupiter
lab
call.
Today
we
have
20
people
on
McCall
if
you're
just
joining,
please
find
the
agenda
and
oops
that's
a
PR
in
the
chat.
Let
me
paste
the
real
link
and
set
the
link
to
the
pr
and
please
sign
in
if
you
haven't
already
and
if
you'd
like
to
talk
about
something
for
about
five
minutes.
Just
add
a
bullet
point
in
the
agenda
and
if
you
got
something
longer
than
that
added
at
the
additional
discussions
near
the
end,
and
why
don't
we
begin?
B
Hello,
everybody
weekly
update
on
YouTube
360.,
a
big
thanks
to
Mike
for
solving
the
the
style
issue
I've
seen
in
the
release
candidate,
one
I
think
we
are
good
to
go.
So
this
is
the
right
time
to
tell
me
if
it's
not
if
somebody
has
seen
a
blogger,
otherwise
I'll
link
the
and
some
some
draft
posts
that
usually
is
posted
on
Discord
when
we
do
a
minor
release
feel
free
to
like
place.
C
The
the
weird
blue
formatting
issue,
I
think
we
all
solved
it
and
and
I
I
commented
on
the
pr
and
I
think
that
the
pr
seems
to
have
fixed
it.
But
I
don't
know
if
it
was
merged
and
if
it
was
back
ported.
Two
three:
five,
because
that
that
that
looked
really
ugly
did
that.
Go
in
put
that
in
after
the
RC.
C
D
Thanks
I
believe
it
only
affects
us
3.6
release,
convenience.
B
E
B
Okay,
thanks
Mike
I
will
look
down
at
the
issue
you
just
mentioned,
and
I
would
just
move
on
the
other
points
so
that
we
have
time
for
triage,
so
I'm
coming
back
to
the
point
I
erased
last
week,
so
for
four:
oh
we're
gonna
extract
everything,
that's
related
to
corroboration
UI
element
in
a
separate
project
and
the
work
has
started
to
move
that
coding,
Jupiter
server.
B
Why
doc
the
that
project
was
already
existing
because
it
was
providing
the
back
end
for
collaboration
element
and
so
the
question
I
raised
last
week
and
I.
There
is
an
issue
on
the
typical
pasta
starting
created
discussion.
It's
should
we
move
it
on
the
Jupiter
lab
GitHub
organization,
because,
basically
you
know
it
starts
to
be
really
related
to
Jupiter
lab
and
notebook
7,
of
course,
but
and
yeah.
And
if
there
is
a
consensus
to
to
today,
then
we
can
directly
move
on.
B
A
F
B
Yeah,
so
the
idea
is
that
we
will
not
have
the
flag
Corporation
collaborative
anymore,
but
you
will
have
to
install
that
one
to
to
enable
collaboration,
and
so
that's
mean
that
when
using
when
installing
Jupiter
app
for
it
won't
be
in-
and
so
you
also
don't
have
like,
for
example
the
FI
ID
service.
But
if
you
decide
to
say
I
want
RTC,
then
you
will
need
to
install
that
package
and
it
will
pull
all
the
dependency,
including,
for
example,
server,
fi
ID.
F
Got
it,
can
we
make
a
an
optional.
G
B
C
And
in
terms
of
potential
loss
of
functionality,
I
just
want
to
understand.
If,
if
that
package
is
not
installed,
is
there
is
there
anything
in
the
single
user
experience
that
might
be
powered
by
the
RTC
Machinery?
B
I,
don't
I,
don't
think
it's
gonna
break
break
the
single
user
experience,
knowing
that
even
with
RTC
there
is
a
a
known
limitation
is,
for
example,
if
the
two
view
of
the
document
are
not
the
are
not
using
the
same
model.
So
sorry,
so
it's
going
to
be
a
technical,
but,
for
example,
if
you
open
the
notebook
with
the
notebook
View
and
if
you
open
the
same
notebook
with
the
file
editor,
for
example,
if
you
use
RTC,
it's
not
gonna
work,
because
the
the
the
Delta
are
not
consistent
between
them.
F
B
C
That
that
has
been
one
of
the
longest
and
like
most
severe
complaints
of
people
about
about
Jupiter,
like
literally
we've
been
hearing
that
one
for
10
years
and
we've
been
telling
people
it's
going
to
be
fixed
one
day,
it's
going
to
be
fixed.
Relegating
that
to
a
secondary,
like
non-sort
of
non-standard
installation,
seems
like
an
issue
to
me.
At
least
it
would
give
me
a
pause
actually.
F
So
what
but
my
question
then
is:
do
we
expect
to
fix
that
during
the
4.x
release
cycles
and
if
so,
I
I
think
that's
a
reason
to
consider
having
this
be
the
default
like
it
should
still
work
without
it.
But
if,
if
we
can
fix
that
long-standing
bug
in
the
4.x
series,
Without
API
breakage
I
I
mean
it
literally,
is
the
top
bug
in
all
of
Jupiter.
B
B
H
B
I'm
probably
leaking
the
the
all
the
old
knowledge
on
the
Kernel
site,
because
that's
that's.
Basically.
The
trouble
for
now
with
the
output
is
that
we
send
from
the
front-end
the
request
to
execute,
and
then
it's
only
the
case
that
is
asked
for
execution.
That
will
get
the
reply
from
the
kernel,
and
so
when
you
get
the
reply,
it
creates
the
output
and
then
the
output
will
be
sent
to
the
RTC
Machinery
to
dispatch.
The
update
of
the
document
to
everybody.
H
William,
of
course,
has
lots
of
things
to
say
about
this.
He's
been
struggling,
I
mean
not
struggling,
he's
been
working
and
implementing
this
for
years
now,
without.
E
I
E
But
my
number
one
remark
is:
make
sure
you
try
to
add
in
some
sort
of
logging
Beyond
what's
there
so
that,
inevitably,
when
issues
arise,
you
can
get
more
than
it's
broken.
You
can
get
like
some
actual
trace
of
what
happened.
People
can
provide
you
something
and
then
you
can,
you
know
kind
of
fix
it.
A
E
But
it's
worth
that's
exactly
what
code
calc
does
like
any
random
client
just
changes,
something
saying
I
would
like
this
cell
to
be
executed
and
then
the
back
end
server
tells
the
kernel
to
actually
do
the
execution,
but
it's
a
little
tricky
to
do
efficiently.
There's
a
whole
bunch
of
subtleties
there,
just
making
sure
that
you
don't
sort
of
waste
time
in
in
the
back
and
forth.
That's
involved
for
a
while.
It
felt
pretty
laggy
and
could
talk
the
way
I
had
done.
That.
A
Hand
is:
can
we
move
this
package
from
Jupiter
server,
the
Jupiter
lab
on
github.org.
A
F
A
B
So
thanks
for
the
vote,
the
the
other
question
I
have
before
four
is
also
about
the
two
render:
let's
collect
it
like
that
that
are
in
the
in
core.
Currently,
it's
a
vegan
package
and
video
package
I
propose
to
move
them
to
the
Jupiter
renderer,
not
for
the
same
reason.
Vidam
is
because
vidam
is
not
maintained
any
longer
and
there
has
been
a
boat
update
for
I
widget
eight,
and
it
seems
that
when
we
update
to
ipy
widget
8,
then
we
don't
doesn't
work
any
longer.
B
I
didn't
look
at
the
detail,
but
that
one
seems
to
not
be
compatible
any
longer
and
Vega.
It's
it's
related
to
some
recent
contribution
that
proposed
to
update,
because
there
is
a
new
version
of
Viga
via
I,
think
we
call
it
five.
I
I
can
help
out
with
this,
since
I
did
something
I
am
doing
something
very
similar,
which
is
kind
of
doing
the
inverse
of
that
moving
the
math
Jax
3
package
away
from
Jupiter
renderers
into
Jupiter
lab
so
with
the
Vega
extension,
where
we
need
to
add
a
version
number
to
the
package
so
that
way
users
can
choose
which
version
they
want
to
install
or.
F
What
would
are
we
also
proposing
that
as
a
Jupiter
lab,
4
jupiterab
will
ship
with
the
Vega
extension
by
default.
B
That's
my
question:
yes:
should
we
ship
one
or
not,
basically
well.
F
I
have
mixed
feelings
on
this,
my
I'm
one
of
the
co-creators
of
Altair,
and
so
we
we
really
wanted
to
advocate
for
Vega
as
a
multi-stakeholder
open
specification
for
visualizations
that
it
was
not
wasn't
sort
of
owned
by
a
single
party,
and
we
felt
like
that
was
important.
At
the
same
time,
Jupiter
traditionally
is
not
bet
on
a
single
I.
Think
that
there's
a
wrong
word:
it's
not
a
product
I.
F
D
F
F
Yeah
I
I
think
I
mean
that
Altair
is
very,
very
popular
these
days
inside
I'm.
Guessing
it's
getting
a
lot
of
usage
at
the
same
time.
I
do
know
that
that
the
Altair
folks
have
worked
hard
to
make
sure
that
Altera
Works
through
other
display
mechanisms,
the
the
existing
Vega
extension
Works
through
a
mime
type,
whereas
Altair
now
has
pretty
solid
support
for
playing
being
able
to
display
using
a
standard
JavaScript
one
type
as
well.
So
maybe
it's
not
the
end
of
the
world.
B
So
I
I
will
open
an
issue
on
the
video
specifically
too
so
that
we
can
can
discuss
and
Trace
that
thanks
on.
F
D
A
Fred
sorry
I'm
losing
my
voice.
Actually
I
don't
want
to
mask
Brian
Europe.
F
Yeah
hi
everyone,
so
we
have
I'm
gonna
paste
this
into
the
chat.
My
team
at
AWS
has
been
working
on
this
generative
AI
extension
for
Jupiter
lab
that
enables
end
users
to
perform
tasks
using
these
generative
AI
models;
tasks
such
as
selecting
a
block
of
code
and
reef
having
the
model
refactor
that
code
or
debug
it
or
create
a
markdown
cell.
That
explains
the
code.
F
Originally
we
were
proposing
to
Incorporated
sort
of
permanently
into
the
Jupiter
lab
org.
After
talking
more
with
a
number
of
you,
and
also
some
folks
from
the
software
steering
Council
in
EC,
our
sense
is
that
for
the
time
being,
the
best
place
for
this
is
the
Jupiter
incubator
program.
This
is
certainly
early
stage.
F
Exploratory
work,
it's
not
stable
mature,
so
we,
the
the
only
hitch
with
that,
is
that
the
incubator
program
is
sort
of
on
pause
as
and
hasn't
been
updated
to
the
new
governance
model,
and
so
the
EC
and
SSC
are
going
to
work
over
the
next
few
months
to
relaunch
that
program
and
I
think
it.
That
would
be
a
good
place
for
this.
F
Our
what
I've
done
is
I've
changed
the
vote
to
be
a
temporary
adoption
of
Jupiter
lab
of
this
repo
to
enable
us
to
work
in
the
open
and
collaborate
with
folks,
but
only
until
the
incubator
program
is
launched,
at
which
point
we
would
move
it
over
to
the
the
incubator
program
and
yeah
any
questions.
Thoughts.
J
As
one
of
those
people
in
that
long
discussion
about
whether
it
should
be
in
the
incubator,
thanks
for
the
update
it's
now,
a
yes
for
me
and
I
was
definitely
leaning
towards
a
solid
no
before
so.
C
And
perhaps
that
have
a
context
for
folks
who
were
not
yesterday
in
the
office
hours
of
the
executive
Council,
we
discussed
this
quite
a
bit
there
and
the
this.
The
idea
is
that
this
should
serve
as
a
good
kind
of
prompt
to
really
update
the
incubator
like
model
and
Machinery.
It's
not
quite
ready.
So
there's
no
point
in
really
sending
like
sending
this
package
out
there
when
that
mechanism
right
now
has
a
bunch
of
things
that
like
are
not
even
viable
because
they
refer
to
the
previous
governance,
but
it
we.
C
We
do
think
that
that
mechanism
makes
sense
for
the
project
at
large,
and
this
will
be
an
incentive
to
do
that.
On
Friday
we
have
the
first
joint
meeting
of
the
EC
and
the
new
SSC.
This
would
fall
kind
of
squarely
under
the
the
purview
of
activities
of
the
SSC,
so
it'll
it'll
be
a
good
motivator
to
to
get
get
that
particular
Wheel,
In,
Motion
and
and
the
mechanisms
for
the
EC
to
coordinate
that
with
the
SSC
are
like,
as
I
said.
C
Our
first
meeting
is
in
two
days,
so
I'm
kind
of
optimistic
that
this
is
a
pretty
reasonable
thing
to
do
and
that
it's
a
good
forcing
function
if
nothing
else
to
say
well
again,
let's
get
this
done
so
that
this
one
and
others
another
one
was
already
mentioned
in
the
issue.
How
Steve
mentioned
mentioned
two
packages
from
Jupiter
server
that
are
being
handled
similar
so
that
they
all
have
a
good
home.
So
that's
just
kind
of
to
give
a
bit
of
context.
A
Just
a
procedural
question:
can
we
consider
an
alternative
to
the
list
of
everyone's
names
and
yes,
no
abstain
way
of
voting,
even
something
that
is
as
simple
as
switching
to
emojis
or
something
I?
Don't
know?
I
just
I've
always
found
this
watching
the
votes
trickle
in
process
a
little
discomforting
myself.
F
K
J
I
I
actually
just
thought
of
like
two
points
around
the
where
it
goes
side
of
things,
while
it's
temporarily
in
Jupiter
lab
org.
Do
we
want
to
Market
in
some
way
as
incubating
like
in
the
name
in
the
docs.
F
I
was
thinking
that
I
think
both
in
the
subtitle
of
the
repo
at
the
top
of
the
readme
yeah.
We
should
Mark
it
as
incubating
an
experimental
that
that
completely
makes
sense.
J
And
then
the
second
thing,
I
guess,
should
be
forwarded
to
the
steering
Council
when
they
discuss
this,
whether
or
not
incubating
projects,
because
I'm
assuming
they're
going
to
be
under
Jupiter
incubating
if
they
should
be
somehow
earmarked
to
which
org
they're
trying
to
get
into
out
of
the
incubator.
F
Yeah
we
so
they
the
interesting.
The
previous
incubator
required
the
incubating
project,
to
name
a
sponsor
on
the
Jupiter
steering
Council
and
something
that
we
started
to
talk
about
in
the
executive,
Council
Office
hours.
Yesterday
was
exactly
what
you're
saying
to
say:
let's
replace
that
by
a
sponsoring
sub-project
that
would,
in
most
cases,
probably
be
the
eventual
Target
of.
F
J
Yeah
I've
worked
with
both
Apache
and
Linux
Foundation
incubator
projects
and
I
know
in
Apache.
It's
just
all
one
tunnel
goes:
there's
there's
no
sub
anything,
but
with
Linux
each
project
has
its
own
incubator,
which
is,
in
my
opinion,
total
Overkill,
and
it
has
its
own
problems.
So
I
do
like
this
idea
of
having
one
incubator
but
making
sure
they
have
paths
towards
specific
projects
and.
F
If,
if
folks
have
input
about
the
incubator
program
or
how
it
should
work,
we
will
likely
be
discussing
that
in
the
executive,
Council
Office
that
are
Tuesday
Mornings
at
10,
A.M
Pacific,
and
so
you
know,
feel
free
to
jump
on
there
and
chat
with
the
EC
and
others
about
this.
C
And
Brian,
it
just
makes
me
think
that
we
should
and
I
can
take
care
of
that
open,
an
issue
kind
of
about
the
future
of
the
incubation
product
programming
in
there,
so
that
people
can
leave
comments
asynchronously
and
whatnot,
but
I'm
going
to
drop
I'm
going
to
drop
the
link
to
the
office
hours
so
that
people
have
it
handy.
It's
in
the
calendar,
but.
F
And
I
I
think
we
we
would
love
to
see.
I
think
this
is
the
I'm
speaking
on
behalf
of
the
EC
we'd
love
to
see
the
incubator
program
be
fairly
lightweight,
so
that
the
the
bar
for
for
projects
to
incubate
is
not
really
high
and
people
don't
do
it
I
think
in
the
past.
It
probably
was
that
way,
and
people
felt
like
it
was
too.
F
And
you
know,
I
think
a
model,
that's
more
closely
aligned
with
how
some
of
the
contrib
orgs
are
working
where
it's
lightweight.
F
G
A
That
how
about
we
do
a
thumbs
up
smell
test
and
if
it
turns
out,
we
have
consensus,
we
don't
need
to
know
if
we
don't.
We
do
so.
Who
is
okay
with
the
proposal
to
take
the
generative
AI
repos,
moving
to
Jupiter
incubator,
with
the
understanding
that
we,
the
Jupiter
lab
project,
are
sponsoring
that
and
that
the
graduation
from
the
incubator
will
be
to
one
day
reside
in
the
Jupiter
lab.org.
A
D
F
A
F
All
right,
so
all
I've
actually
already
updated
the
vote,
so
it
will
close
next
Wednesday
into
the
day
anywhere
on
Earth.
So
thank
you.
Everyone
really
appreciate
the
the
input
and
helping
us
get
through
this.
A
Great
thanks,
Brian,
okay,
let's
move
on
the
next
person
on
the
agenda
is
Jason
Moyle.
K
Hey
good
morning,
good
evening,
everyone,
so
we've
got
right
now
over
a
hundred
issues
that
are
open
against
Jupiter,
Labs,
Milestone
400.,
we're
hoping
to
have
this.
The
the
Milestone
I've
been
using
or
the
the
timeline
I've
been
using
is
that
we
want
to
have
4-0
out
and
generally
available
by
jupitercon,
which
is
in
early
May.
K
So
the
question
then
becomes
like
how
are
we
gonna
decide?
What
is
gonna
go
in
and
what
is
not-
and
we've
talked
last
week
about
devoting
a
half
hour
of
this
hour-long
session
to
triage
we're
already
at
35
minutes
past
the
hour,
so
I'm
not
sure
how
much
time
people
want
to
dedicate
for
real,
because
we've
also
got
some
some
other
items
on
the
agenda.
What
do
folks
think
want
to
do
like
15
minutes,
10
minutes.
K
Let's
proceed,
then,
with
Darian
your
next
step
on
the
list.
A
Okay,
I
linked
a
PR
that
has
a
change
in
the
running
sidebar
for
kernels.
That
I've
been
working
on
recently
with
a
lot
of
help,
and
the
main
thing
to
point
out
is
that
the
concept
of
what
you
see
in
the
running
kernels
drop
down
in
Jupiter
lot.
3.X
has
always
been
active
sessions
connected
to
a
kernel.
So
here's
a
notebook
connected
to
a
kernel.
A
So
you
would
see
the
notebook
and
here's
a
console
connected
to
a
notebook,
sorry
tool
kernel,
and
you
see
that
and
if
you
have
two
active
sessions
which
you
can
do
by
opening
a
notebook,
then
independently
opening
a
console,
then
going
to
that
console
and
saying
switch
to
using
the
kernel
for
the
notebook
that
I
opened
previously
you'll
see
two
items
under
the
list
of
kernels,
even
though
it's
actually
one
kernel,
and
so,
if
you
hit
the
X
on
the
console,
you'll
actually
kill
the
kernel
in
your
notebook
as
well.
A
So
now
this
is
the
UI.
Instead
shows
a
list
of
running
kernels
and
there's
there's
an
indented
tree
structure
where
beneath
each
kernel
shows
the
active
sessions
for
that
kernel,
and
you
can
kill
the
kernel
with
the
understanding
about
everything
in
that
all
of
its
child
activities
are
going
to
go
with
it
and
also
now.
This
isn't
a
feature
that
ships
with
future
server.
A
But
if
you
have
a
server
like
say
jupiterverse
that
when
you
hit
the
API
kernels
rest
endpoint,
it
lists
kernels
that
might
not
be
ones
created
by
the
Jupiter
server
might
be
running
in
another
application.
If
it
has
some
way
of
picking
those
up,
it'll
display
those
kernels
for
you
too,
and
there's
a
new
context,
menu
on
running
kernels.
You
can
right
click
on
a
running
kernel
and
say
give
me
a
new
notebook
connected
to
this
kernel
or
give
me
a
new
console
connected
to
this
kernel.
A
This
isn't
a
feature,
we're
gonna,
being
super
quick
to
release.
I.
Imagine
we'll
have
to
talk
about
it
in
Jupiter,
server
and
it'll
at
the
very
least,
be
a
a
point
release
a
minor
release
because
it's
strictly
speaking,
it's
a
Well.
Actually,
it's
not
an
API
change,
but
it's
a
behavior
change,
so
we
probably
put
it
behind
a
2.1
or
something.
But
you
don't
need
that
feature
to
still
see
a
newly
reorganized.
A
Kernels
panel
in
the
left
hand,
sidebar
and
that's
it
it'll,
be
in
the
next
Alpha
and
you
can
play
with
it
and
if
you
don't
want
to
play
with
it,
if
you
go
to
my
PR
I
made
a
screencast,
so
you
can
see
a
video
of
what
it
looks
like
and
that's
all
I
had
so
Andrew.
You
are
up.
L
Okay,
so
there
was
a
live
discussion
with
Mike,
so
Mike
says
that
he,
like
already
replied
and
as
far
as
I
understood,
feels
so
self-requested
the
review
just
now
so
I
think
we
can
skip
my
topic
and
we
can
take
it
offline
and
rather
just
to
spend
time
on
Treasury
4.0
issues.
Maybe
so
thanks.
Everyone.
A
Okay,
great,
we
do
have
a
bullet
point
under
additional
discussions
who
who
added
this,
and
is
this
something
that
we
should
talk
about
now?
Should
we
leave
it
for
next
week?
What
do
you
think.
K
Frederick's
name
is
on
13746
the
issue.
That's
listed,
I,
don't
know
whether
Frederick
actually
added
that.
A
Okay,
why
don't
we
take
this
up
next
week?
Then
it's
no
one's
claiming
it
now's,
not
the
time
so
I'll
pass
it
over
to
Jason
w.
K
Hello
everyone
once
again,
as
I
mentioned
before,
we
had
135
issues
that
are
open
against
the
400
Milestone
and
we
are
using
Jupiter
Khan,
which
is
in
early
may,
as
our
kind
of
guidepost.
For
when
we'd
like
to
have
a
releasable
version.
Releasable
version
doesn't
necessarily
mean
we
have
zero
bugs
in
it
or
zero
items
in
it.
But
we
might
want
to
take
a
look
I've
been
kind
of
focusing
on
the
bugs
first,
but
there
are
also
maintenance
and
enhancement
requests
in
there.
K
I've
done
a
little
bit
of
spot
triage
myself,
but
I'm
open
to
suggestions
as
to
how
we
should
attack
some
of
the
items
that
are
in
need
of
review
on
this
call,
and
by
review
I
mean,
should
we
accept
them
continuing
to
be
on
400
or
should
we
move
them
to
another
future
milestone.
L
K
A
K
The
the
the
search
Syntax
for
is
unassigned
if.
K
Good
idea,
so
yeah
I
I
actually
went
through
some
of
the
really
old
ones
a
while
back,
and
some
of
them
were
actually
already
fixed,
but
many
of
them
are
not
so
this
is
the
oldest
one,
which
is
tagged,
still
needs
discussion,
even
though
it's
over
five
and
a
half
years
old
at
this
point,
but
versioning
document
models.
So
is
this
still
something
that
we're
targeting
for
4-0?
K
K
Okay,
thank
you
so
keyboard
shortcuts
this
person,
Wes
Turner,
wants
a
a
list
of
keyboard
shortcuts
to
pop
up.
Someone
is
working
on
integrating
a
more
graphical
keyboard.
Shortcut
editor,
more
similar
project
would
be
to
make
make
a
table
existing
keyboard
shortcut.
So
this
is
Jason
crowd's.
Last
comment
from
September
2019.
K
there's
a
pull
request
that
was
opened
in
draft
in
August
of
last
year
that
fixes
4004
and
I
asked
last
week.
Is
this
still
targeted
for
this
version,
but
I
didn't
get
any
responses.
H
No
there's
your
response.
I
can't
pick
it
up,
so
if
somebody
else
can
pick
it
up
great,
but
it
looks
like.
K
The
request
is
to
add
a
keyboard
shortcuts
user
interest
that
displays
in
the
same
way
as
the
famous
question.
I
G
B
H
And
the
idea
here
is:
these:
are
keyboard
shortcuts
that
are
context
context
aware
of
wherever
you
are
in
the
UI,
so
wherever
you
are
in
the
UI,
you
can
hit
this
question
mark
and
it'll
say
what
keyboard
shortcuts
are
available
to
you
right
now,
since
we
have
keyboard
shortcuts.
That
can
depend
on
where,
where
you
are.
I
H
C
K
Okay,
all
right,
so,
let's
move
on
to
four
two
five,
four
another
one
by
Jason
grout
by
the
way,
if
you
enjoy
this
plug,
for
we
do
triage
every
week
for
new
issues
on
Thursdays
at
10,
A.M
Pacific
would
happy
to
have
would
be
happy
to
have
you
there
offer
open
with
in
the
tab
context
menu
for
documents.
K
We
can
click
on
the
document
Tab
and
get
a
new
view
with
the
same
viewer
Matthias
points
out.
It
would
be
a
very
natural
workflow
to
get
an
open
with
option
to
easily
look
at
an
open
file
with
a
different
viewer.
C
Which
is
horrible
ux,
because
if
your
top,
if
your
file
browser
is,
if
you
have
a
document
from
a
different
location
of
the
file
system,
you
basically
need
to
go
to
the
file
browser
re-navigate
to
the
place
where
that
thing
was
opened,
originally
open
it
and
then
you've
lost
your
location
in
the
file
browser.
This
has
been
a
big
pet
peeve
of
mine
when
using
it
and
teaching,
and
so
I
I
would
root.
I
would
root
very,
very
partly
for
this
getting
fixed.
It's
kind
of
a
massive
pain
in
my
ass
okay.
K
So
with
that
endorsement
in
mind,
Jeremy
commented
last
March
if
anyone
would
like
to
work
on
this
included
in
the
40
Milestone.
So
now
comes
the
conversation
of
like.
Would
anyone
like
to
work
on
this.
A
A
H
A
H
Give
you
my
guess,
is:
there's
a
lot
of
these
issues
like
this,
and
maybe
it's
better
for
us
to
asynchronously,
say
hey.
Does
anybody
want
to
pick
up
any
of
these
issues
in
the
next
week
and
if
no
and
whoever
has
not
picked
up
an
issue
in
the
next
week,
gets
moved
to
the
future
Milestone
or
something.
K
H
Between
now
and
next
week,
if
they're
committing
to
work
on
it
in
the
next
four
weeks,
I
think
that
would
be
the
first
sign
of
commitment
to
work
on
it
in
the
first,
the
next
four
weeks.
We
could
just
take
these
issues
and
and
mass
move
them
to
a
new
Milestone
which
is
4.1,
probation
or
provision,
or
something
provisional
or
something
I.
K
J
D
K
H
Right
so
so
I
think
that
I
mean
we've
got
four
weeks
to
work.
If
you,
if
you
know
you
can
commit
in
the
next
four
weeks,
then
great,
because
really
that
means
maybe
two
weeks
and
then
fixing
bugs
and
stuff
like
that,
so
I
would
suggest
having
an
opt-in
versus
an
opt-out
sort
of
philosophy.
H
K
I
I
K
A
E
A
Oh,
we
should
do
this
open
with
context
menu
for
the
tabs
in
Jupiter
lab.
Whoever
raises
that
issue
as
a
5.0
like
to
have
is
going
to
find
it
as
a
pre-written
issue
sitting
in
that
Milestone.
So
you're
right,
there's
a
lot
in
there
and
many
will
never
actually
be
implemented.
But
almost
all
of
the
next
versions
features
are
probably
also
in
there
and
they
will
be
extracted.
L
I
had
a
question
from
this
110
open
issues
are
all
issues
like
a
priority
for
a
4.0,
because
this
is
the
biggest
question
for
me
whenever
I
look
through
the
issues
to
choose
an
issue
to
work
on
like
I.
Imagine,
for
example,
like
mobile
problems
are
as
priority
as
like
keyboard,
shortcuts
problems.
So
I
don't
know
from
my
point
of
view.
It
would
be
really
useful
to
throw
out
issues
that
are
not
priority
for
a
4.0.
I
L
A
So
I
think
the
ones
if
there
are
any
left
that
have
feature
parity
labels.
It
was
a
pretty
high
priority
because
we've
always
intended
Jupiter
laptop
feature,
parity
with
notebook,
and
so
this
keyboard
overlay
thing
has
been
an
oversight
for
quite
some
time,
simply
because
notebook
has
had
the
feature.
I
don't
know
if
we
have
other
feature
parody
ones
there.
But
that's
that's
a
good!
That's
one
good
metric,
okay.
C
If
I
may
have
a
comment,
I
I,
when
I
think
of
this
problem,
I
actually
think
of
two
different
criteria.
Priority
is
a
ranking
decision,
you're
ranking
things,
and
this
is
hyper
and,
as
Vixen
said,
different
people
have
different
senses.
But
I
think
to
me,
what's
important
is
to
know.
Is
it
is
an
issue
high
impact
for
someone.
Now
the
team
may
say,
look
we're
sorry.
C
We
can't
fix
it,
but
knowing
whether
it
has
high
impact
for
a
specific
use
case
helps
inform
whether
to
prioritize
it
or
not,
right
and
so
I
think
it
would
be
valuable
to
say
some
of
these
issues
we
get
I
mean
would
be
nice
for
that.
But
it's
going
to
be
others.
They
are
very
high
impact.
It
would
be
good
if
it
could
be
prioritized.
It
doesn't
make
an
automatically
high.
C
It
signals
and
helps
making
that
decision
and
I
think
understanding.
The
difference
between
those
two
is
useful,
for
example,
in
the
open
with
one
like
I
can
demonstrate
that
it
is
high
impact
for
a
use
case.
It
may
not
matter
for
a
different
one
right,
but
signaling
that
can
help
them
make
the
the
prioritization
decision
for
the
entire
team
on
like
what
is
considered
a
blocker
and
whatnot.
Where
should
we
like
all
hands
on
deck
and
fixed
ad
versus?
You
know,
it'll
come
when
it
comes
at
nice
for
a
Sprint
or
whatever.
K
Foreign
okay,
we
are
currently
at
53
minutes
past
the
hour.
I
wanted
to
close
out
this
conversation
that
we'd
started
about
4254,
and
then
we
can
consider
going
forward
with
like
this.
This
Mass
move
proposal
of
like,
if
you
haven't,
claimed
something
we're
just
going
to
move
it
to
this
provisional
Milestone,
I'll,
say
six
days
from
now,
so
that
we
can
revisit
this
next
Wednesday.
K
Also
Jason
like
how
do
we
inform
everybody
in
this
community?
Hey
you
have
six
days
to
to
claim
something
by
assigning
it
to
yourself.
B
It
should
be-
and
it's
a
bit,
but
it's
a
bit
tricky
because
no
GitHub
allows
a
user
to
be
assigned
if
he
has
at
least
commented
once
so.
For
example,
if
you
are
a
new
contributor
and
you
never
have
commented
on
an
issue
you
want
to
work
on,
you
have
at
least
to
make
one
comment
on
the
issue
and
then
we
may
create.
We
may
allow
the
boat
to
to
assign
issue.
H
K
H
G
Have
a
4.0
issue
already
for
Jupiter
lab
that
people
watch
even
if
they
don't
comment,
it
won't
catch
everyone,
but
it'll
catch.
Some
asynchronous
people
do
okay,.
K
Can
you
drop
a
link
to
that
in
the
chat
please
Isabella
yep?
Thank
you
so.
M
H
M
Or
so
like
in
these
hundreds
issues,
hundreds
of
issues
there
are
like
maybe
100
issue,
they
are
probably
going
to
be
very
few
that
are
going
to
be
picked
up
and
result
by
the
50
freeze.
So
jupitercon
is
not
just
a
deadline
for
releasing
Jupiter
lab
four
like.
Presumably,
you
should
release
Jupiter
for
long
enough
before
GP
to
con
that
we
can
also
have
notebook
7
by
then
and
a
bunch
of
extensions
so
like
for
me,
the
calendar
is
really
like.
M
We
mostly
work
on
consolidation
right
now,
I
have
a
beta
in
the
end
of
the
month.
Then
we
have
a
few
weeks
to
go
through
the
release
cycle
through
the
you
know,
release
candidate
Cycles
so
that
in
end
of
March
earlier
Pro
we
have
a
final,
four
and
and
move
on
to
upgrading.
All
of
the
extensions
and
release
notebook
7.,
like
the
the
the
schedule,
is
very
tight,
and
if
any
of
these
issues
require
significant
work,
then
it's
very
very
unlikely
that
any
of
this
will
Land
by
four.
A
So
I
think
that
what
Mike
is
saying
in
the
chat
makes
some
sense.
Basically,
we
should
go
through
the
issues
painfully
on
a
call,
because
that's
how
to
get
the
most
exposure
to
them.
A
That's
how
to
find
out
that
there's
no
actual
interest
in
doing
them,
and
we
should
probably
do
this
a
couple
more
times
between
now
and
the
end
of
the
month
and
at
some
point
switch
not
just
to
looking
at
who
wants
to
claim
this
for
the
end
of
the
month
to
these
open
issues
that
have
assignees
hey,
you
get
a
finish
by
the
end
of
the
month.
You
know
maybe,
like
we
have
a
February
four,
it
was
the
15th
check-in
or
something
that
is.
You
have
two
weeks
until
future
freeze.
A
Do
you
think
you're
gonna
make
it
if
you're
not
that's
totally
fine
just
do
it
either
in
a
patch
release.
If
that's
what
you
think
or
if
it's
a
feature
and
you're
happy
to
wait,
the
4.1
move
it
to
that,
but
but
we
probably
ought
to
go
through
the
slow
process
of
lots
of
us.
A
Looking
at
these
because
I
don't
I
also
am
skeptical
that
people
are
gonna
just
draw
through
that
issue
list
and
and
do
it
on
their
own,
and
we
are
talking
a
tight
time
frame
that
Sullivan
says
yeah,
so
I
propose
next
week.
We
basically
do
this
exercise
again
and
the
week
after
we
do
this
exercise
again
and
the
week
after
I'm
off,
so
you
guys
can
do
whatever
you
want
and
the
week
after
that.
A
We're
done
right
because
that's
the
end
of
the
month.
K
So
to
be
clear
because
we're
nearing
the
top
of
the
hour
here,
I'm
going
to
move
all
unassigned
400
issues
to
an
I'm
going
to
call
it
new,
Milestone,
provisional,
Milestone
and
then
next
week
we're
going
to
review
that
at
some
point,
we're
gonna
have
to
start
reviewing
presently
assigned
issues
on
400
as
well.
A
Well,
maybe
4.1
is
a
good
place
to
put
in,
because
anything
in
4.1
is
something
that
by
Nature
can
go
into
4.0
before
we
finalize.
So
all
these
issues,
we
can
optimistically
say
they're
4.1
issues
unless
somebody's
volunteering
to
do
them
in
the
next
27
days,
and
we
don't
need
to
create
a
new
milestone
for
that
and
once
4.0
is
out,
they'll
already
be
in
the
right
place.
K
A
I
M
Yeah,
okay,
okay
and
if
any
of
you
feel
like
taking
on
one
of
these
like
prioritize
books
because
like
there
is
a
temptation
to
say
yeah
I'm
going
to
squeeze
in
some
new
feature
by
4.0,
but
I.
Think
if
we
continue
delaying
I
mean
it's
been
two
years
two
and
a
half
years
in
the
making.
If
we
continue
delaying
4.0,
the
pace
of
development
is
just
going
to
slow
like
at
some
point.