►
From YouTube: IPython/Jupyter Dev Meeting, June 21, 2016
Description
Meeting of the IPython/Jupyter development team.
Meeting Notes: https://jupyter.hackpad.com/June-2016-Q7Kx6WRqYKg
A
Hello,
everyone:
this
is
the
June,
twenty-first
Jupiter
ipython
dev
meeting
and
we'll
get
started
with
sort
of
in
order,
whoever
has
anything
that
they
want
to.
They
want
to
bring
up.
First
of
all,
I
wanted
to
Peter
thanks
for
the
graduation
of
or
the
proposal
or
for
Jupiter,
an
incubator,
graduation
of
the
dashboards
and
and
the
dashboard
releases
that
you
made.
That's
that's
actually
very
exciting,
so
we'll
be
working
on
that.
But
I
really
appreciate
it.
You
guys
have
done
a
terrific
job
on
that
front
and
I
encourage
folks
to
comment
and
whatnot
I.
A
Think
Peter
and
his
team
have
been
really
really
very
diligent.
With
with
the
incubation
process
and
I
think
it's
been
an
example
of,
I
hope
the
success
of
that
model
so
far
I
think
it
has.
It
has
worked
well
reasonably
well
with
them,
even
if
we
haven't
always
been
as
fast
that
we
should
so
I
really
appreciate
it.
I
don't
know
if
you
have
any
comments,
anything
that
you
wanted
to
discuss
on
that
front.
Peter
and.
B
C
A
D
Sorry,
probably
the
only
thing
is
I'm
going
to
try
and
talk
their
culture
tomorrow,
because
he
reached
out
there
trying
to
submit
a
sloan
grant
proposal
for
read
the
docs
supporting
science
better.
So
I
will
send
an
invite
along
to
you
and
Brian
and
Jamie,
and
whoever
else
has
interest,
but
this
is
just
sort
of
a
very
much.
This
is
just
a
very
introductory
thing
just
to
get
a
sense.
They
want
to
get
a
sense
of
what
we
would
like
so
oh
yeah,
and
that
is
about
it.
A
D
A
A
Ok,
I
guess
I
only
had
a
couple
of
a
couple
of
things
to
mention
will
be
finishing,
will
be
trying
to
finish
the
Jupiter
hot
mini
workshop
planning.
I
posted
a
link
on
that
to
the
to
the
on
the
on
the
hack
pad
I,
don't
know
if
anyone
has
has
anything
else
of
note
on
that
front,
there
will
be
I.
Think
we'll
try
we'll
try
to
do
it.
I
looked
into
both
blue
jeans
and
zoom
that
have
have
up
to
a
hundred.
The
zoom
account
that
we
have
here
at
lbl
has
up
to
50
people.
A
Bye-Bye
video
and
blue
jeans.
I
found
has
up
to
100
people
by
video,
so
what
I'm
thinking
is
actually
doing
it
that
way
doing
it,
either
via
zoom
or
blue
jeans,
and
maybe
we'll
do
some
tests
before
then,
because
it
it
might
actually
work
better
than
then
trying
to
do
a
hangout.
That
way
with
with
one
of
these,
everyone
who
joins
in
to
the
call
can
actually
type,
and
if
we
say
we
can
have
up
to
50
people
that
50
50
people
live.
Obviously
the
recording
afterwards
can
be
watched
by
anyone.
A
We
lose
live
streaming,
but
what
we
gain
instead
is
actually
many
more
people
simultaneously.
So
I
was
thinking
that
it
might
actually
be
a
better
solution
than
trying
to
do
a
weirdo
hybrid
between
Google
Hangouts
and
this,
and
that
I
was
starting
to
feel
that
we
had
too
many
too
many
moving
parts
in
the
air.
So
I
looked
a
little
bit
into
our
options
and
I.
Think
blue
jeans
resume
will
probably
work
better.
A
And
then
we
have
we
have.
We
continue
to
work
on
Jupiter
hub
in
education
here
at
Berkeley,
with
the
ramp
up
of
the
data
eight
stuff.
Here,
that's
that's
becoming
sort
of
a
significant
effort.
I
will
have
kind
of
more
to
report
and
then
I
just
wanted
to
put
out
there.
We
had
another
chat,
just
randomly
on
other
stuff.
Jamie
and
I
were
walking
with
Brian
and
we
sort
of
came
back
to
our
our
never-ending
communications.
Questions
and
and
I
am
brought
up
again
the
idea.
A
Well,
maybe
we
should
just
actually
kill
Gator
all
together
and
force
ourselves
to
work
on
on
something
like
the
Jupiter
help
panel
and
maybe
potentially
what
I
was
thinking
was.
Maybe
we
should
make
and
we
could
even
make
a
new
org
called
Jupiter
channels
or
something
like
that,
where
literally
every
repo
is
just.
You
want
to
chat
with
someone
open
a
thread
on
it,
King
them
and
close
the
thread
when
you're
done,
and
it
might
actually
give
us
a
combination.
I
know
the
javascript
world.
A
Does
it
that
way,
or
at
least
parts
of
the
know,
Jas
community?
Do
it
that
way
and
they
say
that
it
has
actually
helped
them?
It
gives
threaded
conversations,
because,
basically,
you
open
a
chat
on
anything
and
you
put
it
at
least
you
say
I
want
to
talk
about
this
with
you
and
you
and
it's
just
typing
a
title,
but
it
means
that
conversation
is
recorded.
You
can
refer
to
it
by
URL
later
and
tell
Slater,
hey,
Jason
and
I
are
Carol.
A
Jason
and
I
discussed
this
thing
and
it
turned
out
that
you
might
be
interested.
Here's
the
URL
and
Carol
could
go
a
month
later
and
read
it,
whereas
taking
that
action
is
completely
impossible
on
here.
We,
you
can't
tell
anyone,
go
fish
in
the
Gator
intertwined
logs
of
a
month
ago,
and
so
I
don't
know.
We
don't
need
to
make
a
decision.
A
I
just
wanted
to
put
this
back
in
people's
brains
for
us
the
simmer
for
a
few
weeks,
but
I
think
that
I
was
rereading
a
little
bit
called
Fogle's
open
source
book
a
few
days
ago,
and,
and
one
of
the
points
that
he
really
stresses
is
the
avoid
private
conversations
as
one
of
the
key
issues
that
is
sort
of
a
community
killer
in
open
source
projects
and
I.
Think
in
in
practice.
A
Gator
has
the
unintended
effect
in
practice,
even
if
it's
not
private
and
principal
of
creating
the
fact
of
private
conversations
just
because
of
the
the
velocity
of
them
and
the
being
buried
in
a
hard
to
disentangle
log.
For
all
intents
and
purposes,
a
chat
between
a
two
or
a
few
people
on
gara
that
goes
by
quickly
becomes
a
private
discussion
where
people
might
have
made
decisions,
and
we
haven't
been
successful
at
not
doing
that.
A
And
so
maybe
the
only
solution
really
is
to
just
call
it
off
completely
and
force
ourselves
to
say
now:
use
use
github
if
you're
chatting
with
someone
actually
being
parked
on
a
page.
It
refreshes
almost
at
the
speed
up,
get
up,
get
er,
maybe
not
identically,
but
almost
as
fast,
and
it
does
also
have
the
advantage
that
if
you
get
pinged
on
it,
you
get
email
on,
so
you
can
respond
by
email
and
if
you
decide
to
mute,
you
take
yourself
out
of
the
conversation
and
that
threat
is
gone
from
your
inbox.
A
I,
don't
know
something
to
think
about.
We
don't
need
to
make
a
decision.
I
know
this
is
contentious.
Uv
told
us
how,
in
other
organizations
he's
seen
these
discussions
become
very
contentious,
but
I
just
wanted
to
bring
that
up.
I
think
we're
hurting
our
collective
community
community
memory
and
community
engagement
by
by
by
our
communications
practices
a
little
bit
in
this
might
include.
So
if
anybody
has
some
thoughts
on
it,
if
not,
we
don't
need
to
make
a
decision
right
now,
but
I
wanted
to
put
it
in
people's
brains.
I.
E
A
F
Not
uncommon
for
it
to
happen,
I've
seen
it
happen,
probably
three
or
four
times
now
sullen
did
it
I
did
it
once
where
we're
just
typing
along
and
all
of
a
sudden
half
a
heck
bed
goes
missing
and
I'm
also
typing
in
the
headline,
or
something
and
I
try
to
you
know
command
C
in
case
it
was
me
and
there's
no
undo
no
undo.
Oh,
that's
terrible!
Oh
well,.
A
D
E
A
A
C
A
A
B
A
I
sort
of
give
this
is
the
feeling
Bedford
for
actual
note,
taking
something
like
this
when
you're
kind
of
it
so
I
think
for
discussion.
The
idea
of
using
the
I'm
willing
to
go
over
to
the
issue
tracker
as
a
conversational
thing,
where
you
really
are
we
flying
one
after
the
other,
but
I
think
when,
when
a
bunch
of
people
are
collectively
like
adding
to
a
document
at
the
same
time,
it
really
is
useful.
A
A
A
H
Know
thanks
for
those
that
attended
the
tutorial
last
week,
and
thanks
for
now
no
for
sort
of
holding
us
the
fire
on
that.
One
I
think
that
went
well
and
I
appreciate
any
feedback
from
them.
Just
working
with
Darion
this
week
on
refactoring
in
test
suites
for
the
various
notebook
components
in
Jupiter
Lloyd
great.
A
No
I
wanna
I
first
I
want
to
thank
you
all,
because
you
guys
actually
took
a
lot
of
time
to
put
that
together
and
you
put
a
lot
of
thought
into
it
and
I
think
it
was
very
useful
and
second,
I
want
to
encourage
people
to
to
go
to
go
over
it
on,
because
it
gives
a
good
overview
of
how
things
are
put
together
and
I
definitely
actually
helped
me
quite
a
bit,
and
so
you
guys,
you
guys,
put
a
lot
of
thought
into
it.
A
I
hope
I
hope
even
the
exercise
of
going
through
that
was
useful
to
hell.
Do
I
salute
organize
the
mental
model
in
your
head
and,
I
think
will
be
very
productive,
but
but
I
encourage
others
to
to
pitch
in
watch
it
and
then
provide
feedback
on
the
on
the
issue
tracker
on
how
the
documentation
ought
to
be
structured
and
what
wasn't
clear
and
what
either
what
wasn't
explain
it
or
if,
if
the
conversation
brings
up
technical
questions
and
ideas,
obviously
bring
them
up.
I
So
one
minor
thing
is
I'd,
like
a
graphic
for
the
council
to
put
in
Jupiter
lab
the
way
we
have
four
new
notebook
and
new
terminal,
so
that
I
could
put
that
on
the
front
page
I
wasn't
exactly
sure
who
to
ask
about
that.
So
I
figured
I'd,
bring
it
up
at
the
meeting.
Yeah.
A
Carol,
do
you
think
on
some
of
your
some
of
the
folks
over
there
who
are
more
focused
on
design,
might
might
take
up
that
idea
that
it
would
be
something
that
would
represent
and
the
columns
what
we're
calling
the
console
should
have
an
icon.
That
is
a
little
bit
different
from
the
terminal,
because
one
of
them
is
really
an
actual
terminal
emulator,
where
the
console
is
meant
to
I.
Don't
know
it's.
It's
like
the
old
cutie
console
right.
A
A
What
yeah
that
could
be
that
that
could
be
a
good
starting
point,
but
it
might
be
a
nice
time
to
refresh
it
in
and
bring
it
bring
it
kind
of
give
it
a
given
a
design
refresh.
What
do
you
think
Carol?
Do
you
think
there's
some
bandwidth
over
there
from
the
design
folks,
I
think.
I
Great,
maybe
make
it
a
little
more
open
to
other
people's
thoughts.
I
could
post
it
as
an
issue
on
Jupiter
lab
and
then
email
with
a
little
issue.
Yeah
perfect.
A
Posted
as
an
issue
and
just
ping
ping,
ping,
Brian
and
Carol
directly
I.
Would
things
like
this
that
may
be
of
interest
to
a
broad
audience
had
never
heard,
send
an
email
to
the
list?
Hey
folks,
anyone
who's
interested
here?
If
you
want
to
we're
beginning
to
get
the
foundations
for
the
sort
of
the
evolution
of
the
Qt
console
attention
over
here
on
this
issue
and
then
whoever
whoever
cares,
but
you
can
definitely
think
directly-
are
Carol
and
Brian.
A
E
Right
side,
it's
more
writing
documentation
on
how
to
migrate,
because
we
had
so
many
change.
That
thing
is
that
if
you
write
it
as
we
go,
we
have
to
anyway
DJ
okay,
and
there
is
a
feature
you
will
not
go
even
aware,
existed
in
I
bite
on
that.
Finally,
we're
not
working
for
a
few
years
and
so
that
we
started
working
again
and
have
better
yeah.
So.
A
A
little
bit
of
it
is
actual
release:
time,
documentation,
okay,
okay,
well,
a
big
release,
there's
always
a
little
bit
of
that,
but
that
has
to
happen.
But
but
one
thing
that
I
want
to
make
sure
is
that
we
we
really
move
towards
a
practice
of
documenting
code
as
we
go
when
it
makes
sense
and
I
see
your
point.
Sometimes
there
are
things
that
really
need
to
happen
at
least
Im,
because
you're
kind
of
collecting
the
aggregate
that.
E
E
F
Think
nothing,
nothing
beyond
what
Steve
was
other
than
just
mentioning
that
were
to
have
a
huge
PR
into
widgets,
to
transfer
it
over
to
typescript
and
building
on
the
framework
that
the
Jupiter
lab
and
the
phosphor.
Our
projects
are
using
for
testing
and
things
like
that,
just
FYI,
so
that's
6,
27,
yes,
o'sullivan
and
I-
are
in
someone's
reviewing
it
and
and
I'm
finishing
up
cool.
J
F
So
so,
I'm
not
sure
exactly
what
your
question
is,
but
what
I'm
doing
is
copying
over
the
testing
infrastructure,
so
the
directory
structure,
dependencies,
etc
and
then
I'll
be
moving
that
the
existing
testers,
not
very
many
tests
existing
right
now,
but
moving
the
existing
app
I
widgets
test
into
the
new
directory
structure,
and
things
like
that,
so
that
there's
consistency
between
Jupiter
lab
and
iPad
widgets
for
how
testing
is
done
and
the
directory
structure
when
dependences
we
have
etc,
design.
Ask
a
question
normative.
F
Yeah,
that's
a
really
good
idea:
I,
don't
know
if
the
cookie
cutter
has
testing
built
in,
but
if
it
doesn't
belkin
doesn't
okay,
oh
yeah,
we're
happy
to
put
it
in
the
cookie
cutter.
I
mean
we're
hammering
out
I,
pretty
heavily
with
phosphor
and
Jupiter
lap
and
then
hopefully
spend
with
I
high
widgets
to
sort
of
prove
out
the
infrastructure,
and
then
we
could
put
it
in
the
cookie
cutter.
That'd
be
a
good
idea.
The.
J
A
Okay,
anyone
else
anything
that
might
have
come
up
come
up
doing
conversation
any
other
thoughts.
I
wanted
to
congratulate
Ginny.
She
gave
her
first
project
Jupiter
talk
at
the
NSF
Western
Western,
big
data
hub
uh-huh,
when's,
Big,
Data,
innovation,
west
we're
gonna
I
can
never
remember
the
name
about
yeah.
K
K
The
the
workshop
about
a
third
of
the
live
audience
had
heard
of
or
had
is
Jupiter
and
then
a
handful
of
people
will
approach
me
after
the
talk
and
set
what
I've
been
using.
Jupiter
ipython
notebooks
for
a
long
time,
but
I
didn't
know
that
it
did
all
this
other
stuff,
so
I
presented
on
MV
viewer
and
some
of
the
other
projects
that
run
under
the
veggie
better
umbrella
and
yeah
people
were
excited
about
the
capabilities
and
then
the
development
is
happening
with
sugar,
lab
and
dashboards
for
sure
people
love
their
dashboards.
Awesome
awesome.
A
K
D
K
K
A
K
A
End
of
no
end
of
energy
and
talent,
the
executive
director
is
here
at
Berkeley
and
she's,
very
good.
All
right
folks!
Well
we'll
see
you
all
on
on
friday
for
the
Jupiter
lab
meeting
for
those
who
can
join
and
then
next
Tuesday
again,
unless
anybody
has
any
last
last
comment:
yeah
thanks!
Everybody
happy
summer.