►
From YouTube: Jupyter/IPython Dev Meeting, January 10, 2017
Description
Meeting of the Jupyter/IPython development team, January 10, 2017
Meeting Notes: https://etherpad.wikimedia.org/p/Jupyter-Weekly-Team-Meetings-2017(Jan)
A
Hi
good
morning,
Jupiter
and
good
morning,
Internet,
it's
a
Tuesday
January
tenth.
This
is
the
weekly
Jupiter
def
meeting.
Today
we
have
a
number
of
things
going
on.
We
have
the
data
carpentry,
reproducible
research,
hackathon
happening
here
in
bids
right
behind
us
here,
so
I'll
be
stepping
out
shortly
for
that
on
the
project
management
side,
we
are
starting
to
experiment
with
new
venues
for
taking
our
weekly
notes.
We've
been
using
hack
pad
for
quite
a
while,
but
we've
been
experiencing
some
data
loss
when
the
hawk
pads
trying
to
sync
all
the
different
users.
A
So
we're
we're
trying
out
etherpad
this
this
week.
We
might
try
it
out
for
the
whole
month
just
to
see
if
we
can
get
a
run
going.
This
instance
is
hosted
by
wikimedia
or
Wikipedia.
So
thanks
to
them
for
letting
us
play
around
in
their
sandbox
and
I'll,
be
posting
the
link
both
in
the
existing
hack
pad
Jupiter
hack
pad
and
in
down
in
the
notes
of
this
video
on
youtube.
So
I
need
to
get
there
myself.
A
So
that's
those
are
the
two
main
things
and
the
project
management
area,
the
data
carpentry
hackathon
and
moving
over
to
etherpad.
It
looks
like
the
for
governance.
The
code
conduct
PR
was
merged
and
there's
a
link
there
in
the
in
the
etherpad.
So
you
can
check
that
out
when
you
get
a
chance
and
then
it
looks
like
the
next
folks.
Folks
on
the
list
in
the
etherpad
are
Darien
and
Ian,
with
the
Jupiter
lab.
B
C
Right,
yeah,
a
lot
of
the
stuff
that
was
a
almost
in
a
state
of
readiness
before
the
end
of
the
year,
we've
kind
of
put
together
in
the
last
few
days.
So
it
was
this
a
big
abstract,
editor,
revamping
that
Steve
was
working
on
that.
Just
as
of
now
the
last
a
bit
of
it
went
in
where
we
could
get.
D
C
Of
we
get
rid
of
a
lot
of
things
that
were
called
more
specific
and
in
fact
we
don't
really
need
a
cell
editor
widget
it
just
that
comes
for
free
based
on
what
an
editor
all
the
editors
need
to
do,
and
we
have
now
got
almost
all
of
session
restoration,
except
for
the
layout
of
the
main
area.
So
that
means
is
that
the
sidebar
restoration
not
works
as
well.
C
It's
a
little
bit
easier
to
use
than
it
was
before,
and
there's
a
class
called
instance
restore
if
you're,
if
you're
a
plug-in
author-
and
you
don't
want
to
just
copy
and
paste
some
other
plug-in-
which
you
could
totally
do
there
there
there
there
all
the
store
in
your
state
go
to
that
class,
and
it
has
big
docstring,
showing
you
exactly
what
the
lifecycle
it
is,
but
the
for
end-users.
I
would.
C
I
would
ask
that
you
just
pay
attention
to
what's
happening
and
see
if
it's
actually
restoring
in
a
usable
way,
see
if
there's
any
annoying
things
going
on,
because
this
is
one
of
people
will
definitely
know
this
as
we
transition
to
pay
this,
we
want
it.
We
want
it
to
be,
not
annoying,
don't
see
what
else
isn't
I
have.
B
So
this
is
I.
Ok,
ok,
duck
cool
on!
This
is
from
doing
a
demo
off
of
the
off
of
the
demo,
repo
that
that
Jason,
hadn't
I
I
did
this
demo
yesterday
and
I
was
using
old
versions
of
lab
that
don't
have
the
restored,
but
I
was
wondering
if
the
restore
will
interfere
with
this
pattern,
which
is
a
pattern
that
I
know,
jason
has
also
used,
which
is
to
open,
simultaneously
open
multiple
views
of
the
same
super
lab
effectively
the
same
URL
on
different
desktops,
but
we're
in
one.
B
You
may
have
loaded
a
couple
of
notebooks
for
this
part
of
the
demo
and
another
he
may
have
loaded.
Instead,
a
couple
of
markdown
files
and
some
consoles
and
another
one.
You
may
have
loaded
some
of
the
geo
jason
examples,
and
but
you,
since
at
the
time
there
was
no
concept
of
concession
state,
you
could
just
reloj
kind
of
go
and
load
each
of
them
and
there
would
be
no
collisions
whatsoever
because
there
was
nothing
to
collide
on.
B
There
was
no
state
management
at
all
and
so
I'm
wondering
if
what
if
Jason
was
doing
and
I
was
doing
with
these
demos
of
effectively
hitting
the
same
thing
in
parallel,
multiple
times
is
going
to
become
kind
of
a
pattern
that
won't
work
and
that
may
be
fine
but
then
kind
of
what
should
we?
What
should
we
do?
I
guess
and
pull
in,
and
we
don't
have
to
solve
the
question
right
now,
but
it
occurred
to
me
yesterday
because
I
remember
Jason
demoed
at
that
way
and
I'm
using
it
in
a
similar
manner.
B
C
C
So
if
you
have
multiple
windows
in
the
same
browser,
then
the
next
time
you
refresh
every
single
document
you
opened
in
all
your
sessions
combined
will
reopen,
but
there
are
things
that
are
unique,
so
current
widget
and
the
collapsed
state
of
the
right
sidebar
and
the
collapsed
day
of
the
left
sidebar,
those
are
actually
unique.
So
they
will
be
the
last
one
that
updated
and
it's
written
in
a
fairly
a
user-friendly
way
in
that,
if
it
can't
run
the
command,
that
is
meant
to
restore
something.
C
It
just
deletes
that
key
from
the
database
and
says
all
right,
something's
wrong,
I'm
not
going
to
try
to
restore
so
chances.
Are
you
shouldn't
see
any
errors,
but
it
might
not
restore
in
the
way
you
would
have
thought
when
you
refresh
the
specific
tab
that
you
refreshed
now,
if
you
want
to
work
around
it,
you
want
multiple
views.
You
could
always
use
multiple
browsers
or
you
can
go
to
the
help,
menu
and
use
the
clear
application.
C
State
anytime
you
like,
or
if
you're
feeling,
extra
hacky,
you
could
open
up
the
console
and
do
window
that
local
stores
not
clear
arm,
especially
if
you
find
yourself
in
a
weird,
broken
state
that
I
haven't
managed
to
predict,
but
going
forward.
I,
don't
actually
know
what
the
best
thing
to
do
is
we're
building
an
interface
that
has
multiple
views
on
multiple
documents.
C
It
might
be
reasonable
for
us
to
say
the
the
canonical
way
of
using
this
is
in
a
single
window,
except
that
we're
planning
on
supporting
tearing
stuff
out
into
multiple
windows
and
that
sort
of
thing.
So
it's
not
it's
not
clear
I
think
we
have
to
define
these
behaviors,
probably
as
a
result
of
some
research.
I,
don't
I,
don't
know
if
we'll
just
by
thinking
real
hard
come
up
with
a
great
thing
will
probably
have
to
see
what
people
need
and
want,
and
also
what
other
apps
do
and.
D
F
You
could
also
think
about
being
able
to
launch
Jupiter
lab
with
a
particular
session
name
or
something
like
that,
and
we
incorporate
that
into
the
big
challenge
like
x,
star
trooper
lab
under
like
this
end,
like
a
I,
don't
know
some
kind
of
you
know
virtual
him
for
Jupiter,
lab
storage
or
something
like
that.
Then
we
could
do
it
that
way.
Potentially
so
there
I
mean
there
are
multiple
ways
to
go
about
that.
He.
B
Said
yeah
no
III
wasn't
asking
because
kind
of
form
I
was
neither
complaining
or
asking
for
an
immediate
solution,
but
kind
of
flagging
as
I
use
it
in
some
of
these
scenarios,
I
did
I
think
it
would
be
good
to
put
that
on
your
way
out.
But
I
know
it's
a
comfort
one.
So
I
wasn't
expecting
a
solution
right
away.
One.
G
Of
the
things
that
our
near
future
roadmap
is
a
concept
of
settings
and
storing
settings
on
the
server
for,
along
with
that
comes
and
concept
of
server
sessions
and
a
better
communication
with
the
server.
And
so
maybe
your
URL
is
a
random
URL
mr.
each
other
session
and
then
or
you
said
it
could
be
or
something
that's
through
server
session
and
you
know
maybe
get
request
to
the
same
URL.
The
session
be
stored.
C
Yeah,
you
could
imagine
something
like
the
running
tab,
that's
a
sessions
tab
and
you
can
like
actually
clone
a
session
that
you
used
to
be
in
before,
like
there
there's
lots
of
interesting
possibilities
once
its
server
side.
That
would
actually
transcend
the
limitation
that
right
now
we
only
restore
state.
You
know
if
you're
in
the
same
browser-
and
you
haven't-
cleared
your
browser
setttings
and
all
that
so
yeah
there's
definitely
some
flexibility
once
we
move
to
the
server,
but
there's
also
latency
right
now,
it's
pretty
much
guaranteed
to
save
everything.
C
But
if
it's
an
API
call,
you
know
you
can
imagine
something
sometimes
don't
get
saved,
but
anyway,
yeah
I
mean
it's
a
it's
a
it's
a
big
topic,
but
at
the
very
least,
I
think
what
we've
removed
now
is
the
annoyance
of
having
to
reopen
all
the
things
you
had
open.
If
you
want
to
continue
and
you've
just
refreshed
the
page,
which
I
think
is
like
a
bare
minimum
for
releasing
to
end
users
who
aren't
just
working
on
the
project
itself,
I.
B
H
Yeah
I
was
just
gonna,
paying
everybody
that
we're
getting
closer
to
a
real
time.
Collaboration
is
being
active
in
Jupiter
lab,
so
if
people
want
to
take
a
look
at
that,
there's
a
couple
links
that
are
pasted
in
the
pad
I've
demoed
it
a
little
bit
too
to
a
few
people,
but
without
a
really
good
user
experience.
H
So,
for
the
past
couple
days,
I've
been
porting
and
updating
some
of
the
work
that
Matthias
did
with
Jupiter
drive
to
try
to
get
a
more
integrated
experience
with
a
Google
Drive,
specifically,
so
part
of
that
has
been
reimplement
a
file
browser
in
Jupiter
lab.
So
this
is,
to
my
knowledge,
I
think
the
first
alternative
file
browser
in
lab.
So
that's
been,
there's
been
a
bit
of
kicking
the
tires
of
the
file
browser
model
to
see
how
well
that
works
and
I'd
say
it's
going
reasonably
well,
but
that's
something
to
keep
an
eye
on.
B
I
D
G
Or
I
can
Stratus
at
minus-4
Sullivan,
who
makes
the
beta
6
of
the
widgets
lots
of
Polish
cleanup,
continuing
ongoing
polish
and
clean
up
doubling,
got
custom,
widgets
working
and
then
be
viewer,
I'm
sure,
he'll,
post
the
URL.
Some
of
the
examples
he's
been
showing
after
that
first
bullet
point
and
exported
output
widget
to
notebook
for
right.
Now
it
was
only
number
five
we
made.
G
We
may
release
a
notebook
4.4
if
I
was
delayed
after
/
widget
six
comes
before
don't
put
five,
then
the
big
release
Norfolk
4.4
with
this,
so
the
f
entities
consider
could
never
live
before
and
we
talked
about
documenting
the
if
it's
a
lot
better
and
how
to
make
it
work
better.
With
open
collaboration.
B
What
are
I
guess:
I'm
I'm
wondering
what
are
your
concerns
with
the
kind
of
4x
5x
lab
api's
with
the
widgets.
It
seems
like
right
now.
We
do
have
kind
of
three
parallel
lines
with
these
api's
and
extension,
installation
and
whatnot
that
the
widgets
have
an
impact
on
and
I'm
just
curious
kind
of
what
your
your
take
and
understanding.
Is
it
getting
too
complicated
as
a
fine
as
it
I'm
to
pass
them?
Are.
B
No,
no
specifically
about
specifically
about
widgets,
specifically
about
widgets
working
across
across
kind
of
notebook
versions
for
users
for
users
who
are
not
completely
up
to
speed
on
the
details
between
the
low
level
differences
in
extension,
management
and
JavaScript
integration
on
the
4x
line
on
the
5x
line
on
supra
lab,
basically
what
they
want.
What
users
think,
if
I
have
a
notebook
and
I'm
using
widgets,
does
it
work
right
so.
D
It's
pretty
easy,
it's
pretty
easy.
If
you're
using
stable
versions,
Devon
store
is
complicated
because
you've
got
to
do
develop
instead
of
hype.
I
would
subject
edges
with
jess
and
everything.
So
there
are
multiple
moving
pieces
if
you
want
to
have
the
dev
install
of
everything
at
the
same
time,
I
I
think
that,
as
if
we'd
before
we
have
the
proper
Jupiter
server,
that
with
comedy
approach
for
extensions,
we
are
still
have
a
2
comments,
installation
for
widgets.
So
that's
the
main
issue,
I
think
for
both
lab
and
ended
book
and.
B
D
That's
right,
but
right
now
we're
concentrating
on
I
by
which
at
six
and
which
is
the
first
version
that
we
all
have,
or
these
different
contexts
working
like
the
web
contacts,
Jupiter,
lad,
notebook
and
strings
and
Andy
viewer.
But,
to
be
honest,
other
thing
it
complicates
the
picture
actually
makes
things
simpler,
though
yeah
I.
G
Think
the
thing
that
we're
concerned
mainly
about
right
now
is
making
sure
that
I
five
digits
five
when
I
hi
widget
six
are
compatible
because
there
was
pretty
major
rewrites
between
the
two
etc
so
I
by
widgets.
That
should
work
across
at
least
notebook
five
in
Jupiter
lab
and
if
we
put
this
one
back
border
PR
in
notebook
for
as
well-
and
we
just
want
to
make
sure
that
people's
code-
that's
in
that's
using
iPad
widgets
works
the
same
across
I
hi,
widget
6
upgrade
yes,.
B
Just
yesterday
we
have
questions
here
at
the
refusal
research
hackathon
about
the
use
of
widgets
and
they're
building
curriculum
for
teaching,
and
they
were
very
interested
in
that
and
so
I
be.
Our
answer
was
as
long
as
you're
using
released
versions,
it's
fine
and
there.
No.
There
is
no
breakage
kind
of
across
across
release
versions
for
end-users,
and
we
want
to
continue
be
able
to
tell
users
who
are
in
education
and
research,
kind
of
that
that
that's
that's
a
true
story.
B
G
There
are
some
breakages,
it
is
major
release,
so,
for
example,
the
styling
is
done
a
little
bit
differently.
All
we
have
a
new
layout,
huge,
etc,
but
we're
trying
to
minimize
those
and
documents
and
really
love
and
making
sure
the
little
relieved
a
common
use
cases.
Four
digits
the
derp
backwards,
compatible
more
advanced
use
cases.
There
are
breakages,
it
is
a
major.
It
is
what.
D
Thinking
about
is
make
the
current
Veda
a
public
beta
and
send
an
announcement
that
specifically
for
for
the
people
who
are
authoring,
custom
widgets,
because
there
are
folks
out
there
who
are
building
really
sophisticated
things
on
the
top
of
widgets.
And
we
don't
want
to
break
their
libraries.
And,
and
so
we
want
to
be
able
to
communicate
clearly
with
them
about
what
they
need
to
do
to
migrate
to
the
latest.
B
B
D
B
That
is
still
early.
The
the
danger
is
putting
stuff
and
not
telling
people,
but
what
its
limitations
are,
but
as
long
as
we
get
people
fair
warnings
and
then
I
think
that's
fine
and
so
I
would
say,
yeah
go
for
it
make
the
announcements,
the
mail,
our
our
main
meal,
which
doesn't
really
see
that
much
traffic
and
especially
about
releases
and
all
every
time
we
put
something
out.
There
should
be
a
quick
note
on
amendments
hey.
B
D
B
K
So
this
is
something
I've
been
working
on
for
the
last
couple
months,
and
so
the
mime
render
cookie
cutter
is
in
a
pretty
stable
state.
So
I
encourage
people
to
start
using
it.
I've
used
it
to
make
the
Jupiter
lab
JSON
extension,
which
renders
JSON
as
a
tree
as
interactive
tree
I
just
recently
updated
the
Geo
JSON,
which
I
think
Jason
started
arms.
K
Now
it
it
was
rewritten
using
the
cookie
cutters,
so
it
supports
both
lab
and
notebook,
and
then
I'm
working
on
a
json
schema
table
and
that
it's
currently
working
on
and
now
I'm
working
on
the
panda
side
and
trying
to
figure
out
a
way
to
make
it
work
with
pandas
really
seamlessly
and.
B
K
That's
the
idea:
there's
Kyle
opened
up
an
issue
on
pandas
on
a
and
it's
dev
and
it
there's
a
discussion
going
on
so
I'm,
not
sure
exactly
what
the
result
is
going
to
be,
whether
it's
going
to
be
like
a
referee
method
or
whether
it's
going
to
be
some
new
like
like
the
way
I'm
currently
doing
it
as
I'm.
Just
taking
I'm
taking
a
penas
data
frame
and
doing
a
to
dick.
You
know
that
that's
just
now,
I'm
currently
doing
it,
but
maybe
there
will
be
something
added
to
pandas.
K
B
Yeah
that'll
be
useful,
I
mean
obviously
wes
wes
is
now
at
2
Sigma
2
Sigma
has
their
data
bus
for
a
Jason
serialization
of
common
data
structures
across
languages
in
their
notebook.
So
I'm
gonna
I'm
going
to
guess
that
whatever
they've
done
they've
done
for
that
for
the
the
beaker
notebook
data
bus,
whence
it's
probably
already
on
top
of
that
and
that
in
that
discussion,
ok,
what's
was
his
last
name.
B
West
McKinney
is
the
he's
the
original
author
of
pandas
and
he's
now
at
two
sigma,
which
is
the
company
that
developed
the
beaker
notebook.
We
has
this
concept
of
this
sort
of
cross
cross
language,
common
data
bus,
where
they
serialize
various
data
structures
and
that
thing
that
common
data
bus
is
in
Jason,
so
all
languages,
basically
dumped
dumped
their
common
variables
in
Jason
into
this
common
date
of
us
and
so
I
would
imagine.
I
would
imagine
that
they
they've
already
done
a
fair
amount
of
work
in
this
direction.
Okay,
cool
I'll,
look
into
that
grant.
G
K
Cookie
cutter,
yes,
I'm,
attended
you
burlap
org
yeah,
so
that's
pretty
these
pretty
self-explanatory.
But
if
you
have
any
questions,
just
you
know
feel
free
to
contact
me
and
yeah.
Then
I've
done
a
couple.
Little
experiments
with
these
mime
renderers
nothing,
nothing
really
worth
showing
yet
but
yeah.
What
little
there
is
something
we're
sharing,
sharing
I
will
do
so
and
I
think
that's
it
for
the
memberid
or
extensions
grant.
I
Hard
on
everybody,
most
of
the
work
that
got
done
this
past
week
for
Jupiter
hub
has
to
do
it,
deploying
for
education
and
getting
Brian
ready
for
another
class
at
Cal
Poly
and
one
of
the
things
that
we
did
was
release
0.72,
which
has
some
cool
support
for
running
a
shared
single
user
server
as
a
service
as
well
as
authenticating
small
groups.
So
there
are
examples
in
the
repo
of
how
to
do
that,
and
we've
also
updated
a
lot
of
the
ansible
scripts
for
the
teaching
reference
deployment
and
working
on
the
online
tutorial.
I
B
I
actually
played
with
envy
over
the
last
few
days,
because
I
was
giving
a
talk
yesterday,
I
wanted
to
at
least
have
a
little
highlight
of
it,
and
it
looks
really
great.
Do
you?
I
B
L
B
I
It's
a
pretty
straightforward,
like
we
cleaned
up
a
lot
of
the
usability
of
it.
So
it's
pretty
straightforward,
like
most
of
the
configuration,
has
now
been
in
one
vial
and
I'm
working
on
trying
to
get
a
vagrant
instance
so
that
you
can
just
type
biggrin
up,
and
you
know
once
you've
configured
it
so,
but
yeah
I
mean
it's
simpler
than
the
docker
at
this
point,
I
think
so.
B
There
are
a
lot
of
context
where
you
have
people
who
are
saying
I
wanted
to
play
for
my
class.
For
my
group,
I'm,
not
a
sysadmin
I,
don't
have
a
sysadmin,
but
if
you
give
me
some
kind
of
easy-to-follow
steps,
I
can
manage
that
right
and
that
that
use
that
kind
of
sweet
spot
of
small
to
medium
sized
classes
without
professional
dev
ops
support
is
actually
pretty
that
market
in
a
sense
is
very
large
in
education
and
so
making
making
people
aware
that
on
the
education
mailing
list
might
be
might
be
useful.
M
Ditto
lots
of
some
small
things
and
then
we
convert
testing
improvements.
One
thing
that
people
should
probably
know
is:
if
you
read
the
docs,
if
you
have
Python
greater
than
or
equal
to
3.5
in
your
environment,
yeah,
Mel
and
you're,
using
a
condo
environment,
it
right
now
is
going
to
start
requiring
3.6.
I'm
not
sure
why
it
does
this,
but
that
was
what
was
breaking
our
read
the
docs
most
recently.
So
just.
B
You
know
Mike
I
noticed
that
I
think
it's
a
problem
with
Conda
is
my
impression
or
something
in
the
pipeline
of
repose
that
come
from
continue
that
when
you
cry,
if
you
install
an
akan
mini
Conda
like
I,
have
to
do
it
this
week
and
I
might
contain
solution,
broken
so
I
wiped
it
and
made
a
new
one.
There's
a
very
weird
effect
right
now.
B
If
you
install
anaconda
and
then
you
make
a
new
environment
and
you
say,
conduct
create
dash,
enthu,
Python
or
ipython,
or
anything
that
new
end
becomes
a
36
environment,
but
the
36
repo,
the
36
channel,
doesn't
have
the
complete
stats
set
of
packages
it's
very
incomplete.
So
you
know
what
you
have
to
do
is
you
have
to
manually
say
in
the
create
call,
you
have
to
manually
say,
bite
on
equals,
3.5
to
get
force
on
that
to
stay
and
I.
Think
this
is
a
car
I
think
this?
M
That
sounds
exactly
like
what
was
happening.
I
already
have
an
issue
open
with
them.
I
already
fixed
up
for
the
for
an
env
convert.
I
was
just
warning
other
people
if
they
had
this
and
they
hadn't
been
watching
that
read
the
docs
and
suddenly
they
see
that
they
were
failing.
This
is
probably
going
to
be
the
cause
I.
B
M
I'll,
do
that
and
then
there's
a
bit
of
a
discussion
going
on
about
whether
or
not
we
should
be
replacing
some
of
the
so
basically
matplotlib
back
end,
some
of
them
only
work
in
the
notebook
or
that
kind
of
an
environment
in
line
and
notebook
in
particular,
and
so,
if
you
try
to
currently
convert
any
notebook
with
that
in
it
to
a
script,
it
will
fail,
and
this
is
suboptimal.
M
This
is
even
if
you
use
ipython
and
so
I've
been
working
on
getting
a
fix
where
you
just
pop
in
there
rip
out
the
back
end
so
that
you
fall
back
on
the
default.
If
you
have
only
those
two
back
ends,
but
then
I
min
Thomas
and
Matthias
had
pointed
out
that
the
goal
had
been
to
move
away
from
Python
and
I
Python
specific,
like
code
and
implementation
details.
On
the
other
hand,
this
is
a
common
enough
use
case
and
also
we
already
have
an
ID
on
diff
Python
function,
which
is
specific
to
ipython.
B
M
I
also
had
sort
of
a
meta
question
about
the
organization
of
the
notes.
I
have
a
little
I
have
a
little
project
that
I'm
working
on
on
the
side.
I
don't
want
to
put
it
in
there
yet,
but
there's
no
current
way
to
show
like
to
describe
and
talk
about
how,
like
things
that
are
outside
of
the
core
packages
right
now,
but
they
would
still
be
relevant
to
other
people.
So
where
should
we
do
that?
Should
we
just
have
a
miscellaneous
thing
at
the
end
of
the
notes.
B
Yeah
I
would
encourage
you
to
put
yet
to
make.
Maybe
we
can
have
a
miscellaneous
section
or-
or
you
just
put
a
little
section,
basically
with
just
with
your
name
on
it.
If
you
want
to
like
report
report
from
Mike
on
other
things
and
any
anyone
should
feel
comfortable,
but
but
it
would
be
good
if
it's
captured
in
the
notes
as
well.
I
think
I
really.
I
really
want
to
support
kind
of
much.
B
M
A
medium
burning
bug
with
Matthew
Brett
that
I
would
like
to
fix
because
it's
a
night
it's
we
killed
the
ability
to
sideload
exporters
into
the
slide
load
templates
in
memory.
So
that
is
one
of
the
big
fixes
that
would
be
coming
with
this,
but
I
think
that
we
can
probably
figure
out
this
thing,
because
the
actual
implementation
of
it
it's
just
a
matter
of-
is
fairly
straightforward.
It's
just
a
matter
of
deciding
what
to
do.
Okay,.
A
You're
here
he's
here,
thanks
for
joining
so
looks
like
he's,
not
sure
how
the
existing
work
with
the
paper
fits
with
the
new
procedure
for
writing
mini
papers.
Do
we
still
want
to
do
an
overview
paper
or
discard
it
and
write
many
jobs,
pay
pers
for
individual
repos?
So
that's
the
question
that
I.
B
My
personal
opinion
is
that
now
now
that
we're
kind
of
seven
on
this
we
should
we
should
still
do
it
over
a
larger
overview
paper.
I
think
the
josh
papers
and
meet
a
different
goal.
A
joss
paper
is
meant
to
be
very
short.
It's
it's
basically
sort
of
almost
an
enhanced
it's
kind
of
a
DOI
with
an
abstract,
almost
right.
B
It's
really
more
of
a
a
checklist
of
the
state
of
the
payoff
of
the
software
release,
along
with
a
short
summary,
so
that
kind
of
a
visible
identifier,
a
joss
paper
is
not
a
place
to
discuss
the
ideas
about
the
project
and
I
think
it's.
It
is
time
that
we
do
that,
and
that's
the
kind
of
thing
that
that
you
do
in
an
overview
paper
where
you
kind
of
summarize
the
thinking
about
what
what
is
it
that
we're
doing?
What
are
the
key
ideas
and
in
that?
B
That
is
a
separate
kind
of
writing,
so
I
do
think
we
which
is
kind
of
what
what
the
same
pie
paper
did,
that
we
copied
the
procedure
from
which
was
really
kind
of
describing
simply
as
a
project
where
the
ideas
off
of
sin
PI
as
a
library
rather
than
simply
flagging,
simply
zero
point
whatever,
which
is
more,
what
the
josh
releases
are
so
I
yeah
I
think
they
do
need
different
needs,
but
we
can
continue
discussing
this
I
mean
obviously
that's
I'm,
not
making
a
decision.
I
am
stating
my
opinion
here.
B
A
I've
been
experimenting
with
the
github
projects,
features
and
the
best
way
to
see
what
I've
been
working
on
is
to
just
go
to
our
main
org
github
com
forward,
slash
Jupiter,
and
hopefully
all
of
you
can
see
the
projects
when
I
tested
this
with
another
person
who
was
not
part
of
the
project,
he
couldn't
see
our
projects
tab
at
all,
so
I
think
it's
oh.
A
B
A
A
No
worse,
all
right,
so
at
the
org
level
we
should
be
looking
at
the
project:
Jupiter
org,
with
our
logo
logo
at
the
top
interactive
computing
or
destination,
or
location
as
the
future
under
projects
that
you
should
see
three,
and
if
you
click
on
that
I've
created
three
projects.
The
first
is
a
dummy
project
text
test
one.
A
Can
everybody
see
that
or
if
you
can't,
can
you
just
say,
no
use
your
voice
and
you
know
I'm
going
to
assume:
that's
all
yeses,
so
the
ones
that
are
actually
active
right
now
that
I'm
working
on
is
a
2017
Jupiter
team
conference
and
event
calendar.
So
this
is
taking
the
data
from
the
google
calendar
that
Ana's
put
together
and
created
a
calendar
with
in
github.
So
if
you
click
on
it,
you
should
see
all
the
months
of
the
year
for
2017
and
then
the
different
events
that
that
I
found
on
on
my
calendar.
A
A
Thanks
so
that
may
be,
because
all
these
cards
live
in
a
new
repo
that
I
created,
which
is
a
project
dash
MGT,
so
project
management.
A
G
A
Alright,
so
cool,
so
if
you
go
into
so
that
was
the
the
high
level
org,
the
org
level
projects
and
both
of
the
projects
that
were
listed,
our
calendars,
one,
is
like
a
project
calendar
that
includes
like
milestones,
we're
working
on
some
more
the
dev
work
and
the
other
one
is
like
event,
conferences
and
events.
So
I
thought
it
may
be
useful
to
have
that
events
calendar
in
github.
So
you
know
there
are
fewer
places
where
we
leave
the
github
space
to
find
information
about
the
project.
So
we
can
try
that
out.
A
If
it,
if
you
guys
like
it,
we
can.
You
know,
think
about
moving
into
that.
If
not,
we
can
keep
the
google
calendar.
That's
fine!
So
let
me
ask
you
all
to
just
go
into
that
project:
ash
MGT
repo,
the
readme
is
just
you
know.
The
definition
of
a
project
from
the
project
management
institute,
which
is
you
know,
project,
is
something
with
a
beginning
and
end.
So
the
things
that
I'm
going
to
list
in
this
repo
will
be
things
that
will
start
and
end.
So
we
don't
want.
You
know
Jupiter
itself.
It's.
A
A
I've
created,
11
kind
of
sample
projects
of
things
that
have
been
on
my
plate
in
some
way
shape
or
form
and
I've
created
a
couple
different
types
of
workflows,
so
some
of
them
will
have
the
calendar
where
their
actual
issues
that
are
that
are
fit
into
a
month
of
the
year.
So
I
think
the
Jupiter
lab
released
Jupiter
lab
1
point
0
on
my
screen
is
3
up
from
the
bottom
and
there
are
a
couple
issues
in
there
for
milestones
that
we
we've
identified
that
we
want
to
hit
over
the
next
six
to
eight
months.
A
A
Yep
so
so,
like
I'm,
looking
a
tad
contributor
map
to
website
that's
a
very
small
project
that
we've
been
working
on
for
a
while.
So
for
these
I
just
created
three
different
work
states
to
do
in
progress
and
done
and
yeah.
So
the
way
this
works
is
I
create
an
issue
in
the
project
management.
Repo
is
you
know
something
high
level
piece
of
work
that
is,
you
know,
meant
to
gonna.
Keep
the
project
moving,
but
not
super
technical
and
I
create
an
issue
and
then
I
pull
that
issue
into
the
board
as
a
card
a.
A
That's
a
great
question,
so
everything
that's
in
so,
as
I
mentioned,
I
created
projects
at
two
different
levels:
the
the
boards
and
the
projects
that
are
in
this
project
management.
Repo,
are
only
for
the
issues
created
in
the
project
management,
though
so
I
have
to
create
a
special
project
manager,
tat
task
within
this
project
management
space.
But
if
you
go
up
a
level,
if
you
go
to
the
work.
B
A
A
Yeah
right
right
right,
so
that's
that's
what
I've
been
working
on?
If
you
guys
like
it,
you
know
you
can
create
an
issue
and
tell
me
you
like
it
and
then
thumbs
up
it.
If
you
have
things
you'd
like
to
see
in
this,
you
can
also
create
an
issue
in
that
project:
management,
repo
and
I'll,
just
kind
of
start
picking
away
at
building
this
out,
especially
for
some
of
the
larger
scale,
projects
that
we're
doing
things
that
are
our
grant.
So
the
real
time
collaboration
that
you
put
a
lot
as.
A
A
Any
game.
B
B
It's
not
a
secret
and
I
think
we
can
make
the
issue
because
because
I
mean
we
weave,
it's
not
a
secret
that
it's
happening.
What
we
haven't
done
is
kind
of
a
big
PR
blitz
about
it,
because
we
do
want
to
have
a
nice
website
up
the
moment
we
make
the
big
PR
blitz.
But
yes,
it's
not
a
secret
that
that
it
is
happening.
I.
C
Just
want
to
ask
about
the
issue
for
Jupiter
lab
1
point
0,
since
this
is
publicly
accessible
and
since
people
have
publicly
asked
this
issue
might
be
something
that
they'll
take
as
the
official
like
release
date
or
something
so,
oh
they
might.
We
might
want
to
put
a
note
in
there
about
what
the
actual
thinking
on
when
the
one
point
0
release
is
happening.