►
From YouTube: Opening Remarks
Description
Rollin Thomas and Fernando Perez opening remarks, July 11, 2019 Jupyter Community Workshop
A
Kelly
is
in
charge
of
zooms
here
she
set
it
up,
so
all
right.
So
what
I
have
to
do
a
little
introduction
here,
including
a
little
bit
of
safety
stuff,
especially
for
people
who
have
been
here
the
first
time
and
then
we'll
talk
about
the
discourse,
so
there's
a
topic
on
discourse
where
we're
collecting
ideas
for
breakouts
and
things
that
people
want
to
organize
together.
A
So
if
you
want
to
go
out
for
dinner
or
whatever,
that
would
be
a
great
place,
also,
it's
a
kind
of
put
that
stuff
together
talk
about
the
agenda
a
little
bit
and
go
over
the
code
of
conduct.
So
I
was
curious
before
we
got
into
the
safety
stuff,
though,
to
find
out
how
many
people
kind
of
consider
themselves
to
be
from
the
science
user
facility
side.
So
that
would
be
like
people
who
are
part
of
an
experiment
or
a
place
that
you
know
a
telescope
or
something
like
that.
A
So
this
is.
This
is
great,
like
I,
think
most
everybody
raised
their
hand,
at
least
once
so.
There's
like
I
mentioned,
there's
like
45
people,
I
think
when
we
had
an
initial
expression
of
interest
from
people
we
had
like
over
80
people
interested,
and
so
we
you
know
we
have
only
a
finite
number
of
seats
for
people,
so
it
we
really
kind
of
had
we
had
to
turn
people
away.
It
was
it's
pretty
rough
to
figure
that
out.
A
A
The
format
is
we're
gonna
have
talks
in
the
morning
and
then
breakouts
in
the
afternoon.
We
have
about
15
talks
over
three
days.
So,
like
five
each
day
we
have
lightning
talks.
Also,
we
have
a
few
spaces
left
for
lightning
talks.
Lightning
talks
are
like
10
minutes.
So
if
you've
got
an
idea
for
a
lightning
talk,
I
think
we
have
maybe
two
or
three
slots
left.
A
There
is
time
in
the
afternoon
for
breakouts
and
hacking
effectively
that
can
start
at
12:00
noon.
If
you
want
to
grab
lunch
and
then
go
start
hacking,
you
can
at
the
end
of
the
breakout
sessions,
we're
gonna
meet
back
in
here
and
then
have
report
outs
from
the
breakout,
so
just
show
and
tell
what
you
did
during
the
breakout
and
in
terms
of
breakout
topics
we're
collecting
those
on
discourse
I'll
get
to
that
in
a
second,
but
also
there's
a
box
I
drew
on
the
wall
back
there.
A
We
have
these
cool
walls
where
you
can
write
with
with
markers,
was
not
any
kind
of
marker,
but
you
know
whiteboard
marker
and
you
can
say:
hey
I
have
an
idea:
I
want
to
talk
about
pre-spawn
hooks
or
something
like
that.
You
can
you
can
that
there
and
then
people
could
like
put
a
check
mark,
and
then
we
can
write
before
going
to
breakout,
decide
which
ones
we're
going
to
have
enjoy
the
weather
if
it
doesn't
heat
up
or
if
we
have
a
fire.
Hopefully
we
don't
have
like
a
fire.
A
We
can't
even
yesterday
in
nearby,
but
what
you
should
do
here
is
meet
everybody
and
try
to
include
everybody.
In
your
conversations
we
had
this
cool
pac-man
rule
last
year,
if
you
remember,
which
was,
if
you're
standing
in
a
circle
and
there's
somebody
over
here
kind
of
pac-man,
open
the
pac-man
Mouse
and
well
that
person
in
and
if
there's
old
project
that
you
just
haven't
been
able
to
finish.
You'll,
try
to
revive
those
and
get
those
going
and
we're
going
to
have
a
breakout
on
helping
to
prepare
a
white
paper.
A
A
When
we
have
stickers,
okay,
great
gotta
have
stickers
all
right,
so
the
safety
minute
safety
minute
is
should
be
like
a
minute
or
whatever
there's
a
fire.
It
will
be
obvious
because
there
will
be
horns
and
strobes
and
they'll
be
very
loud,
and
what
you
should
do
is
evacuate
to
any
of
the
exits
on
this
side,
the
east
side,
and
if
we're
obviously
we're
in
this
room,
just
shook
out
that
door.
If
you're
on
the
third
floor,
you
can
use
that
door.
Okay,
turn
on
the
fourth
floor,
you're
not
supposed
to
use
that
word.
A
Okay
and
there's
another
assembly
area
which
is
out
that
way
which,
if,
for
whatever
reason,
you're
on
that
side
of
the
building,
just
go
down
and
cross
the
road
that
you
came
up
on
and
go
down
the
stairs
there,
you
can
wait.
Anybody
with
limited
mobility
should
should
go
to
one
of
these
areas
of
Refuge,
which
is
outside
the
northeast
or
South
East
stairwell,
either
side
there
and
then
let
other
people
who
are
leaving
get
in
contact
with
the
build
emergency
team
and
tell
them
that
that
you
might
need
help.
They
have
radios
okay.
A
A
Get
under
the
desks
here
and
hold
on
to
them
and
wait
until
the
shaking
is
all
done,
and
then
you
know
come
out
and
they'll
tell
you
what
to
do
leave
the
building
is
going
to
be
what
they
tell
you
to
do.
Okay,
restrooms
are
right
over
there.
Okay
and
then
recycling
is
a
challenge
here
at
Berkeley
Lab.
It's
a
test
for
people,
usually
there's
like
five
different
ways
that
you
can
throw
away
trash,
but
we
have
compostable
paper
landfill
stuff.
A
If
you
have
questions
just
ask
ask
somebody
that
looks
like
thinking
of
attention
all
right
so
in
terms
of
the
agenda,
that's
all
on
our
website
and
so
I've
included
the
website
URL
up
here,
but
also
I,
think
I've
sent
it
a
few
times
in
an
email.
So
you
should
be
able
to
find
one
of
my
many
emails
just
to
remind
everybody.
First,
two
days
are
gonna,
be
up
here
at
nurse
and
then
the
third
day
Thursday
is
gonna,
be
down
at
bids
which
is
on
the
UC
Berkeley
campus.
A
So
you
won't
ride
the
shuttle
up
and
if
you're
staying
at
the
guest
house,
you
need
to
ride
the
shuttle
down
to
the
north
side
of
campus
and
tomorrow,
we'll
have
instructions
on
how
to
get
two
bids.
If
you
need
them,
but
there's
there
should
be
some
instructions
on
how
to
get
there
on
the
website.
Right
now,
it's
in
the
library,
it's
it's
a
kind
of
bigger
space
than
this
I
think
so
we're
gonna
have
I
mentioned
talks.
There
are
30
minutes
lightning.
Talks
are
10
minutes.
A
The
way
that
we're
gonna
do
the
talks,
the
AV.
It
is
a
little
complicated,
but
we've
over
years
figured
out.
This
is
what
works
best
for
our
system.
We
have
an
AV
system
where
there's
a
Mac
Mini
somewhere
in
the
room
and
the
podium
locks
in
to
that.
Basically,
and
then
it's
projected
here
on
these
screens
and
so
if
you're
giving
a
talk,
what
you
should
do
is
you
should
take
your
laptop
and
you
should
dial
into
the
zoom,
which
is
here
but
don't
join
the
audio,
that
Kelly's
turned
off
all
the
audio
stuff.
A
And
then
you
just
share
your
screen,
and
then
you
can
give
your
talk
over
over
over
the
Mac
Mini.
That
way
and
we're
also
recording
the
talks
and
so
we'll
have
those
posted.
Once
we
get
all
the
recording
worked
out
on
the
agenda
page
same
with
slides
like
to
collect
everybody,
slides
of
course,
and
so
we'll
come
after
you
for
those
of
you.
If
you
haven't
sent
them
already,
I
mentioned
discourse.
A
We
have
a
discourse
topic,
there's
a
link
here
on
the
agenda
where,
if
you
have
an
idea
for
breakouts
or
you
want
to
collaborate
with
people
or
you
want
to
organize
like
I'm
going
to
dinner
or
something
like
that-
and
you
can
do
that
there,
but
we've
seeded
it
with
a
few
ideas
of
things
to
talk
about
breakouts,
but
we're
looking
for
more.
So
please
you
know,
and
if
you,
if
you
like
something
there,
you
know
make
sure
to
say
you
know
thumbs
up
to
that
whatever
and
then
for
the
breakouts.
A
We're
gonna
have
two
or
three
breakout.
Today
we
have
three
rooms
that
we
can
use
in
this
building
for
breakouts
tomorrow.
We'll
only
have
two
because
there's
a
competing
meeting,
there's
also
some
room
across
the
way
in
building
50
that
Shreyas
here
this
is
raise
your
hand
shreya,
so
shreya,
so
we'll
take
people
over
to
breakouts.
A
Yeah
today
is
better
than
yesterday,
all
right,
so
I'm
almost
I'm
almost
perfectly
on
time.
So
this
is
kind
of
the
start
of
the
morning
agenda
we're
gonna
have
Fernando
is
gonna.
Do
a
welcome
in
an
introduction
of
our
keynote
speaker,
who's
Michael
over
there
and
then
Rick.
Wait,
no
Rick's
here,
alright
good,
so
I
don't
have
to
adjust
the
schedule
on
the
fly
here
and
then
we'll
have
a
break
after
those
talks,
there'll
be
two
more
talks
and
then
lightning
talks
and
I
think
we
have
room
for
one
more
lightning
talk
here
today.
A
So
if
you
want
to
propose
lightning
talk
at
the
last
minute,
just
set
shoot
me
an
email
and
I'll
I'll
take
care
of
it.
So
are
there
any
questions
any?
If
you
have
any
questions
at
all,
you
know
you
can
find
me.
You
can
find
any
of
the
other
committee
people
so
just
to
make
sure
everybody
knows
who
those
are
occurs:
Shane
raise
your
hand
Shane
and
Kelli
and
Trey
us
and
me
so
we're
from
lvl
and
then
Dan
back
there
from
from
Brookhaven.
A
So
there's
any
questions,
let
us
know
and
then
finally,
I
need
to
mention
that
we
do
have
posted
on
the
website
a
code
of
code
of
conduct
and
you'll
read
to
this
when
you
showed
up
so
if
you
haven't
read
it
or
in
a
while,
you
should
reread
it
again,
but
if
there
are
any
issues,
please
feel
free
to
contact
any
of
the
of
the
organizers.
This
is
my
my
phone
number
that
goes
straight
to
my
my
cell
phone
and
then
we
can
can
start
working
it
out
from
there.
B
For
that
morning,
everyone
I'm
going
to
keep
this
brief
I'm
not
going
to
use
15
minutes.
My
name
is
Fernando.
Perez
I
have
a
dual
appointment
here
at
the
lat,
maybe
I'm
at
all,
but
now
I
spend
most
of
my
time
down
on
campus
in
two
Statistics
Department
I
want
to
start
by
by
thanking
you
all
for
coming,
in
particular,
by
thanking
the
the
organizing
committee.
Roland
has
done
a
ton
of
work,
but
also
Dan.
B
It's
great
to
have
participating
Chris,
McGrath
I
think
joins
us
tomorrow
from
from
bids
in
from
the
north
Debbie
Shane's,
Reyes
and
Kelly
I
really
appreciate
it.
It's
fantastic
to
see
to
see
this
coming
together.
This
kind
of
began,
probably
what,
like
two
years
ago
at
a
lunch,
a
lunch
at
Caelius
I
think
was
kind
of
one
of
the
first
times
when
we
wouldn't.
We
had
a
conversation
to
try
to
get
sort
of
a
community
around
this.
B
This
path,
big
thing
that
produces
that
right,
big
thing,
maybe
telescope,
maybe
high-throughput,
genome
sequencing
facility,
maybe
light
source,
maybe
particle
accelerator.
Big
thing
wants
and
needs
compute
next
to
it,
because
big
thing
isn't
quite
just
producing
raw
data.
Now
you
actually
want
compute,
potentially
in
the
in
a
semi
lab
right,
because
maybe
you
want
to
run
simulations
on
an
HPC
facility
to
monitor
what
you're
doing.
B
So,
there's
a
number
of
reasons:
why
that
big
thing
big
expensive
thing
producing
data
that
doesn't
want
to
stop
the
wait
for
you
next,
the
big
compute
thing
that
is
up
and
since
Ron,
and
that
wants
to
run
at
peak
capacity
with
a
flat
with
a
flat
utilization
curve.
It's
connected
to
a
human
who
wants
to
take
a
sip
of
coffee
is
to
go
to
the
bathroom,
has
a
brain
that
needs
to
think
needs
to
talk
to
their
colleague,
maybe
to
kind
of
news
about
something
on
the
whiteboard
how
to
make
that
work.
B
It's
an
interesting
and
complicated
technical
and
social
challenge
and
I
hope
that
in
this
workshop
will
come
up
ideas
on,
but
it's
because
I
think
there's
a
lot
of
interesting
technical
work
to
be
done
here.
There's
issues
on
deployment,
customizability
access
to
high
high
throughput
streaming
data,
balancing
of
the
very
wheedle
needs
and
constraints
that
the
people
who
run
these
kinds
of
facilities
have
right.
B
They
don't
tell
you
that
they
want
to
keep
running
the
thing
without
something
to
wait
for
you
just
because
they
want
to
annoy
you
it's
because
the
big
thing
be
it
the
Machine,
the
the
HPC
facility
or
the
scientific
instrument,
is
potentially
very
expensive
to
run
in
a
unique
resource.
So
there's
reasons
for
that
for
that,
but
we
need
we
need
to
get
creative
because
at
the
same
time,
simply
using
our
supercomputers
to
heat
up
the
room.
It's
probably
not
the
best
use
of
those
machines.
B
What
Kathy
Alec,
who
used
to
be
director
of
nurse
once
put
it
as
I,
want
nurses
to
be
the
facility
that
produces
the
most
science
per
megawatt
and
I?
Think
that's
the
right
metric,
not
a
flat
Linpack
Linpack
running
like
a
lot
of
health
curve,
but
but
science
food,
and
that
optimizing
for
the
combined
usage
of
the
human
and
and
the
machinery
Jupiter
is
kind
of
at
the
heart
of
that
that
that's
what
we
do.
B
We
are
a
system,
a
system
precisely
for
human-in-the-loop
interact
if
ik
computing,
Jupiter
laughs
just
coming
out,
hopefully
1.0
will
be
in
the
next
few
days.
I
think
app
is
designed
to
be
a
highly
modular,
very
flexible
system,
precisely
for
interactive
scientific
computing,
with
the
potentials
for
lots
of
extensions,
so
I
hope
out
of
this
workshop.
Good
ideas
will
come
on.
How
do
we
use
these
tools?
These
new
tools,
better
environments?
B
It
also
presents
challenges
in
customization
in
single
user
deployment
and
reproducibility
is
an
area
which
is
a
big
problem
in
in
every
context,
in
this
kind
of
environment,
where
the
datasets
are
gigantic
for
the
instruments,
the
physical
instruments
that
produce
it
are
unique
and
where
the
HPC
resources
are
big.
Reproducibility
is
like
difficult
difficulty
cubed
or
something
like
that.
It's
an
awfully
nasty
problem,
I'd
like
to
see
ideas
coming
out
of
here
and
that
at
least
make
make
a
little
bit
of
a
dent
on
that
problem.
B
We
have
a
few
services
that
the
project
runs
and
I
put
them
at
the
top,
because
the
services
really
do
run
on
the
stuff
that
comes
beneath
beneath
and
services
like
binder,
like
Jupiter,
like
the
Jupiter
demos,
like
can
be
viewer,
we
don't
run
a
ton
of
services
as
a
project
or
service
services
are
an
expensive
and
time-consuming
thing
to
do,
or
these
kinds
of
things
definitely
brought
in
services.
Right
facilities
are
very
much
entities
that
run
services,
compute
or
data
or
scientific
instrument
type
services
for
their
users.
B
Those
run
on
top
of
open
source
software,
that's
probably
what
people
think
of
as
Jupiter
is
the
software
layer.
Yes,
it's
artifacts,
there's
github
repos
that
you
can
dip
install,
install
and
whatnot,
but
I
think
that
one
of
the
important
things
is
that
the
the
and
we've
put
a
ton
of
time
in
in
the
Jupiter
project
in
that
in
ensuring
that
that
software
is
actually
the
embodiment
of
ideas
that
themselves
are
openly
documented,
openly
specified,
argued
and
debated
with
the
community
so
that
they're
a
good
fit.
B
So,
for
example,
we
don't
just
have
something
that
caught
spy
fun
code.
We
developed
a
proper
protocol
for
what
does
it
mean
to
have
an
interactive
UI
that
connects
to
a
stateful
kernel
and
we
worked
with
the
community
on
refining
what
that
should
look
like,
and
today
there
is
about
a
hundred
different
implementations
of
kernels
for
the
Jupiter
particles
right.
So
we
took
the
time
in
documenting
that
the
notebook
format
is
itself
separately
documented.
B
The
JSON
schema
is
specified
so
and
think
of
our
software
libraries
as
an
implementation
of
something
more
fundamental,
which
is
the
which
are
these
founded
standards
and
protocol.
There's
other
UIs
for
the
notebook
we've
produced
to
the
classic
one
in
jupiter
lab,
but
other
teams
have
produced
others
interact,
has
its
own
set
of
new
UI
ocal
team
has
produced
effectively
a
brand
new,
separate
UI,
the
Kolkata
machinery,
the
Google
be
a
fork
but
heavily
customized
fork
of
an
old
version
of
ipython
that
run
on
Google
infrastructure,
so
on
and
so
forth,
and
that's
important.
B
B
It
exists
on
top
the
most
important
base
layer,
which
is
a
community,
and
we
really
emphasize
that
we
want
to
build
a
strong
community.
This
is
a
kind
of
space
where
we
haven't
had
historically
a
lot
of
representation
and
I
would
love
to
see
that
change
to
see
that
grow
by
community
I
mean
people
who
work
together.
People
build
things
together.
People
who
respect
each
other
people
who
new
projects
together,
people
who
get
funding
together,
would
be
we
here
how
many
of
you
come
from
from
outside
of
the
u.s.
B
I
think
it
would
be
great
if
we
do
that
if
we
show
funding
upward
work
done
with
with
fund
from
funding
aid
from
funded
by
by
national
or,
as
is
the
case
in
europe,
transnational
funding
agencies
in
operation
and
in
cooperation
also
with
the
u.s.
funding
agencies,
we're
talking
a
lot
more
do
e
to
NSF
to
nih.
This
kind
of
pattern
is
now
hitting
it.
This
kind
of
thing
used
to
be
a
little
bit
more,
the
purview
of
do
we
do.
He
has
been
kind
of
a
this
kind
of
stuff.
B
It
has
been
involved
with
things
like
particle
physics
at
cern,
which
is
the
the
prototypical
example
of
this
kind
of
thing
for
30-40
years.
But
this
pattern
is
now
happening
across
agencies.
All
the
agencies
are
now
having
this
kind
of
flavor
of
problem.
I
think
it
would
be
fantastic
if
effectively,
this
community
emerges
as
a
set
of
leaders
in
that
space
across
national
and
international
boundaries,
I'm
always
happy
to
help
when
I
can.
With
these
things,
I
will
be
in
Europe
in
October.
B
If
anyone
is
if
any
of
the
European
folks
are,
gonna
are
gonna,
be
there
talking
to
Swiss
universities
about
open
science,
I'm
always
happy
to
help
if
I
can,
but,
but
nothing
has
to
bottleneck
on
me
or
nuts.
I
hope
that
lots
of
collaborations
will
emerge
on
in
this
space,
as
the
project
does
need,
need,
resources
and
I.
B
So,
hopefully
we
can
have
an
impact
in
that
space
and
I
do
think.
It's
scientifically
the
right
thing
to
do.
It's
not
just
because
I'm
interested
in
funding
the
project,
but
it's
because
I'm
convinced
it
has
the
right
building
blocks
for
a
class
of
problems
which,
I
think
is
just
emerging
I,
think
we're
kind
of
at
the
beginning.
B
Obviously,
CERN
has
existed
from
I
think
the
explosion
of
this
across
the
sciences
is
something
that
is
just
starting,
so
I
think
this
we're
very
much
at
the
beginning
of
something
that
I
hope
will
be
very
fruitful
and
productive
workshop.
Thank
you
again
for
coming
thanks
to
the
organizers,
thanks
to
Bloomberg,
for
funding
this.
We
we
have
had
a
fantastic
collaboration
with
Bloomberg
as
a
company.
B
They've
they've
put
a
ton
of
effort,
both
human
development
lab
and
this
workshop
is
made
possible
because
Bloomberg
don't
resources
for
actually
a
collection
of
community
workshops
on
some
of
the
folks
in
this
room.
Work
Oh
in
Paris,
working
on
a
different
workshop
around
that
hurting
I.
Think
Lindsay's
gonna
give
a
talk
about
some
of
that.
Some
of
that
work,
lightning
talks
and
there's
another
in
the
funding
that
the
company
has
provided
for
these
kinds
of
community
community
driven
events
around
different
parts
of
the
project
has
been
extremely
helpful
and
useful.
B
So
if
you
see
some
of
the
Bloomberg
folks
like
Polly
when
I'm
in
the
back
thanks
so
anyway,
I
won't
take
more
time.
I
want
to
close
by
what
keynote
speaker
Michael
Milligan
from
the
Minnesota
saluting
Center.
We
first
met
in
2016
online.
You
organized
the
kind
of
a
baby
version
of
something
like
this,
which
was
a
one-day
workshop,
basically
to
talk
about
Jupiter
hub
deployments
and
for
people
to
discuss
how
they
were
trying
to
use
Jupiter
hot
in
their
own
environments
and
Michael
had
a
great
presentation
about
the
work
he
was
doing.