►
From YouTube: SunPy developer Discussion
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
E
C
E
A
A
A
A
E
A
A
A
A
A
Seriousness:
okay,
I'd
like
to
bring
up
something
that
we
were
discussing
at
the
end
of
the
last
meeting,
but
didn't
really
come
to
a
conclusion
on
the
living,
which
is
the
things
that
we
are
my
lost,
owning
the
0.7
and
the
more
more
than
this.
What
we
are
using
our
milestones
for
on
get
off,
evimer
0.7
being
the
case
in
point,
but
not
actually
the
core
topic
of
this
discussion.
A
E
A
F
A
I'm
not
sure
that
we
need
that
differentiator
on
and
that
oh
well,
the
other,
so
I
don't
think
we
need
something
that
I
would
be
nice
to
have
this
in
your
point,
seven
because
it'd
be
nice
to
have
any
open
issue
in
nor
point
seven
like
there's,
none,
no,
no
open
issue.
Would
we
go
or
no?
We
don't
want
that,
no
point,
seven
at
all
or
want.
Why
would
why
would
we
merge
that
would
be
stupid
and.
A
A
A
A
A
B
A
A
It
like
you,
just
need
to
go
through
every
method
in
map
and
make
a
set
of
relatively
straightforward
changes
like
because
at
the
moment,
what
we're
doing
in
lots
of
places
in
the
map
methods
is
copying.
The
input
instance
and
then
modifying
the
matter
of
tribute
in
the
data
review.
What
we
need
to
be
doing
is
copying
the
input
instance
modifying
matter
and
data
into
a
separate
into
a
separate
variable
and
then
creating
a
new
map.
Instance
on
the
way
out,
rather
than
modifying
data
and
map.
A
E
A
Don't
think
so
in
the
moment
by
okay,
so
that
then
having
had
some
level
of
agreement
that
and
probably
definitely
make
those
changes
about,
add
that
to
my
list
or
someone
else
could
do
it.
Let
me
see
if
I
can
find
see.
Astro
play
they've
list,
but
one
of
the
things
that
are
currently
milestone
does
not
point
seven.
Are
we
actually
going
to
block
nor
points
ever
more?
I
can
see
a
beautiful
one.
I
would
get
rid
of
two
that
I'll
get
rid
of.
D
F
F
Them
and
then
they
become
this
thing,
we
should
just
focus
in
on
like
curve
improvements
and
ion
tube
right.
Is
that
basically,
what
people
want?
Oh.
A
Yes
and
the
definition
of
the
word
related
in
that
sentence
needs
to
be
strict
right,
so
that's
something
that
is
definitely
related
and
not
like.
This
may
come
up
in
the
course
of
doing
that,
or
we
may
implement
this
in
the
course
of
doing
that,
like
say,
for
instance,
a
meta
data
object
for
like
of
yes,
we
may
find
that
we
need
that
to
do
the
mat
factory,
but
there
are
gonna
be
might
not
so
it
shouldn't
be
blocking
up
with
seven
in
and
of
itself
if
it
gets
pulled
in
than
fine
yeah.
F
A
15,
don't
15
15,
15
15
14
see
now.
This
is
the
other
thing
like
bugs
in
theory
block,
although
I
would
argue
that
15
14
no
15
14.
If
in
itself
is
a
book,
we
shall
be
returning
a
ninja
turtle.
Takedown
bugs
can
block
releases,
that's
fine!
Why
is
this
I
mean
we
have
the
Bible
English,
so
I,
some
of
the
other
bugs
should
probably
be
blocking
it
as
well,
like
quite
a
few
I
would
argue
but
victims,
we
add
more
things
to
this
in
a
minute,
so
range
object
can
go.
F
A
F
A
Yes,
but
that
is
kind
of
what
is
currently
happening
with
no
point:
seven
right,
because
there's
a
lot
of
bugs
that
are
open,
that
we
aren't
popping
water
because
they're
not
we've
deemed
to
be
low
priority
yeah.
So
we
we
definitely
need
a
higher
priority
low
priority
thinking.
So
I
put
a
link
to
the
Astro
pie
wiki
in
the
chat,
which
explains
their
labeling
convention.
I
am
effectively
proposing
that
we
adopt
this
along
with
priority
high
median
and
mode
I.
B
F
F
F
A
A
F
A
B
F
A
Really
should
do
more
work
on
our
test
coverage.
It's
assured
it's
not
know.
I
know
I
would
say
that
we
ship,
we
didn't
do
anything.
We
should
make
that
like
I
can't
even
rude
where
we
are
low,
because
coveralls
been
down
for
so
long
I
would
say
all
right,
I
would
remove
that
one,
but
replace
it
with
another
one
which
is
to
have
a
reasonably
comprehensive
set
of
figure
tests
or
even
extra
lives.
Now
we
have
now
we
have
a
figure
testing
infrastructure.
A
B
B
F
F
A
But
ok:
well,
we
drove
it
for
the
moment,
but
I
would
argue
that
we
should
have
some
kind
of
test
increase
in
north
point.
Seven,
I
created
a
set
of
those
issues
that
incremented
it
by
like
a
section.
Each
release,
I
didn't
leave
the
idea
to
sort
of
forces
to
move
forward
right
tests
for
code
is
in
there
with
no
test,
so
I
would.
B
F
A
F
But
think
of
it
from
the
users
perfected
I
mean
if
you
can
go
around
and
say
you
know,
this
release
is
going
to
be
the
light
curve
release
and
we're
gonna
have
a
you
know,
increase
in
functionality
like
her
right
and
then
you
know,
and
that's
all
that's
in
there
and
it
you
know
we
don't
take
a
really
long
time
because
we
get
stuck
on
other
stuff
like
like
which
mccauley
it'll
it'll
come
out
on
time
and
people
will
be
happy
with
it
and
then
the
next
release
we
could
say,
oh
as
next
release
is
actually
we're
doing
a
lot
of
back
end
stuff,
and
so
we
manage
expectations
from
our
users
that
well,
this
release
is
actually
you
know
a
test
test
release.
F
It's
not
gonna
affect.
You
know
your
your!
It's
not
going
to
give
you
a
lot
of
features
so
I
have
like.
We
should
run
things
like
that,
a
little
bit
more.
I
think,
because
if
we
there
are
all
these
things
that
we
can
always
say
that
right
like
like,
why
not
you
know
squash
this
one
buddy
alright,
like
oh
there's,
a
bike
here?
What
about
all
the
other
buttons
why
these
bikes
yeah,
we
could
say?
F
H
F
Yeah
I
think
it
would
be
nicer
if
we
just
had
a
theme
for
the
release
and
that
aim
would
just
be
like
herb
and
Python
3.
That's
how
you
tell
people
what
we're
doing
right.
The
next.
The
next
theme
release
could
be.
You
know,
test
coverage,
that's
what
we're
doing
this
release.
It's
just
half
cut
this
covered.
A
I
think
that's
the
thing.
I
think
you
need
to
sort
of
I'm,
not
disagreeing
we're
going
to
say,
but
I
wouldn't
have
a
whole
list
of
the
theme
of
which,
which
was
test
coverage,
because
that
would
be
boring
and
how
many
weeks
that
has
less
features
slightly
less
features
but
he's
also
test
opera.
But
oh
yeah.
F
A
About
we,
okay,
so
the
reason
that
that's
really
all
right,
though
so
the
thing
is,
if
we
have
I'm,
not
necessarily
disagree
with
you
for
a
start
I'm
just
saying
we
should
include
tests
coverage
issues
in
milestones,
because
it
gives
us
a
deadline,
arbitrary
deadline,
given
by
says
that
it
forces
us
if
we
obey
our
own
walls
if
force
is
asked
to
increase
the
test
coverage,
so
anything
in
the
general.
We
should
have
test
coverage
milestones
in
releases
because
it
will
kick
us
into
writing
more
tests.
A
Then
releases
are
supposed
to
be
bug,
fixes
they
not
I
mean
you
could
you're,
not
wrong,
you
could
you
could
you
could
increase
test
coverage
and
appointments
but
I?
Don't
we
don't
really
need
milestones,
point
releases
because
there's
nothing,
there's,
never
anything
blocking
appointment,
because
the
point
release
just
fixes
a
bug.
So
let
you
release
a
point
release
for
every
both
we
fix
in
the
null
point.
Six
runs
from
now
almost
yeah,
but
that
means.
A
All
right,
but
then
we're
holding
up
bug,
releases,
bug
fixes
which
could
potentially
be
critical
bug
fixes
that
we
should
just
push
out
as
soon
as
we
fix
them,
because
they're
stopping
users
doing
something
they
should
be
able
to
do
so,
which
we
should.
My
point
is:
if
we
need
to
do
a
point
release,
we
should
never
be
holding
it
back,
because
it's
point
release
it's
designed
to
fix
an
existing
usability
issue
or
some
other
on
Quora
behavior.
That
should
be
burned.
A
F
Well,
ok,
I
knew,
we've
probably
talked
about,
there's
no
I,
don't
think
I've
convinced
enough
people.
No
that's
fine.
We
can
move
on
I
would.
A
As
I
say,
I
think
I
think
it
would
actually
be
very
easy
to
achieve
that
goal.
I
think
writing
m
for
tests
for
map
and
like
Irv
and
map.
You
would
probably
fix
that
by
right
over
there
is
the
plotting
code
by
line
length
is
long
and
it's
not
tested
at
all
at
the
moment.
So
I
think
that
actually
just
writing
a
few
podcasts
would
probably
achieve
eighty
percent
coverage.
A
B
A
F
Ready
later
where's
a
link
to
the
point
release
framework,
they
are
using
us.
A
F
If
know
where
it
is
right
on
the
website
again,
Oh,
Andy
and
I
went
through
this,
and
we
found
some
problem
no
and
I
thought
we
should
discuss
how
we
really
want
to
handle
this.
So
as
a
current
lease
is
set
up,
the
calorie
it
is
its
own
repo
and
in
head
and
therefore
home,
hosts
its
own
website.
So
when
you're
clicking
on
the
gallery,
you're
actually
opening
a
new
website,
it
looks
exactly
like.
F
Nice,
yes
and
terrible
the
terrible
way
that
we
do
that
right
now,
but
any
case
nope.
The
way
it
works
currently
is
that
inside
the
gallery,
people
are
supposed
to
provide
notebooks
with
all
the
content
removed,
all
the
spell
outputs,
the
loop
and
then
Adam.
So
that's
so
that
you're
not
committing
like
figures
right
images
with
can
be
included
and
inside
the
side
you
know,
but
could
be
brand
something.
F
That's
essentially
binary
data,
so
we
don't
want
that
in
there
and
then
there
is
a
script.
This
is
no
Stewart
stuff.
This
is
Kirk
call
deploy
and
it
actually
runs.
This
thing
called
run:
I,
hi,
Steve,
yes,
actually,
no
go
keuka!
You
want
this
quit
the
script.
Deploy
which
concludes
this
thing
called
run.
I
pi
and
run.
F
I
pi
will
actually
it
all
since
you
compile
your
notebook
and
set
all
the
output
and
then
frigid
HTML
file
as
and
dump
I
ipython
notebooks
can
do,
and
then
that
gets
pushed
up
into
the
web
page
parked
up
the
of
this
repo
now.
So
the
idea
for
that
would
that
you
know
you're
forcing
them
to
be
actually
ran,
run
to
run
each
of
it.
I
find
a
notebook,
great
and
then
you're
also
not
including
any
binary
data
in
your
notebooks,
which
is
also
nice.
F
There
is
one
problem,
at
least:
is
that
run
I
pi
seems
to
be
broken
and
probably
I
don't
know.
Maybe
they've
stopped
development
on
it,
because
it's
been
a
long
time.
That's
the
update
anything
and
it
does
not
seem
to
work
on
the
new
super
things:
q,
brunette
books.
F
The
other
thing
is
that
we've
actually
failed
in
addressing
the
main
issue,
which
is
that
which
is
not
not
committing
binary
files
through
the
repo
you're
not
being
committed
to
the
master
repo,
but
they
are
being
committed
in
the
HTML
outputs
to
the
e
web
part.
The
repo.
A
Yeah,
that's
a
problem
that
you
that
we're
all
right,
fine,
but
that's
a
problem
with
using
github
pages
as
a
hosting
mechanism
is
that
you
have
to
you-
have
to
commit
binaries
to
the
get
up
pages
branch.
But
the
way
there's
some
play
gallery
scripts
work.
Is
it
doesn't
store
history
for
github
pages,
so
every
time
you
push
to
get
up
pages,
it
complete,
destroys
the
history
of
the
branch
and
starts
again,
so
it
doesn't
balloon
the
size
of
the
repo,
because
there's
no
history
and
you
get
our
pages
brunch
I-
see.
F
I,
don't
know
okay
well,
so
ronaiah
pies
broken,
and
so
what
we
did
was
we
essentially
just
went
in
and
started
editing
the
pages
themselves,
which
means
that
nothing
is
in
sync
anymore.
If
you
ran
deploy,
it
would
destroy
the
sump
I
gallery
page.
Currently,
if
you
got
it
running,
which
I
don't
think
you
will
ill
to
another
all.
A
A
Except
not,
apparently,
a
vision.
Has
anyone
got
that
link
I've
sent
it
twice
now
all
they
do
is
come
up
and
at
the
root
directory
of
the
somme,
pyar
mein,
Sohn
file
repository
and
which
contains
Python
scripts,
and
then
we
have
a
sphinx
extension
that
runs
those
Python
scripts
and
compiles
them
into
our
main
documentation
in
the
form
of
a
gathering.
A
A
A
A
A
A
F
This
is
not
exactly
what
we're
talking
about
there
in.
This
is
essentially
a
gallery
more
like
a
map
plot
lib
type
gallery,
which
is
like
a
plot
and
the
code,
a
plot
and
the
code,
as
opposed
to
what
we've
been
using
so
far,
which
is
ipython,
which
is
like
here,
is
a
bunch
of
different.
You
know:
here's
like
a
almost
like
a
guide.
You
know
like
here's,
a
workflow,
and
so
this
workflow
would
not
fit
in
this,
and.
A
So
have
a
look
at
the
notebook
styled
example
section:
if
you
go
back
from
that
back
up,
I'm
trying
to
I
haven't
actually
looked
at
it
yet
yeah.
If
I
don't
know,
I
think
this
knees
looking
at,
but
I
think
I'd
see
I'm
not
convinced
by
the
model
of
using
actually
using
the
ipython
notebook
to
run
the
gallery.
I
think
it's
potentially
slow,
because
I
think
we
have
the
problems
that
we
have
just
run
into
I.
Think
it's
not
necessarily
easy
to
maintain.
A
A
A
C
D
We
have
a
more
basic
question.
We
need
to
address
really
as
regards
to
gallery,
which
is
not
actually
hosted,
but
what
should
be
in
it
so
bring
this
up,
because
what
we
have
right
now
or
essentially,
some
little
guides,
but
we
have
a
user
guide
in
the
dots.
So
the
gallery,
to
my
mind,
should
be
something
especially
different
from
the
user
guide.
It
has
to
be
something
else
and
its
own
right
and
right
now,
I,
don't
I
don't
feel
like
it
is
like
that.
So
it
is
a
more
fundamental
question.
It's.
D
Not
different
I
mean
there's
like
one.
It's
like
curved
tutorial.
I
was
like
a
match
each
other,
but
we
got
rid
of
it.
Okay,
there's
some
PI
over
you,
which
is
a
bit
different
again.
You
know,
I,
don't
know
like
I,
don't
feel
like
the
gallery
is
sufficiently
like
it's
duplication
of
effort
right
now,
so
it
needs
to
be
something
that
stands
on
its
own.
So
oh.
A
Hello
scikit-learn
is
currently,
of
course,
is
I
should
have
guessed
that
so
I
can't
learn
is
currently
using
this
Sphinx
gallery
extension
to
build
its
examples
by
rotary
so
yeah.
We
just
find
you
no
no
wait,
I'm
coming
back
to
what
Andy
said,
give
me
I'm,
just
sorry
I,
because
I'm
using
two
computers
and
they're
not
linked
together,
it's
difficult
for
me
to
generate
links
for
you,
so
they're
using
it.
The
the
things
in
the
psychic
learn
gallery
and
also
to
a
lesser
extent.
The
psychic
image
gallery
are
singular
trained
examples
right.
A
This
is
not
a
gallery,
isn't
it
these
examples,
so
have
a
look
at
this
scikit-learn
gallery
and
pick
out
one
of
them
going
to
go
with
face
completing
it.
That
is
a
whole
load
of
code.
That
does
one
thing,
so
it
would
be
I'm
trying
to
think
of
simply
so.
It
would
be
like
and
be
a
Lolena
system
like
plotting
a
composite
map
from
downloading
data
and
plotting
a
competent
map
or
plotting
a
map
koupe,
making
a
map
tube
and
a
light
curve
of
the
same
flare,
oven
or
I'm.
A
D
I
think
in
my
mind,
I
thought
Laura's,
the
gallery
yeah.
You
have
some
sort
of
like
not
a
user
gay,
but
you
have
some
sort
of
like
workflows
like
things
that
people
would
actually
want
to
try
and
do
that
there
are
things.
I
belong
in
a
basic
guide,
but
yeah.
You
know
things
that
people
ask
you
about
a
lot
like
iTunes.
Yes,.
A
Yeah
kind
of
like
it
kind
of
like
a
so
the
map
that
live
gallery
is
really
good
as
well,
because
you
so
the
map
obviously
map.
Well,
it
is
a
little
bit
different
because
everything's
our
plot.
Obviously
everything
is
a
plot.
So,
like
the
matplotlib
gunnery,
you
can
scan
down
it
and
see
a
plot
that
looks
kind
of
like
the
one
that
you
want
and
then
plot
it,
and
then
you
get
a
whole
self
contained
code
from
making
that
kind
of
plot.
A
It's
quite
easy
to
inflect
your
data
to
get
that
not
so
I
think
we
can
kind
of
pull
that
over
for
some
pie
in
some
ways
like
how
to
plot
a
aia,
HMI
image
and
an
aia
in
which
next
one
another
having
downloaded
the
data
or
how
to
create
a
animation
or
how
to
you
know
these
kind
of
things.
More
self-contained,
singular
examples.
Yeah.
F
F
H
A
F
A
F
F
Next
conference
comes
up
and
I
need
my
notebook,
so
what
do
I
do
I
think
you
already
lived
in
it
right.
It's
your
notebook
like
that
it
so
yeah,
but
somebody's
gone
ahead
and
improved
the
gallery
netbook
for
a
little
bit
and
I'd
like
to
have
that
now,
I
have
to
like
firsta,
to
my
hand,
it's
pain.
Yes,
it's
so
are
they
didn't?
F
You
have
and
then
the
other
things
that
people
no
one
in
their
right
mind
their
providers
with
with
these
you
know
who
the
docs
these
galleries
in
the
docks,
but
people
are
already
writing
stuff
in
notebooks
and
it'd,
be
really
easy
for
us
to
lick
and
just
say:
oh,
could
you
give
us
this
as
a
notebook
right
and
they
would?
They
would
be
very
easy
able
to
provide,
but.
A
That's
the
same
argument
we
just
made
a
minute
ago,
which
was
its
we're
going
to
have
notebooks
and
we're
going
to
want
to
put
them
in
the
gallery.
That's
easy!
That's
run
it
through
mb
convert
or
even
mb
convert
you
saved
it
out
with
the
web
interface
as
a
Python
script
and
I,
don't
think
porting
into
this
would
be
difficult.
So
now,
let
me
counter
point
with
some
advantages
of
doing
it.
This
way.
Firstly,
Python
scripts
are
substantially
substantially
easier
to
manage
your
version
control.
A
A
You
can't
do
it
and
you
better
have
to
treat
them
as
binary
files
yeah,
and
you
can
probably
get
it
right
and
it's
different
and
it's
difficult
to
reef
it's
difficult
to
automatically
run
them,
as
proven
by
the
fact
that
run
I,
pythons,
broken
Ron,
I,
PMPs,
broken
so
difficult
to
automatically
bond
them.
So
but
it's
very
very
easy
to
ultimately
by
napoleon
python
script.
Obviously,
but
I
don't,
I
think.
A
Yes,
it
does
I
your
galleries
broken.
All
of
my
people,
brief
run,
I
play
ymp
is
broken,
I,
don't
know
if
we
ever
manage,
there's
no
point
in
starting
up
a
gallery
where
we're
managing
it
by
hand
when
asked
that,
because
that
is
stupid,
because
that
is
a
way
that
then
gallery
is
going
to
get
neglected.
The
gallery
is
already
gone
neglected.
It's
got
an
automatic
builder
running
on
it.
F
Right
but
they
got
neglected
because
the
unmet
Beltre
was
kind
of
a
pain
right
and
only
fun
writing
an
easy
one
based
on
looks
yeah.
You
just
sank,
you
know,
send
us
your
network.
Somebody
has
to
look
it
over
anyway,
because
it's
a
notebook
is
served
like
a
it's
almost
like
a
paper
right.
You
just
gonna
have
to
be
a
little
bit
back
up
for
it
with
like
what
content
is
in
the
notebooks.
I'm
gonna
have
to
reason.
B
E
F
And
we
really
don't
I
mean
as
it
stands.
I
agree
with
you
love.
We
should
probably
sell
it
but
third
term.
This
is
not
an.
A
E
A
Volunteering,
oh
yeah,
don't
volunteer
to
do
it
because
all
right,
so
the
other
advantage
to
this
is
that
all
comes
back
into
one
repository.
It's
all
back
in
Maine,
some
pipe
and
you
download
some
time.
You
get
all
the
examples
and
we
can
link
to
the
examples.
I
think
it's
much
it's
much
more
useful
waffle
like
this,
the
motorola.
C
A
Part
of
the
main
documentation,
which
means
they
won't
get
forgotten.
When
we
do
da
three
factors
they
we
can
run
them
as
tests
and
make
sure
that
they
actually
execute
in
the
sizes
a
test
script.
But
if
that
property
we're
using
something
we're
using
a
gallery,
build
infrastructure
that
is
maintained
not
by
us.
This
is
a
big
advantage.
It
reduces
the
amount
of
time
we're
spending
liking
a
gallery
when
we
could
be
writing
code.
A
F
A
F
A
Support
where's
that
you
can
have
this,
so
markdown
is
exceedingly
limited
in
what
you
can
do,
so
it
does
titles
which
are
supported
by
RS,
see
ya.
Markdown
table
is
an
extension,
and
but
it's
also
the
same
as
our
to
you,
so
I,
don't
even
show
them
that
supported
in
high
five
book.
Bold
is
the
same.
Italic
is
the
same
strikethrough
same
most
of
the
stuff.
This
markdown
is
also
rst.
F
A
F
Know
Dennis
divorced
from
the
actual
notebook,
so
yeah
I
think
it's.
So
what
start
messing
around
with
the
ipython
notebook?
You
just
split
the
files.
You
pray
more
work,
you
haven't,
you
haven't
solved.
The
problem.
You've
now
created
two
different
versions
of
the
same
thing,
one
of
which
is
actually
being
tracked
and
possibly
be
changed
by
other
people,
one
of
which
is
like
sitting
on
your
computer.
Yes,.
A
The
ipython
notebook
but
you'll
see
your
release
from
your
perspective
of
running
stuff,
writing
notebooks
for
specific
things
and
also
using
them
as
an
example
gallery.
I'm
saying
that
you
should,
just
if
you
teaching,
you
should
have
a
life
in
notebook
and
you
should
use
that
teaching
version
for
teaching.
A
If
I'm
saying
that
we
should
have
a
coherent
gallery
that
is
version,
control
and
managed
and
dipped
and
built
into
the
main
Sun
pipe
repo
and
part
of
our
documentation
as
part
of
some
pie,
these
two
things
all
right
sure
it
would
be
nice
if
they
could
be
the
same
thing.
But
I'm
saying
we
have
a
solution
that
does
this
perfectly
and
you're
cutting
it
down,
because
it
doesn't
also
to
fill
your
personal
requirement
for
being
able
to
combine
your
teaching
material
with
the
Songhai
cover.
A
People
to
contribute
right,
everybody's
this.
This
makes
it
easier
for
other
people
to
contribute
because
you're
dealing
with
a
text
file
and
not
an
ipod
and
notebook.
If
you
already
have
an
ipod
on
notebook
right
into
text
file
contributing
if
you
don't
already
have
an
eye
by
the
notebook
and
you're
modifying
it,
this
is
substantially
easier.
A
F
A
A
Notebooks,
which
I
think
is
a
failing-
no,
no,
don't
get
me
wrong
with
my
love.
I
Python,
notebooks
I
use
them
a
lot
for
lots
of
different
things,
I'm
just
saying
that
for
the
except
having
sort
of
set
of
coherent
examples
of
how
to
use
something
that
you
may
not
be
the
perfect
format
for
how
we
want
to
use
them.
In
fact,
I
would
argue
that
they're
not
the
perfect
format
for
how
we
want
to
use
them,
and
don't
you
think
that
somebody's
gonna
solve
this
issue
of
fi
thunderpuss,
and
maybe
someone
will
solve
it.
A
Maybe
someone
will
solve
it
in
six
months
before
we
get
around
to
implementing
this.
Maybe
someone
will
solve
in
five
years
time,
and
then
we
can
convert
all
our
Python
scripts
and
notebooks
once
and
move
into
that
happy
note
program
I'm
just
saying
there
isn't
the
solution
at
the
moment
and
I'm
suggesting
that
we
don't
try
and
make
one.
A
E
D
A
A
A
Would
be
yes,
it
would
be
better
if
that
was
split
up
into
separate
things
like
small
standalone
sections,
because
I'm
thinking,
I'm
thinking
of
like
I'm
thinking
of
likely
that
Yoga
is
my
providing
something
which
is
more
useful.
Oh
I'm,
thinking
about
a
user
that
comes
out
by
two
you've
got
a
user
who's
using
sunlight
the
last
month,
and
they
want
to
know
how
to
create
a
lack
of
maybe
they're,
not
answerable.
A
A
Things
it
workflows
something
maybe
something
that
takes
like
two
pages
in
the
documentation
where
they
explain
all
the
sections,
and
you
can
always
go
through
and
read
that
problems
and
it's
very
useful
prost
agreement.
But
the
example
is
a
script
that
will
do
exactly
what
you
want
and
you
can
look
at
it
and
you
can
just
run
it.
D
Yeah
well
I
think
we'd
like
to
get
more
so
set
more
opinions
of
it
from
the
rest
of
the
dev
team
as
well
about
gallery
content.
A
A
A
A
That
that
involves
having
a
selection
of
like
I,
think
that
involves
being
out
in
this.
G
A
G
F
Yes,
but
I
mean
the
problem
is
also
that
you
won't
get
informed
this.
You
won't
get
informed
answers
because
people,
it
I,
don't
think
people
understand
I
mean
we're
having
our
time
scribing
exactly
what
we
want,
each
other
and
and
what
people
will
do
is
default
to
what
they
know
best
right,
and
that
is
you
know
they
know,
I
find
the
notes.
If
I
go,
I
want
to
buy
them
if
they
want.
You
know
they're
really
familiar
with
the
map.
Public
galleries.
I'll
just
say:
oh
I
want
something
like
the
puzzle,
killer,
Oh.
F
G
F
Don't
know
but
but
I
think
fundamentally,
Stewart
I
think
what
we
should
do
right
now
is
before
we
start
figuring
out
exactly
what
we
need.
I
think
we
maintain
the
gallery
as
it
is
with
ipython
notebooks.
F
We
were
going
to
add
the
a
rep
comparison
to
book.
My
brew,
the
workshop
know
folks
that
I
put
together.
I
think
that's
my
hand
is
like
not
a
big
deal.
They
already
exist.
Mines
will
make
them
available
in
a
place
where
people
can
find
them
and
and
then
I
think
we
should
put
together
like
a
gallery
like
you
were
suggesting,
like
your
show
here,
I
think
that's
a
really
great.
A
Scikit-Learn
examples
and
the
Astro
by
tutorials
they
have
to
apply
tutorials
are
long
like
em,
like
which
one
was
it
by
Bishop.
Viewing
a
manipulating
fits
files
is
long.
It's
got
lots
of
description,
it
does
multiple
things,
it
does
image
stacking,
plotting,
a
histogram
and
writing
black
out
again,
and
it
has
exercises
at
the
end
right,
whereas
the
sky,
kid
learn.
Examples
are
one
relatively
short
script,
which
does
a
thing.
A
Nice,
so
we
could
greet
it
and
I
think
the
tutorials
work
better
out
of
a
notebook,
because
it's
easier
to
download
the
for
people
to
download
and
run
it
in
places
you
I
know
if
it
works
so
well
for
teaching
in
that
kind
of
style,
right
I'm,
but
the
example.
The
better
for
like
sort
of
having
a
script
is
better
for
examples,
because,
most
of
the
time
people
are
doing
the
kind
of
work
in
a
reproducible
context
where
they've
running
scripts
to
do
different
ya.
F
A
I
think
there
is
yeah,
so
I
think
like
the
first
two,
you
could
combine
those
into
one
thing,
but
you
could
have
short
running.
You
could
have
very
short
things
and
slightly
longer
things
in
a
example
gallery.
If
you
wanted
to
and
like
but
yes,
I
agree,
aerosoft
I
agree
with
your
cat,
categorization
I.
Just
think
we
could.
So
most
of
some
of
your
first
first
point
is
for
us
specifically
I.
Think
some
of
the
first,
like
very
short,
things,
are
in
the
docks,
but
should
probably
be
in
the
docks
and
the
example
Gary.
A
If,
depending
because
then
it
depends
where
someone
comes
to
look
for
something
right.
So
something
might
be
easier
to
find
in
the
example
gallery
than
the
documentation
and
and
then
tutorials
/
long
talk
so
on
the
topic
of
tutorials.
If
we
wanted
to
create
a
collection
of
tutorials,
then
drew
Sam,
Nabil
and
I.
Did
that
workshop
and
there's
our
notebooks
for
the
tutorials?
For
that?
Some
of
those
are
not
some
by
the
koala
on
some
pipe
but
I
feel
as
though
they
should
probably
he
advertised
somewhere
someplace
like.
F
A
A
A
Or
done
first
yeah
also
that
he's
not
a
high
priority.
Does
discussion
I,
put
it
out
there,
because
it's
something
I.
Couldn't
I
created
an
issue
for
it,
but
it's
something
that
we've
been
talking
about
a
bit
before
and
Jack
and
I
were
having
a
convo
started
to
have
a
conversation,
value
and
another
issue
and
I
wanted
to
aggravate
it
properly,
and
but
it's
not
something
that
we
need
to
worry
about
too
much
of
them.
For
a
start,
the
functionality
that
we
want
doesn't
exist
in
that
for
quite
able
at
the
moment.
A
So
it's
it's
actually
a
completely
irrelevant
discussion
from
that
point
because
we
couldn't
do
it
tomorrow,
even
if
you
wanted
to
but
I'm
as
I
have
made
clear
in
that
issue.
I
have
reasons
why
I
think
it's
a
good
plan,
so
perhaps
it
is
something
that
we
could
just
do
over
text.
Is
it
it's
not
urgent?
So
I
see
you
later
a.