►
From YouTube: SunPy Developer Discussion
Description
Agenda:
* 0.5.0 wooo yeah party
* SunPy Gallery!
* 0.6.0 - 0.8.0 roadmap
* PR swatting?
* AoB
A
A
B
You
know
I
mean
yeah,
the
rotating
really
pisses
me
off
lack
of
a
better
term,
and
obviously
I
am
not
without
fault
there.
I
should
have
I,
never
did
actually
check
out
that
PR.
So
you
know
it,
you
know
it
got
merged
because
it
had
to
get
merged
or
it
got
the
approvals
and
everything
me
it
followed
our
normal
procedure
right.
So
the
fact
that
I
didn't
check
it
against
the
stuff
is
partially
my
fault
for
sure,
but
we
have
to
make
sure
that
we
have.
B
This
is
a
problem.
We
broke.
Existing
functionality
function,
I
the
only
existed
for
a
week
grant
I
granted,
but
it
was
still
working
functionality
that
we
broke
and
that's
that
points
toward.
We
need
to
be
more
careful
about
testing
and
then
the
problem
that
rotate
is
a
little
hard
to
write,
good
unit
test
for
that,
and
so
it's
easy
to
miss
this
kind
of
stuff.
But
that
means
that
we
have
to
be
we're.
B
Certainly
aware
of
that
we
knew
rotate
was
broke
and
we
knew
how
was
broken
when
you
is
fixed
and
then
got
rear
oaken.
So
we
have
to
be
more
careful
about
catching
those
cases,
and
so,
when
someone
suggests
a
change
to
rotate,
it
would
have
been
nice.
If
you
know,
one
of
us,
including
me,
had
suggested
a,
but
we
check
it
against
the
earlier
PR.
Those
test
cases
that
already
lined
out
and
demonstrated
to
you
know
how
it
worked.
B
It's
good
to
go
back
and
just
to
make
sure
everything
still
works,
I,
don't
know
whether
there's
some
way
you
know
like.
I
said
it's
hard
to
do
a
good
unit
test
on
this,
because
it's
hard
to
compare
what
what
you
mean
when
you
do
a
rotate,
but
certainly
if
we
had
some
kind
of
gallery,
I
guess
like
the
kind
of
gallery
are
talking
about,
maybe
like
just
some.
You
know
like
a
human
could
just
look
like.
Oh
okay,
so
are
like
examples.
B
Oop
that
example
is
no
longer
appear
in
the
way
it
should,
or
something
like
that,
something
that
we
could
see
a
problem
and
part
of
my
concern
about
that
or
I.
Think
part
of
the
issue
I
have
with
the
release
and
in
Stuart
I
know,
I
I,
certainly
I'm
fully
behind
releasing
frequently
and
everything
like
that,
but
yeah
the
gap
between
when
the
pr
was
merged
into
the
master
branch
and
then
the
release.
I
don't
remember
how
long
the
gap
was,
but
imagine
it
was
not
very
long.
We
were
talking.
B
So
it
it
didn't,
have
the
chance
for
things
to
kind
of,
settle
or
check
whether
another
PR
conflicted
with
it
either.
You
know
I
at
PRS,
in
there
for
draw
a
grid
and
draw
a
limb
that
were
merged
into
the
contraction.
There's,
no
reason
that
should
have
broken,
but
you
know
there
is
no
opportunity
really
test
everything
in
mass
right.
B
Fixed
okay,
I,
didn't
mean
break
like
that.
That's
a
that's
emerging
is,
of
course
an
issue,
but
I
just
mean,
like
you
know,
non
subtle
interactions
between
a
code
after
it's
been
put
all
together
in
the
branch
to
be
released.
It
would
have
been
nice
if
there
have
been
a
couple
days,
for
you
know,
people
major
developers
to
go
and
say:
okay,
this
is
the
branch
is
going
to
be
released.
No
more
features
are
going
to
go
into.
B
It
will
do
any
fixes
that
need
to
be
necessary,
but
you
have
to
look
for
look
for
anything
outstanding
now.
Even
that
may
not
have
caught
this
kind
of
problem,
because
if
I
wasn't
paying
attention
and
other
people
didn't
think
of
it
either,
then
we
might
not
have
run
by
these
test
cases.
But
the
fact
of
these
test
cases
were
out
there
is
that,
because
a
we
were
aware
that
rotate
could
break,
and
none
of
us
all
of
us
kind
of
fail
to
just
check
those
test
cases
and
identify
this
problem.
C
I
totally
agree
with
you:
Albert
I,
think
I
think
we're
pushing
for
releases
I
think
we're
being
a
little
bit
too
aggressive,
and
maybe
the
key
is
to
like
the
pr
being
accepted
the
time
between
a
new
feature
being
accepted
as
a
PR,
and
when
we
release
should
be
like
a
minimum
amount
of
time
like
a
month
or
something
just
so
that
everyone
it's
in
master
everyone's
like
used
it
for
a
little
while
and
so
that
we
don't
impede
progress.
We
could
create
a
new
branch
called.
C
A
B
To
this
day,
I
Greer,
that
I
mean
that
was
the
whole
scheme
of
how
we
do
our
branches,
and
so
we
that's.
What
we
should
have
done.
Mastership
in
put
to
0.5
0.5
would
be
relatively
locked
and
then
people
all
the
developers
ago,
deliberately
pull
that
branch,
try
out
the
0.5
branch,
and
then
people
can
still
continue
to
work
on
the
main
master.
B
Do
new
features
do
what
a
hell
they
want,
but
0.5
is
locked
at
that
point
and
then
once
we're
comfortable,
as
your
point,
five
looks
stable,
I,
don't
think
it
needs
to
be
necessarily
be
a
month.
Of
course,
I
I'm,
a
little
more
I
want
to
push
for
releases.
I.
Don't
want
us
to
get
too
conservative
on
that,
but
at
least
a
week,
maybe
or
something
like
that,
where
05
everyone's
had
an
opportunity,
we've
had
a
death
meeting
in
between
to
say:
okay
guys,
we
really
gonna
check
this
out,
and
then
we
place
it.
B
A
B
C
A
Okay,
the
thing
is,
we
need
to
be
more
okay,
I,
agree,
that's
the
way
it
should
work
it's.
Why
should
it
work?
I?
Think
we
need
to
be
more.
Oh
I,
don't
know.
Maybe
I
need
to
stay
in
to
the
office
to
ihop
update
just
to
get
the
feature
freeze
done
tonight.
Yes,
we
seem
to
be
things,
don't
seem
to
move
like
that.
A
C
A
A
C
E
A
C
C
B
A
Will
have
it
patched
well,
I'm
going
to
be
in
the
same
place
is
drew
on
Wednesday
for
a
week
and
a
half.
So
at
that
point
it
will
get
fixed
if
it's
not
that
by
thursday
morning.
Uk
time
I
was
stunned,
so
okay,
okay
and
we
need,
since
we
need
to
add
some
math
tests,
because
to
the
rotate,
because
it
would.
A
B
A
A
B
E
C
Yeah
I
would
also
suggest
that
I
think
we've
all
been
fairly
bad
at
looking
at
the
PRS
that
Stewart
flags
is
for
review.
I
would
I
would
encourage
everyone
to
you
know
every
once
in
a
while,
like
you
go
and
look
in
to
come
and
give
your
thumbs
up,
because
it
seems
like
very
few
people
are
doing
that
which
we.
C
B
Yeah,
so
we
have
to
be
a
little
careful.
You
know.
Sometimes
it's
easy
to
appraise
with
our
code
change.
You
know
you
don't
actually
pull
the
branch
and
you
can
approve
it,
but
something's
extensive,
the
aia,
prep,
and
that
was
the
reason
why
I
kept
putting
off
actually
giving
my
approval,
because
I
was
like
I
have
to
test
the
branch,
but
I
don't
have
time
and
right
I
mean
it's
what
it's
great
if
people
approve
stuff,
but
sometimes
you
really
need
to
test
it
out
to
back
to
be
a
pain
in
the
ass
kaput,
yeah.
A
C
The
gallery
could
be
used
in
that
way
in
that
going
back
to
whatever
was
saying
that
really,
you
need
to
like
look
at
it
and
see
what
it
doing
and
you
could
people
could
write
galleries
which
you
know
if
things
break.
It
would
be
clear
in
the
gallery
right
like
that
rotate
function
if
it
becomes
all
black
and
you're.
Looking
through
the
gallery,
its
genetic,
like
that,
and
you
know
something
all
right-
the
for
the.
C
F
C
C
A
C
C
A
When
the
when
the
gallery
runs,
he
adds
a
matter
that
it
adds
the
metadata
as
a
version
number
to
the
whole
gallery
metadata.
So
there's
a
deploy
script
in
the
gallery
repo.
So
when
you
run
you
run
that
deploy
one
someone
who
has
Bush
rights
to
the
upstream
branch
runs
the
deploy
script
and
the
disappointed
voice
who
deletes
the
github
pages
branch,
creates
a
new
morph.
Engages
branch
builds.
A
A
C
A
A
But
that's
fine
at
the
moment,
we've
got
three
things,
but
when
we've
got
50
and
word
that
becomes
rapidly
impractical,
so
maybe
we
have
to
think
about
that
as
well.
Well,.
A
A
C
A
C
A
E
C
Well,
at
least
it
would
be
clear,
at
least
you
wouldn't
have
to
go
and
open
the
notebook
and
then
look
at
it.
You
would
actually
just
see
like
like
pretty
much.
You
shouldn't
give
you
much
of
an
output
at
all,
maybe
a
few
plots
and
stuff,
but
you
would
see
oh
there's
a
big
red
thing
that
says
oh
and
something
failed
and
it
stops
at
that
point
right
yeah
I
mean
the
gallery
is
great
and
I.
Think
it's
very
powerful
and
I
think
we
can
use
it
for
a
number
of
things.
C
E
E
A
C
A
A
C
C
G
A
C
F
A
So
knowing
this
works
is
that
there's
a
git
repository
it
contains
each
example,
has
its
own
folder
inside
the
government
directory.
So
you
add
a
directory
in
you,
add
your
notebook
into
that
directory
and
you
have
to
add
JSON
metadata
to
your
notebook
when
you
PR
it,
which
gives
your
author
name
the
date.
You
voted
a
title
for
it,
which
is
generating
links
and
a
category.
A
Then
there's
a
better
prepared
destroyed
as
I
prepare
on
the
store
deployed
action
script
in
the
base
directory
of
the
Rijo.
You
can
run
that
by
doing
Python
prepared
to
boy
space
run
space
convert
which
will
execute
all
the
notebooks
and
convert
the
Malta
HTML
and
build
the
HTML
site
in
a
directory
called
HTML.
You
can
then
open
up
the
index
page
in
firefox
or
whatever
browser
you
use
from
that
HTML.
So
I
can
preview
the
site.
Al
there's,
also
a
published
whovian.
A
The
public
equals
true
and
metadata
flag
as
well,
which
you
can
set
published
a
force,
which
means
you
are
when
you
buy
locally,
but
won't
build.
If
you
run
the
main
to
floods
with
and
and
then,
if
you
run,
run,
prepare
deploy
and
you
can
see
the
site,
then
if
one
of
the
main
taneous
runs
the
vast
ripped
which
is
called
deployed,
it
will
do
why
I
was
describing
earlier,
where
it
pushes
through
bill
result
to
get
our
pages.
F
C
A
A
A
C
Yeah
I
think
I
think
it
might
be
a
good
idea
Stewart
to
start
now
that
we
have
both
the
web
page
and
galleries,
which
are
kind
of
their
own
thing.
I
think
it
might
be
a
good
idea
to
put
somebody
some
people
in
charge
of
those
two
things
so
that
things
don't
get
forgotten
or
lost,
because
it's
sort
of
a
lot
to
do
both
to
handle
the
code
and
then
the
galleries
as
well,
which
you
have
to
watch
another
repo
and
the
the
website,
there's
a
lot
to
keep
track
of.
C
So
it
might
make
sense
to
kind
of
have
people
whose
responsibility
is
to
essentially
maintain
that
you
know
the
gallery.
Person
would
be
nice
because
it
would
be
somebody
that
would
go
around
and
encourage
people
to
actually
make
notebooks,
even
if
they
weren't
super
good
at,
like
you
know,
improving
the
gallery
code
itself
and
everything
just
a
thought.
C
C
G
A
A
Mean
for
the
moment,
I'm
happy
to
look
after
it
unless
Yvonne
what's
to
say
it
on
in
that
it's
not
much
work
at
the
moment
to
maintain
it
and
really
pr's
on
it.
But
if
someone
wants
to
tell
you
under
the
wing
right
CSS
for
and
genuinely
dog
after
it,
then
you'll
grow
faster
than
it
will
live.
Someone
oh
I
mean
even
if
you
don't
want
to
volunteer
to
do
it
in
an
official
capacity
or
this
or
semi
official
capacity.
Just
saying
yet
I'll
write
some
code
Oriel
at
some
point
would
be
good,
but
web.
C
C
H
C
Yeah,
remember
that
I
put
together
three
notebooks
for
the
rescue
workshop,
so
those
are
good
resources,
ones
on
numpy
ones,
on
syfy.
Thank
your
side
play
no
Matt
pot,
lid,
sorry,
nope
I,
another
one's
on
my
pot,
lid
and
plotting,
and
another
one's
on
some
pie.
They're
currently
hosted
on.
Well.
If
you
go
to
the
blog,
you
can
get
links
to
those
notebooks.
My.
H
C
Yeah
the
reason
why
this
came
up
was
my
notebook
is
basically
like
an
overview
of
all
kinds
of
stuff.
So
I
had
just
decided
to
add
a
category
called
overview,
which
is
I,
don't
know
we
made
yeah.
I
agree
with
Stewart.
We
want
some
more
so
fiercest
set
of
categories.
You
could
see
the
categories
in
as
they're
displayed
right
in
the
gallery
is
like
curves
and
there's
a
code
like
a
map.
There's
two
notebooks
in
there.
A
A
A
Original
naming
and
oh
look,
he
go
like
the
way
that
I
sort
of
started
going,
which
was
have
a
category
for
me
section
of
some
by
but
habit.
Maybe
you
also
have
an
introductory
/
overview
kind
of
section
and
then
maybe
a
we
could
have
like
an
introductory
overview.
Section
have
like
0
respect
for
a
net
section,
probably
the
most
things
right,
and
then
we
can
have
a
sort
of
I
can't
that
I
can't
think
of
a
name
for
like
doing
science.
C
Would
suggest
that
we
create
an
issue
for
this
and
just
have
the
conversation
there
where
people
can
post
and
say
what
they
think
that
way,
it's
all
maintained
the
conversations
maintained
and
you
can
bring
it
all
together.
I
mean
that
one
one
other
possibility
is
that
we
have
no
categories
of
the
web
at
all.
Going
like
the
matplotlib
gallery,
which
is
really
you
know,
they
just
provide
lots
of
keywords
and
things
and
just
kind
of
browse
and
look
at
look
around
to
see
what
you
want.
A
C
H
E
C
C
Yeah,
so
basically,
I
was
able
to
redo
what
you
did
using
a
PI
FM,
which
is
Python
version
of
ephemeris
pi
FM
is
not
super
great.
It's
not
it's.
Basically,
wrappers
around
see
functionality,
that's
written
for
xfm,
but
it's
all
open
source
and
available.
Somebody
wrote
a
script
around
that.
So
the
idea
for
this.
C
Which
is
has
a
higher
accuracy
than
what
we've
been
using
so
far
right
and
up
high
so,
but
it
turns
out
that
pie
FM
provides
that
similar
functionality
and
lots
of
other
stuff.
So
I
wanted
to
see
what
people
thought
about
having
an
optional
dependency,
and
this
is
way
off
that
we
don't
have
to
figure
this
out
now,
but
I
think
in
the
future.
It
would
be
nice
to
have
pie
FM
in
there
for
lots
of
other
reasons
as
well
again,.
C
But
as
we
kind
of
get
closer
to
Esther
pie,
it
would
be
nice
to,
for
example,
like
here's
one
you
use
case
if
you're
looking
at
last
coat
was
it
last
go
c.
3
is
a
wide
angle,
one
right
yeah
and
you
would,
for
example,
you
would
do
a
plot
and
it
would
know
what
hi
FM
would
automatically
know
what
planets
you
have
in
the
field
of
you
and
would
label
them
automatically.
For
example,.
E
E
C
A
G
H
Yahoo
remove
incomplete
they
don't
get
in
the
other
function
from
a
DPL
also
provides,
and
in
any
case
there's
been
a
couple
of
bug.
Fixes
since
then
Stan
night
to
be
that
kind
of
threw
the
Sun
guy
code.
Is
that
connect
then
slice
that
boob
also
probably
just
need
to
be
done.
The
whole
thing
again,
yeah.
C
A
C
A
E
A
Yeah
I
mean
go
it
as
long
as
you
can
as
long
as
the
ephemeris
package
can
provide
a
coordinate
not
in
a
projective
system
which,
on
a
are
a
debt,
is
but
like.
If
you
can
provide
a
non
projected
coordinate
system
of
any
sort.
In
theory,
it
can
be
arranged
so
that
a
trip
I
coordinates
can
convert
that
into
something
that
we
want
to
use.
A
C
C
A
C
G
H
A
A
E
All
right,
no,
the
job
for
the
exit
of
the
five
thanks
people.