►
From YouTube: 2020-06-22 Crossplane Community Meeting
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
Okay,
the
recording
has
started-
and
this
is
the
June
22nd
2020
crossplane
community
meeting.
So
the
first
section
here
is
on
our
releases
and
our
milestones.
So
we
did
a
0.12
release
earlier
this
morning
that
Dan
ran
and
got
out
four
of
us
note
that
that
release
was
strictly
for
the
core
cross,
flane
repository
so
din
and
Nick.
Can
you
give
me
a
quick
update
to
catch
me
up
on
what
the
plan
might
be
for
if
any
beliefs
is
for
providers,
if
there
just
aren't
updates
that
are
needed
for
this
one
yeah.
B
So
it
kind
of
depends
on
the
provider
which,
in
our
in
our
release,
documentation,
it
talks
a
little
bit
about
how
we
don't
have
to
be
coupled
to
cross
plain
releases
for
the
providers
or
any
other
repos.
Really
that
being
said
in
the
past,
we
typically
have
especially
obviously,
if
there's
some
sort
of
breaking
changes
in
crossplane
that
require
that,
as
we
are
pre
1.0,
so
we
do
have
breaking
changes
from
time
to
time.
This
time
we
did
not
have
any
breaking
changes.
B
We
also
had
a
number
of
PRS
landing,
for
instance,
in
provider
AWS
for
new
types
that
we
wanted
to
get
into
the
next
release,
move
off
as
we'll
talk
about
later
on
added
pub/sub
support
in
GCP,
which
that
has
now
been
merged,
but
all
those
things
were
kind
of
landing
very
recently
or
are
about
to
land
and
then
in
something
like
provide
our
or
Alibaba
I.
Don't
think
we
have
any
new
updates
there,
so
they
don't
really
necessitate
a
release,
so
I'd
imagine
releases
for
AWS.
A
Yeah,
that's
totally
reasonable.
That
is
definite
part
of
our
philosophy,
our
release
philosophy,
that
providers
can
be
released,
a
tab
and
a
sink
along
their
own
timelines,
etc.
It's
been
convenient
to
do
all
of
them
at
once,
along
like
with
core
crossplane,
but
it's
not
a
strict
requirement
whatsoever.
So
yeah
we
can.
You
can
run
those
when
appropriate
when
things
settle
down
or
with
the
you
know,
critical
mass
features
as
they
ended
for
those
say
yeah
we
can.
We
can
follow
up
on
that
f10.
B
Would
be
fine
with
me
AWS,
we
have
a
couple
of
things
in
flight
I.
Think
the
big
thing
that
needs
to
land
is
the
node
groups
PR
for
eks
and
the
other
ones
could
kind
of
flow
in
whenever
I'd
be
fine
with
end
of
the
week
releases
or
just
some
time
this
week
releases
very
either
one
of
those,
but
if
anyone
has
dissenting
opinions
to
that,
I'd
be
open
to
till
later
earlier,
if,
if
needed,
what's.
C
A
C
A
D
C
B
B
A
Yeah
and
I
definitely
think
that
it's
they
choose,
they
think
if
you
just
rolled
things
through
very
easily,
then
that
kind
of
keeps
Croft
accumulating.
But
if
you
have
to
be
very
explicit
about
the
the
ADI
items
and
issues
you
want
to
tackle
in
the
next
release
and
explicitly
add
them,
then
it
makes
it
much
more
crisp
and
clear
about
what
the
goals
are.
Gonna
be
so
yeah
I
like
that.
A
It
will
drop
a
link
in
here
when
that
is
done.
We
could
give
a
couple
of
overall
updates
for
the
group
here
about
basically
the
things
that
went
into
0.12,
but
also
some
other
outstanding
work
that
is
going
into
upcoming
milestones
as
well.
So
Nick
in
case
you
guys,
are
both
on
the
calls.
You
want
to
give
an
update
about
efforts
around
expanding
the
breadth
of
coverage
for
writers
in
their
services
in
Crossman
organization
or
ecosystem.
C
Sure
you
know
try
to
be
very
brief
and
then
hand
over
to
Casey,
as
it
says
who
we
are
investigating
code
generation
at
the
moment
as
a
way
to
speed
up
writing.
Cross-Play
providers.
We're
looking
at
two
fronts
for
this
one
front
is
that
we
are
doing
some
Rd
into
whether
it
would
be
possible
to
reuse
some
of
the
existing
code
for
terraform
providers
and
use
those
with
cross-play,
and
the
other
front
is
that
we
are
looking
at
a
more
sort
of
Michael
like
a
native
code
generation
approach.
C
Either
way
we
do
expect
that
there
is
going
to
be
some
amount
of
code
that
is
not
fully
generated.
That
needs
to
be
filled
in
at
the
end,
so
we're
going
to
look
at
whether
we
end
up
doing
both
tear
fall
and
hand
generation
or
one
or
the
other,
depending
on
which
most
promising
and
meets
our
needs.
There's
no
documents
or
anything
like
that
out
there
that
I'm
aware
of
that
attracting
this
effort.
C
At
the
moment,
it's
still
very
early
Rd
from
from
my
end,
I
have
been
looking
at
compiling
in
terraform
code
to
a
crossplane
provider
which
is
which
is
not
recommended
by
the
tariff
or
maintainer.
So
probably
not
the
direction
will
actually
go,
but
we're
just
looking
at
it
for
experimentation
purposes
and
I
have
an
AWS
I
enrolled
can
read,
read,
read
and
observe.
A
E
So
the
default,
an
enforce
composition,
references
are
implemented.
So
now
you
can
choose
a
default
position
to
be
selected
for
for
your
infrastructure
definition
and
also
you
can
enforce
a
composition
for
your
type
so
that
no
I
can
override
it
and
they
they
want
me
to
you,
know,
provide
any
composition,
selector
or
anything,
and
the
third
up
it
was
disappear.
Was
that
in
composition,
when
you
have,
you
know,
leaves
of
resources,
they
all
got.
E
You
know
as
a
fix
of
six
characters,
you
know
and
five
random
check
characters,
and
this
could
get
you
know
like
really
long,
and
once
you
get
like
a
tall
tree
APR
that
address
that
made
it
like.
You
know,
we
only
have
two
suffixes
in
total,
like
12
characters
added
to
the
actual
name
of
the
requirement,
and
that's
pretty
much
it
like
it
doesn't
suffix
continuously
at
each
level.
It
just
has
at
most
two
different,
so
fixed
to
make
sure
it's
unique
with.
C
Lovak,
did
you
get
a
chance
to
play
the
docs
with
the
details
about
the
default
enforce
compositions
before
the
release
yeah.
E
A
B
References
on
the
package
to
those
CR
DS
that
get
created
so
we're
basically
going
to
be
doing
a
similar
thing,
with
M,
structured
definitions
and
publications
and
also
with
some
of
the
permissioning.
That
happens,
we'll
need
to
kind
of
follow
some
of
that
creation
to
the
CR
DS
that
are
produced
by
creating
those
composition,
pipes
and
in
making
sure
the
right
permissions
are
added
there
and
the
right
labels
and
that
sort
of
thing,
but
we're
basically
just
adding
support
there
and
then
the
other
goal
of
this
is
being
able
to
support
update
of
packages.
B
B
So
we
don't
want
to
have
that
on
our
release,
notes
anymore
essentially,
and
we
want
you
to
be
able
to
upgrade
your
provider
successfully.
Obviously,
if
there
are
breaking
changes,
we
can't
always
guarantee
that
every
upgrade
is
going
to
be
possible
without
doing
some
manual
intervention,
but
we
do
want
to
make
that
process
a
lot
easier,
so
be
watching
for
a
one-pager
on
that
and
also
feel
free
to
weigh
in
and
say
you
know
this
fits
our
use
case.
It
doesn't
that
sort
of
thing.
A
Yeah
that
the
support
for
upgrade
of
packages
in
their
contents
in
series
etc
has
been
a
a
long
time
coming
some
better
making
some
progress
on
that,
for
you
know
as
a
step
towards
reliability
in
your
readiness.
So
that's
great
and
then,
if
you're
going
to
do
a
quick
update
on
any
open
application
model,
support
and
improvements
that
were
included
in
0.12
timeframe.
B
So
you
know
there
wasn't
a
new
release
of
the
any
of
the
add-ons
or
anything
like
that.
I
know,
there's
been
a
lot
of
work
going
on
there
and
we
have
a
couple
folks
on
the
call
they've
been
working
on
that,
so
I
would
defer
to
them
too.
To
talk
about
that
I
know.
There's
been
some
some
things
there
that
I
haven't
gotten
to
review
recently.
So
if
anyone
wants
to
jump
in
and
mention
anything,
that'd
be
great.
D
Yeah
I
think
like
at
a
high
level
just
trying
to
get
like
the
core
of
across
playing
down
to
the
minimum
footprint
and
then,
like,
basically,
you
know
getting
home
into
its
own
like
stand-alone
home
charts
that
can
be.
You
know
like
installed
alongside
the
the
crossplane
core
stuff,
but
then,
as
far
as
the
ohm
content,
that's
going
to
be
going
into
the
next
couple
of
releases.
Does
anybody
else
want
to
share
with
what
they
have
planned
for
that
I've?
Seen
a
lot
of
activity
going
on
so
I?
Don't
know
if
you
know
hello.
F
C
Also
seen
some
Piazza
gonna
be
doing
things
like
some
more.
The
design
face
proposing
the
traits
may
be
words,
but
all
these
traits
have
a
field
that
declares
whether
they,
whether
the
traits
ER,
is
aware
of
OEM
compatible
with
how
I
am
or
whether
it's
something
that
is
being
used
with
I
am,
it
doesn't
know,
is
designed
for
OEM.
If
that
makes
sense
and
I
believe
there
is
a
whole
bunch
of
work
going
on
as
well
for
component
to
revisions.
So
basically
a
roll
forward
and
roll
back
a
story
for
our
opponents.
C
Just
to
the
point
that
Phil
mentioned
there,
which
I
don't
I,
think
that's
the
first
time
we
mentioned
that
we
we
do
need
to
talk
about
it
with
the
OEM
community
folks
at
the
only
meeting.
But
we
are
thinking
that
it
might
be
nice
to
move
the
OEM
app
config
controllers
out
of
cross
lake
or
and
into
a
single
installable
binary.
Maybe
the
only
local
or
something
like
that,
but
we
need
to
talk
to
the
rest
of
the
OEM
folks
to
figure
out
where
the
best
place
for
that
alter
go
is
yeah.
F
I
actually
think
it's
a
good
idea,
so
I
actually
proposed
before
that.
We
may
want
to
make
CrossFit
more
mobilize
the
ball,
so
it's
more
either
for
people
to
install
what
they
want.
Like
stepper
controllers,
they
don't
have
helm,
install,
for
example,
yeah
Oh
across
time,
Cole
even
without
calculator,
and
they
can
try
out
om
functionality,
spy.
C
Yeah,
let's,
let's
figure
out
some
time
to
come
up
with
a
plan
on
this
and
maybe
write
it
down.
I
will
say
that
not
sure
what
the
timeline
is,
but
we
are
leaning
toward
removing,
rather
than
updating
with
the
kubernetes
application
functionality,
which
means
our
OEM
remote,
add
on
one
exists
anymore
point
we
needed
which
basically
we're
banging
moving
pots
under
the
cross
plane
of
OM.
C
It
would
be
the
OEM
runtime
and
the
only
local
add-on
custom,
removing
and
cousin
add-ons,
because
they
never
really
made
it
for
the
the
the
OEM
thinks
that
we
could
maybe
rename
these
things
that
we
might
want
to
merge
those
two
repos
into
one
that
could
be
binary
had
libraries
as
well.
Oh,
we
can
leave
them,
as
is
it
looks
one
but
pretty
flexible
now
as
to
how
we
do
that
I
think
they've
been
realized
as
we
want,
all
of
that
go
and
stuff
to
just
be
installable
as
a
single
binary.
Yes,.
F
Yes-
and
you
can
discuss
about
the
naming
or-
and
the
repo
names
think
so
they
are
generally
open
for
that.
So
let's
keep
I
think
we
can
raise
the
issue
on
the
github
and
we
can
continue.
Discussion
sounds
good,
yeah,
I,
also
post
a
few
links
that
we
are
recently
working
on.
So
the
first
one
is
the
animation
in
the
relation
of
the
components
he
owned
and
it's
generally
ready
for
review.
So
if
someone
someone
can
take
a
take,
a
look
at
it
on
that
PR
and
it
will
be
very
appreciated.
F
A
second
one
is
another
idea
that
we
are
recently
Conkling
on
it's
about
something
like
decomposition,
this
more
like
the
reverse
process
of
the
existing
combination,
which
which
also
have
a
similar
purpose.
Then
we
want
to
create
higher
level
abstractions.
So
if
anyone
interesting
this
idea,
we
are
very
happy
to
talk
about
that,
and
maybe
you
know
future
we
can
have
a
similar
in
annotating
the
course
encore,
which
can
not
only
provide
you
with
the
way
to
do
composition,
but
also
decomposition.
F
So
you
can
actually
create
smaller,
see,
artists
from
given
given
CRV,
for
example,
today
in
Ottawa,
but
we
have
open
crews,
which
is
very
huge.
They
are
including
everything
instead
of
on
CRT,
so
we
are
thinking
about.
Maybe
we
can
use
the
decomposition
to
split
them
into
smaller
pieces,
including
a
simple
workload
and
the
several
roles
chantry's
based
on
that.
So
it's
it's
it's
a
very
simple
idea,
but
very
similar
to
a
combination,
but
we
would
like
to
explore
it.
It's
very
for
402,
higher
level
abstraction
and
talk
of
kubernetes.
C
F
A
All
right
so
yeah,
we
will
then
preserved
a
third
team,
as
we
mitching
will
open
up
a
project
forward
to
start
tracking
issues
for
the
across
main
ecosystem
as
a
whole
for
0.13
time
frame,
and
we
will
open
up
a
poll
request
to
update
the
roadmap
with
latest
desired
items.
Features,
and
you
know,
efforts
for
that
time
for
you
as
well.
A
So
that's
the
milestone,
release
or
the
section
of
the
meeting
here.
So
we
can
proceed
on
to
the
community
topics.
The
first
one
I
wanted
to
share
is
that
we
have
gotten
all
the
post-processing
and
publishing
done
and
cleaning
up
of
all
the
recordings
for
the
community
a
event,
so
the
cross
plane
the
day
stuff.
All
of
the
every
single
talk.
A
A
I,
don't
know
how
Google
came
up
with
that,
as
the
is
the
suggested
name
for
that
link.
That
is
really
weird
interesting,
but
anyway,
so
we
have
an
entire
playlist
here.
The
cross
playing
Community
Day
links
here,
so
the
keynotes
and
you
know
stuff
about
own
with
burner
burns
and
the
rest
of
everybody
who
is
contributing
to
that
all
the
other
cool
demos
rest
of
the
day,
so
those
are
all
up
on
the
cross,
plain
YouTube
channel
and
ready
for
people
to
watch
at
your
leisure.
A
B
For
sure
so,
move
off
it
came
on
and
basically
in
an
hour
and
a
half
implemented
cloud
pub/sub
for
your
provider
GCP,
so
that
was
super
cool
I'm,
starting
with
literally
like
opening
an
issue
and
then
assigning
and
implementing
it
and
opening
a
PR
and
that's
actually
been
merged.
Now
we
had
some
some
good
participation
in
terms
of
viewership
for
that,
so
that
was
great
and
we
think
we
can
use
it
as
something
to
point
to
in
the
future.
A
D
B
A
B
Sure
I
can
go
and
give
it
I,
don't
know
all
the
details
for
it.
But
if
you
are
familiar
with
pop,
who
is
the
field
CTO
over
at
Cystic
he's
gonna
be
joining
me?
He
has
his
own
podcast,
so
he's
very
familiar
with
live
streams
and
that
sort
of
thing,
but
we're
not
exactly
sure
what
we're
gonna
be
doing
yet,
but
he's
gonna
be
on
and
he's
he's
definitely
a
fun
person
to
talk
to
and
has
a
lot
of
technical
expertise,
so
I'm
sure
it'll
be
fun.
Nice.
A
B
Think
it'll
be
focused
pretty
heavily
on
what
Cystic
does
so
we're
talking
about
how
we
want
to
incorporate
that
with
crossplane.
It's
mostly
doing
so,
as
you
might
imagine,
from
their
name
doing
some
like
lower
level
analysis
of
what
like
sis,
calls
are
being
made
and
that
sort
of
thing
which
is
sort
of
related
to
kubernetes
controllers.
Although
you
know
that
there's
not
that
much
analysis,
you
do
on
something
like
our
providers,
but
we
may
talk
about.
A
B
A
Don't
think
I've
missed
at
CBS,
yet
so
I'll
be
up
for
that
one
as
well.
I
will
give
a
just
a
brief
update
on
CN
CF
stuff.
There
I
don't
think,
there's
a
huge
any
big
progress
upstream
with
a
proposal
that's
been
opened,
but
one
thing
to
note
is
that
we
did
take
the
time
to
go
in
update
the
owners
code
files
and
the
maintainer
xand
permissions
and
all
that
stuff
for
every
repo
in
the
crossplane
organization.
A
So
to
my
knowledge,
every
repo
that
crossplane
has
has
been
updated
with
the
maintainer
for
that
repository,
and
they
all
have
permissions
to.
You
know:
merge
pull
request
to
maintain,
administer
the
the
repositories
as
well.
So
if
anyone
has
any
issues
with
access,
they're
they're
expected
to
have
please
let
me
know
reach
out
on
slack
or
whatever
is
easiest
for
you
and
I'll
take
a
look
at
those
permissions,
but
I
went
through
them
a
couple
times.
I
think
that
they
are
up-to-date
and
accurate.
That
did
cause
a
little
regression
on
our
Docs.
A
Sorry,
the
cross
mine
at
I/o
website
repo.
When
I
was
messing
with
the
permissions
there,
I
inadvertently
messed
up,
the
bots
account
that
we
have
that
automatically
publishes
contents
from
crossplane
repo
to
the
docs
website.
Repository
so
Dan
pointed
that
out
during
the
released
this
morning,
and
we
got
that
fixed
up,
so
that
should
be
all
cleaned
up
now.
But
if
there's
any
lingering
permissions
issues,
then
please
please.
Let
me
know
I.
F
F
F
Yes
in
guilty,
meaning
list
and
I
am
prettier
than
before.
I
asked
the
Kiowa
say
they
want
to.
They
claim
that
the
existing
project
should
go
through
the
old
process,
but
now
they
I
think
some
of
them
exchange
idea.
I
think
it's
leads,
because
it's
just
that
they
are
a
bunch
of
project.
This
show
goes
to
the
new
process.
I
see
that
coal
plant
part
of
it.
F
A
Yeah
thanks
for
that
heads
up,
Harry
I
just
took
a
note
on
my
own
to-do
list
here
to
follow
up
and
get
in
get
come
from
it
or
find
that
and
get
operation
of
that,
and
then,
if
we
need
to,
you
know,
follow
a
new
process
or
you
know
whatever
the
appropriate.
Next
steps
are
we'll
we'll
do
that.
That's
you
know.
That's
okay!
Thank.
C
A
You
should
thank
you
for
pointing
that
out:
Harry
mm-hmm
cool
okay
and
then
that
was
the
community
section
there.
So
then
we
have
a
couple
of
total
requests
that
have
been
flagged
for
discussion
today,
so
we'll
bring
up
the
first
one
here:
move
offic!
Oh
sorry,
yeah
those
go
off!
You
wanted
to
discuss
215
here,
yeah.
E
Yeah
so
Alan
is
owning
this
piace
we've
got
a
discussion
with
him
ed,
so
there
are
like
two
main
resources
right
now
hosted
zone
and
resource
record
said,
but
I'm
kind
of
on
the
fence
about
like
treating
research
I
could
set
a
separate
managed
resource
or
not.
So
if
you
go
like
the
bottom,
I
have
left
a
few
few
comments
about.
Like
you
know,
current
status
and
like
you
know
how
the
AP
is
shaped
so
that
people
can
get
an
idea
like
you
know,
without
digging
into
too
much
detail
in
the
AWS
SDK.
C
See
the
tire
skimming
your
comment
here:
I
see
that
you've
compared
it
to
what
platform
ition
in
terraform
do
to
model
this
stuff
and
I
know
that
we
commonly
face
this
where,
especially
in
AWS
and
older
AWS
AP,
is
they
exactly
fit
into
the
declarative
patterns
that
we
want?
We
have
to
kind
of
them
a
little
bit
and
often
what
we
do
is
go
and
look
at
how
other
projects
who
have
had
to
do
this
before
I
have
interpreted
them.
Does
it
is
the
patent
that
we're
proposing?
The
same
is
what
cloud
formation
and
terraform.
E
E
C
C
But
those
resources
might
not
always
be
sort
of
our
idea
of
the
ideal
implementation
of
those
results
in
the
crossplane
world
regards
to
api
shapes
and
babies,
and
things
like
that,
so
we
are
considering
sort
of
having
basically
two
providers,
one
that's
generated
from
terraform
and
one
that's
some
handwritten
to
have
opinions
that
we
feel
better
about.
C
So
if,
if
we
didn't
really
feel
that
this,
this
API
was
much
better,
when
everything
is
all
one
resource,
then
we
could
diverge
there.
I'll
go
have
a
look
at
this
this
afterwards.
In
my
experience,
usually
if
other
systems
like
terraformer
CloudFormation
for
model
plays
as
separate
things,
there's
a
reason
for
it
usually
in
subsidy
was
like
you,
don't
always
want
to
create
them
all.
As
one
results
or
not
but
I
can't
I
have
looked
at
this
one.
So
hopeful
happy
to
have
looked
out
of
bed.
B
Yeah
I
I
also
linked,
underneath
this
PR
and
the
agenda
doc
to
route
table,
which
we
just
merged
a
PR
to
fix
an
issue
with
the
client
library,
but
that
also
kind
of
brought
up
that
it
was
kind
of
taking
the
place
of
multiple
resource
types
which
are
very
closely
associated
with
Ralph's
table,
but
they
do
expose
separate
API
s
and
you'll
see
a
similar
comment.
The
move
off
ik
has
here,
to
the
other
one
I'd
say
I'm,
usually
pretty
in
favor
of
making
as
granular
as
possible,
move
off.
B
They
brought
up
a
good
point
that
like
well,
do
we
want
to
manage
like
tags
as
their
own
resource,
but
that
seems
like
a
pretty
like,
like
I,
guess,
sort
of
like
obvious
thing
that
we
want
to
be
included
with
it,
but
I
mean
I.
Guess
you
could
argue
that
for
any
of
the
others
as
well,
but
generally
I
actually
think
it's
simpler
to
implement
when
we
get
more
granular,
even
though
it's
more
controllers
you're
the
lifecycle
of
the
things.
Basically,
what
you're
putting
into
the
crud
methods
is
a
lot
more
straightforward.
B
In
my
opinion,
you
know
if
you're
just
creating
a
route
and
you're
managing
the
route,
then
you
know
you
just
call
the
create
method
for
that
and
the
delete
method
for
that,
rather
than
in
the
route
table
having
to
iterate
through
the
routes
and
kind
of
like
try
and
manage
them
and
and
understand
changes
to
it.
So
I'm
I'm
definitely
pretty
strongly
in
favor
of
being
more
granular,
pretty
much
all
the
time
and
I
feel
like
that's
kind
of
the
purpose
of
composition
and
those
sorts
of
things
that
were
able
to
make
these
abstractions.
B
E
Mean,
as
you
pointed
out
like,
we
usually
face
this
problem
with
only
eight
obvious
and
mostly
with
their
old
APs,
for
example.
In
this
case,
it's
like
an
only
list
and
change
APRs
exist
and,
like
you
know,
there
are
like
all
intricacies
about
the
API
like
you,
don't
have
any
unique
identifiers
and
it's
like
like
really
closely
coupled
like
tightly
coupled
with
hosted
zone,
but,
on
the
other
hand,
like
you,
know,
a
cloud
formation,
it's
one
of
them
as
separate
and
the
other
one
as
merged
the
toasted
zone.
That
are
like
no,
not
this
one.
E
So
yeah
I
mean
it's
a
bit
hard
to
draw
the
line
because,
for
example,
we
also
have
bucket
API
we'd,
like
for
example,
bucket
policy
is
a
different
API
call
like
create
bucket
policy
and
other
things,
but
you
can't
actually
use
bucket
pulls
here
with
other
buckets,
because
it's
just
a
configuration
of
that
bucket.
So
so
yeah
I
mean
it's
not
clear-cut
and
like
it's
kind
of
subjective
but
I
guess
I
guess
Nick
said
maybe
we
can
just
rely
on
terraform
classification
or
cloud
formations,
but
I
kind
of
prefer,
like
as
a
general
approach.
E
E
Then
I
think
it
could
be
easier
to
fan
out
the
cer.
You
know
as
a
migration
instead
of
you
know,
finding
again,
like
you
know,
if
it
was
separate,
it
would
be
harder
to
come
back
from
that
point.
If
it's
and
you
wouldn't
come
back
because,
like
users
wouldn't
come
and
say
hey,
these
are
like
I'm
separate
see,
ours,
I
can
work,
but
it's
like
you
know
bad
UX.
E
B
Yeah
I
feel
like
that,
in
that
case,
they're
like
I,
definitely
prefer
to
be
on
the
side
of
we're
already
more
granular
and
they're
like
it's
a
tough
UX
and
then
maybe
we
like
publish
some
default
compositions
with
the
provider
or
something
like
that
like
and
one
one
of
the
ways
you
just
have
to
write
some
compositions
and
make
those
available
in
the
other
one.
You
have
to
actually
write
new
controllers
and
break
api's
and
that
sort
of
thing
so
I
guess
that's
kind
of
my
take
on
it.
Yeah.
C
I
think
I
think
my
litmus
test
is
generally
I,
prefer
things
to
be
separate
resources
if
it's
possible
that
one
of
them
could
be
managed
by
a
cross
player,
and
one
was
not
so
I,
don't
like
any
case
I'm,
not
I'm,
not
sure.
If
this
is
the
case
in
this
particular
one
that
we're
talking
about
but
I,
don't
like
any
case
that
requires
you
to
say.
C
It's
used
example
that
I
do
understand
like
UJS.
If,
if
the
UK
is
no
pools
were
all
part
of
the
same
CR,
then
there
is
there
no
way
to
create
a
nucleus.
They
would
pull
with
that
also
pretty
unique,
a
yes
cluster.
So
in
those
cases
you
might
have
an
existing
key
to
get
S
cluster.
That
cosplay
doesn't
manage
one
cross
plane
to
manage
it,
and
then
you
want
to
use
like
to
manage
node
for
those
or
something
like
that.
E
Yeah
I
mean
it's
just
that,
like
you
know
not
that
clear
cut
as
eks
close
right,
not
cool,
because
I
mentioned
four
four
buckets
that
are
like
you
know,
maybe
10
different
structs
different
calls,
but
there
are
actually
under
the
same
API
in
cloud
formation
like
they're,
not
they're.
Only
a
pack
they're
only
a
few
properties
under
the
same
resource,
yeah.
C
I,
don't
I,
don't
think
I'm
making
the
case
that
I
I
think
that
we
mostly
understand
now
that
you
can't
just
say:
okay,
this
is
one
API
objective.
It
will
always
be
its
own.
Cr
I
definitely
agree
that
there
are
cases
where
it
makes
sense
to
interpret
a
bunch
of
different
API
causes,
as
one
so
I
do
understand.
We
have
to
play
this
on
a
case-by-case
basis,
usually.
B
C
I
say
like
if
existing
projects
in
our
area
have
had
some
have
faced
this
decision
and
chosen
to
do
one
thing:
I
might
not
to
stay
in
why
we
would
do
something
differently.
In
that
case,
as
I
say,
I
need
to
go
and
read
up
with
a
specific
reason.
Was
tonight
behind
you'll
actually
happen
about
this
one
as
a
general
direction.
C
A
B
That
being
said,
you
know
the
makefile
and
that
sort
of
thing
is
very
minimal.
It
doesn't
have
the
build
so
module
in
there
and
that
sort
of
thing
I
do
think
that
can
be
sort
of
nice
for
some
folks,
because
it's
very
easy
to
just
read
the
make
file.
I
know
it's
happening
and
know
kind
of
the
exact,
build
and
I
think.
If
it's
in
the
cross
plane
org
that
we
could
probably
be
more
diligent
about
keeping
it
up
to
date.
B
That
being
said,
it
is
another
thing
to
maintain
which
I'm
always
cognizant
of
as
a
function
right
now,
I'm
pretty
much
maintaining
it
outside
of
the
cross,
flane
org
as
people
use
it.
So
maybe
it
wouldn't
be
that
much
different,
but
I
just
thought
I'd
raise
it.
Send
some
folks
requested
that
you
know
it
might
be
helpful.
A
Yeah
I
think
it
general
for
me
it's
a
question
of
like
the
ongoing
value
of
it
and
if
folks
are
using
it-
or
you
know,
that's
serving
as
a
starting
point
for
folks
to
write
new
providers
and
that's
helping
flourish
or
support
that
part
of
the
ecosystem.
Then
that's
definitely
a
positive.
That's
nothing!
A
wins.
You
know
kind
of
reduced
barriers
or
burdens
for
for
people
to
get
into
it
and
get
started.
A
B
B
Think
like
right
now
that
wouldn't
be
super
discoverable,
but
if
we
really
start
pushing
folks
to
say
like
oh
yeah
like
go
over
there
and
will
actually
create
a
repo
for
you,
if
you
want-
and
you
can
kind
of
just
like-
have
a
little
bit
more
freedom
in
that
org
to
hack
on
stuff
and
try
stuff
out,
then
maybe
that
would
be
a
good
place
for
it.
So
either
either
way.
I
think
would
be
fine.
A
The-The-The
crossplane
contrib
organization,
to
talk
about
that's
right,
yeah
I,
don't
know,
I
think.
I
think
that
this
it's
somewhat
of
a
good
mashing
within
the
main
crossplane
organization,
because
it's
you
know
a
sort
of
sort
of
an
educational
record
sort
of
how
to
best
practices,
or
you
know,
structure
or
etc.
For
radical
the
provider
that
it's
not
a
bad
home
in
the
hospitalization.
C
It's
not
something:
it's
not
something:
creating
a
new
provider
isn't
something
that
we
do
super
frequently,
but
when
we
do
having
something
like
that,
bootstrap
that
would
probably
expediate
the
process
quite
a
lot.
It
would
give
us
something
else
to
maintain
a
little
bit
but
yeah
I.
My
inclination
would
be
to
just
put
it
on
to
the
cosplay
movie
I
think
in
practice
we're
using
it
already
yeah.
Not
many
people
think
about
it
is
something
that
the
cool
crossplane
maintain
is
maintained,
so
real
cosplaying
seems
better
than
crossplane
contrib.
C
B
And
if
it's
in
crossplane
org,
you
can
still
like
the
template.
Thing
works
to
any
repo,
so
I
think
it
would
be
more
of
a
tool
for
folks
that
are
like
I
want
to
get
started,
not
necessarily
like
I'm
making
this
like
I'm,
even
contributing
to
crossplane,
like
I,
just
want
to
make
my
own
provider
kind
of
thing
in
hack
around
on
my
counter
and
crosswind
contribute
whatever
and,
and
it
just
gives
them
the
scaffolding.
So
what.
A
A
It
seems
to
me
that,
like
that
in
the
Crimean
cross,
the
ignoring
makes
more
of
a
statement
that
hey
yeah.
This
is
useful.
It's
not
just
like
completely
excu.
Somebody
would
sit
up
like
overnight
and
it's
like.
No,
no
we're
not
gonna
be
a
ton
of
guarantees,
but
it's
it's
more
of
a
statement
of.
Please
do
use
this.
It's
functional!
It's
practical!
A
Yes,
if
we
didn't
follow
up
and
I,
don't
know
that
strip
of
organizations
of
repos
never
works
have
to
be
an
admin
on
both
sides,
but
no
creating
an
org
I
started,
creating
a
repo
and
didn't
then
doing
a
a
push
of
the
history
of
that
repo
into
it
and
as
any
remote
that
works
pretty
well.
I
think
does
it
happen.
Ii
I
can't
open
issues
or
anything
in
that
or
I'm.
Sorry
to
delight,
nor
am
I
mixing
important
people
I,
don't
think
so.
Yeah.