►
From YouTube: 2022-10-20 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
A
You
want
to
discuss
in
the
appropriate
section
and
try
to
put
your
name
attached
to
it
as
well,
so
that
we
we
know
who
is
adding
that
suggestion
and
we
will
get
through
all
the
agenda
today,
all
right.
So
as
usual,
we
go
ahead
and
start
with
our
Milestone
checkup
and
so
recent
releases.
A
There
is
a
very
recent
release
that
was,
you
know,
a
big
official
core
cross
plane
release,
so
the
1.10
release
is
out
as
of
Tuesday
I,
believe
so
mafik
had
the
honors
of
running
that
release,
so
I'm
going
to
pull
up
the
release,
notes
and
mafik.
If
you
want
to
give
us
a
little
bit
of
overview
of
that
release,
go
ahead.
B
There
are
lots
of
changes
in
this
release,
but
not
like
you
know.
Two
big
features
introductions
which
kind
of
slipped
through
the
the
next
series,
so
the
the
biggest
ones
as
listed
here
are
the
are
that
now
we
have
maintainers
and
reviews
groups.
This
is
not
specific
to
this
version,
but,
like
you
know,
since
the
last
one,
that's
that's
a
big
change
and
we
have
upgraded
to
1.19
all
the
dependencies
that
we
have
at
Cross
plane,
including
runtime
and
cross
between
itself
and
there's
now
a
new
way
to
provide
package
pool
Secrets.
B
If
you
provide
package
pool
stickers
during
the
installation
of
cross
plane,
then
it
will
propagate
to
all
providers,
so
you
won't
have
to
deal
with
pull
Secrets
at
all.
Once
you
install,
crossplaying
and-
and
one
of
the
biggest
features
around
this
time
is
the
patching
from
environment
config
and
its
design
dock
has
made
it,
but
implementation
hasn't
so
I
would
say.
Look
for,
like
you
know,
1.11,
which
will
sell
like
a
lot
of
customer
a
lot
of
users.
B
Problems
regarding,
like
you
know,
deploying
the
same
configuration
in
different
different
environments
and,
like
you
know,
having
access
to
static
values
and
there's
also
a
new
field
on
the
claims
now
called
the
lead
composite
policy
which
will
allow
you
to
configure
how
claim
deletion
effect
composite
normally
by
default,
we
would
issue
a
background
deletion
which
will
make
which
would
make
clean
go
away
as
soon
as
it
makes
the
deletion
call,
but,
like
you
know,
compose
it
may
get
stuck
there
so
with
this
composite
policy.
B
If
you
choose
foreground
policy,
then,
unless
composite
is
gone,
claim
is
still
in
your
cluster
existing.
This
is
still
kept
in
your
cluster
and,
like
you
know,
many
goodies
as
listed
there
as
as
for
every
PR.
That's.
A
A
Okay,
so
I
like
the
goodies.
That
sounds
good,
a
couple,
quick
questions
and
Bob.
It
looks
like
you're
on
the
call
here.
So
maybe
you
can
address
this,
you
know
I
I
had
click
I'm,
clicking
and
dragging
I.
Don't
want
to
do
that.
I
had
an
impression
that
you
know
the
garbage
collection.
You
know,
deletion,
Behavior
type
of
functionality
was
it's
still
true.
It
is
a
hairy
problem,
but
we
got
some
pretty
pretty
beneficial.
You
know
high
level
behavior
in
for
1.10.
Is
there?
A
C
That's
a
great
question
it,
so
it
turns
out
that
the
delete
the
the
changes
in
in
the
highlighted
section
there
are
important,
but
for
a
different
reason.
So
what
those
changes
will
do
is
highlight
problems
with
stuck
managed
resources
in
terms
of
you
know,
in
the
in
prior
to
1.10
or
actually
even
in
1.10.
C
If
you
delete
with
background
and
not
foreground,
you
know
the
claim
and
all
the
Composites
will
go
away,
but
you
could
still
have
managed
resources
sitting
out
there
that
are
hung
for
one
reason
or
another,
and
you
would
never
know
that
unless
you
did
a
cube,
control,
get
managed
and
you
know
and
saw
what
is
still
kind
of
hanging
out
there.
C
If
you
set
the
delete
composite
policy
to
foreground,
then
what
will
happen
is
the
claim
when
the
claim
deletes
it
will
trigger
a
cascading
foreground
deletion
in
the
in
the
rest
of
the
resources
and
those
Composites
and
and
at
the
end
of
the
day
the
claim
won't
go
away
until
all
of
the
managed
resources
have
gone
away,
and
so
that
will
that
bring
some
visibility
to
some
stock
managed
resource
issues.
C
The
kind
of
the
the
parallel
issue
or
the
the
related
issue
in
terms
of
deletion
dependencies
and
not
deleting
things
until
their
dependencies
have
gone
away,
is
assisted
by
this.
It's
kind
of
you
know
it
kind
of
makes
it
more
obvious
that
there's
a
problem
there,
but
the
resolution
to
that
is
is
kind
of
a
bigger
problem.
I,
don't
know
if
it's
actually
targeted
for
1.11.
C
Nick
opened
an
issue
on
it.
He's
I
think
he's
calling
it
a
usage
resource
right
now
that
will
track,
allow
you
to
track
dependencies
between
resources
and
actually
block
their
deletion
until
the
dependencies
have
gone
away
and
I
did
model
something
like
that.
C
Using
some
Chi
Vernal
policies,
where
using
a
combination
of
two
policies
and
a
custom
resource,
you
can
simulate
well,
not
simulate,
you
can
actually
make
it
work
that
way
in
terms
of
you
can
block
deletion
of
certain
resources,
while
they're
being
used
by
other
resources,
but
the
official
implementation
that
yeah
exactly
and
I
do
like
I,
say:
I,
don't
know
if
that's
actually
targeted
for
1.11..
C
It
is
obviously
well
described
here
and
I.
Think
Nick
has
a
good
idea
of
of
how
he
wants
to
proceed
on
that.
B
I
was
just
going
to
ask,
like
you,
know,
I
think
we
already
discussed
this
deletion
problem
Bob
one
I
made
a
suggestion
regarding
like
a
connection
secret
usage.
Is
this
a
new
secret?
Is
this
new
usage
type
similar
to
that
or
is
it
more
like
you
know
manually,
you
would.
Is
it
something
that
you
would
create
manually
to
represent
the
dependency.
C
Yeah
I
think
it's
more
something
that
the
comp,
the
composition
developer,
would
would
specify
manually.
Now
you
know
the
the
actual
interface
for
doing
that.
I,
don't
know
that
has
been
worked
out
yet,
but
you
know
I'm
doing
it
with
annotations
and
labels.
C
Just
in
my
POC
grade
version
of
the
of
the
solution,
I,
don't
I
hesitate
to
call
it
a
solution
but
yeah
the
implementation,
but
I
think
it's
it's
more
intended
to
be
to
allow
the
user
to
say,
hey
I
know
these
two
resources
have
dependencies
that
maybe
cross
playing
can't
see
and
and
to
allow
the
user
to
specify
those
manually,
at
least
that's
my
impression
so
far,
better
thanks.
A
Awesome
yeah
so
Bobby
thanks
a
lot
for
driving
a
lot
of
this
general
area
of
functionality.
A
We
are
just
a
note
for
everybody
at
kubecon
next
week,
Bob
will
be
talking
through
some
of
this
in
more
detail
in
the
Deep
dive
session
that
we're
doing
for
the
cross-playing
session
of
the
in
the
maintainer
track.
So
there
will
be
more
more
details
about
this
and
more
areas
to
dive
into
the
other
thing
for
1.10.
That
I
wanted
to
call
out
is
the
new
contributors
list.
Here.
A
That's
a
pretty
healthy
set
of
folks
that
are,
you
know,
identified
as
having
their
first
contribution
to
core
cross
plain
itself.
I
know
that
a
lot
of
these
folks
have
been
working
on
providers
and
then
maybe
some
stuff
in
runtime,
but
within
four
cross
plane.
This
is
a
set
of
new
contributors
here,
so
that's
great
to
see
new
folks
getting
involved
and
new
folks
writing
code,
and
you
know
all
the
ways
that
we
contribute.
A
So
that's
that's
fantastic
to
see
that,
but
all
right
so
great
stuff
there,
thanks
for
writing
that
release
on
Tuesday
thanks
everybody
for
contributing
to
it
definitely
happy
to
get
that
out,
and
so
now
we
start
the
next
release
cycle
towards
1.11.
This
is
you
know,
we're
two
days
into
it.
So
I
cannot
say
that
there
has
been
much
planning
thinking
or
prioritization
at
all
on
1.11
super
early
on
that
the
release
won't
be
until
you
know,
or
on
a
quarterly
basis,
so
the
release
will
not
be
until
next
year.
A
You
can
always
follow
along
on
the
calendar
for
releases
and
upcoming
things,
and
then
the
crossblade
releases
project
board
is
still
you
know,
running
and
serving
as
our
roadmap
here,
or
you
know
more
as
areas
of
feature
Investments
that
maybe
not
maybe
not
completely
committed
to
or
have
owners
for,
but
things
that
are
important
for
the
community
and
things
that
we're
identifying
and
bubbling
up
to
the
top
that
there's
demand
for
so
you
know
as
we
go
through
the
1.11
cycle,
you
know
we're
well,
there's
a
good
opportunity
to
you
know
to
take
in
feedback
and
take
in
you
know,
ideas
or
proposals
for
what
to
work
on
in
1.11
time
frame,
and
you
know
contributors
and
folks
that
want
help
with
making
progress
on
things.
A
A
A
Let's
keep
on
moving
here
then.
So,
let's
move
into
the
provider
section.
So
we
have
a
a
dedicated
section
later
on
in
the
agenda
that
I
wanted
to
get
to
as
well
related
to
providers.
So
we'll
move
quickly
through,
what's
here
now
and
we'll
get
into
that
section
later
on,
but
so
the
there
is
a
in
the
recent
releases
section
I
know
that
there
is
a
recent
provider
AWS
released
since
the
last
community
meeting
and
then
there's
a
set
of
providers.
A
The
providers
from
upbound
that
have
been
released
as
well
we'll
get
into
many
more
details
on
that
later
on
in
the
agenda.
But
anybody
else
who
wants
to
call
out
other
recent
releases
of
providers
feel
free
to
add
them
the
links
to
them
here
in
this
list
or
call
them
out,
and
we
could
definitely
get
those
included
in
there
as
well.
A
All
right,
let's,
let's
keep
on
moving
then
to
the
next
couple
of
sections
here,
so
a
couple
of
cool
pieces
of
contents
that
I
wanted
to
share
with
everybody,
so
at
Google
Cloud.
Next
there
is
in
in
one
of
the
talks
there
might
have
been
a
keynote
I'm,
not
quite
sure
actually
but
they're,
showing
anthos
running
on
eks
and
using
crossblades.
It's
kind
of
like
you
know
a
few
words
that
you
wouldn't
necessarily
see
mixed
or
didn't
expect
to
see
mixed
up
there.
A
But
yes,
that
is
correct,
with
Google
xanthos
running
on
eks
and
then
using
Crossman
to
orchestrates
of
that.
So
that
was
really
definitely
cool
to
to
see
that
come
up.
That
was
a
bit
of
a
surprise
for
me.
I
think
other
folks
might
have
known
about
that.
But
I
was
definitely
stoked
to
see
that
and
then
there's
an
upcoming
talk
from
arsh
and
that's
going
to
be
kubernetes
Community,
Days,
I
I
think
this.
A
Maybe
this
is
the
first
one
they're
doing
in
Dubai,
but
you
know
those
are
kind
of
a
bit
smaller
scoped
than
the
kubecons
themselves,
but
you
know
being
able
to
have
a
bit
more
Global
presence
for
them.
So
Arch
is
going
to
be
presenting
on
Cross
plain
at
this
upcoming
kubernetes
Community
Days
of
Dubai.
A
If
anybody
else
has
you
know
a
blog
posts
or
videos
or
anything
like
that,
feel
free
to
add
those
to
the
agenda
doc
here,
so
they
get
a
little
bit
more
presence
or
been
a
publicity.
And
then,
of
course,
or
you
know,
the
cross-plane
Twitter
account
is
always
happy
to
retreat
and
amplify
stuff
that
you
all
are
doing
and
writing
and
producing
and
creating
with
crossplay
as
well
so
always
feel
free
to
reach
out.
And
we
can
we're
happy
to
retweet
and
amplify
some
stuff.
A
Kubecon
is
next
week,
so
I
can't
really
say
it
snuck
up
on
us,
but
it
feels
like
it
came
up
all
of
a
sudden
a
little
bit.
So
hopefully,
we're
going
to
be
seeing
a
bunch
of
you
folks
next
week
in
lovely,
lovely,
Detroit,
I
know
a
handful
of
folks
that
are
on
this
call
right
now,
I'm
definitely
planning
on
seeing
seeing
some
of
you
already
and
you
know
we'll
be
at
the
the
cross
plain
booth
in
the
cncf's
project,
Pavilion
area
we'll
be
there
all
week
as
well.
A
So
there's
lots
of
opportunities
to
chat.
Ask
questions
provide
feedback.
We
have
a
number
of
talks
on
crossband
as
well,
which
I've
been
talking
about
basically
every
week
in
this
community
meeting,
so
there's
a
direct
link
to
the
schedule
there
for
all
the
talks
related
to
cross-plane
note
to
every
single
one
of
them
is
on
my
schedule
as
well
too
so
I'm
excited
about
all
of
those
obviously
and
yeah.
The
the
in
the
last
one
I
believe,
is
that
maintainer
track
session.
A
That
I
was
talking
about
where
Matthias
and
I
will
be
kind
of
giving
the
introduction
to
cross
plain
and
teaching.
Folks,
maybe
that
are
you
know,
a
newer,
less
familiar
audience
with
crossplay
and
we'll
be
kind
of
teaching
some
of
the
introductory
material,
but
in
the
Deep
dive,
Nick
and
Bob
will
be
going
into
some
pretty
cool
stuff.
Nick
will
be
focusing
on
composition,
functions,
which
is,
you
know,
upcoming
functionality.
A
That
is,
you
know,
being
thought
through
very
thoroughly,
and
you
know
it's
something
that
we're
starting
to
invest
more
and
more
in
to
bring
that
to
the
community,
and
then
Bob
will
be
talking
in
more
details,
as
we
mentioned
about
the
garbage
collection,
resource
deletion,
functionality
stuff.
A
So
definitely
looking
forward
to
that
talk
and
all
the
other
ones
here
as
well
lots
of
opportunities
for
for
folks
to
check
out
cool
stuff,
you
know
Jesse's
talk
is
is
on
here,
Dan's,
my
favorite
I
think
all
those
people
are
on
this
manibu
as
well.
So
a
lot
of
pretty
much
all
the
speakers
are
on
the
meeting
right
now.
So
awesome,
I'm,
looking
forward
to
everyone's
talks
next
week,
really
really
excited
about
that.
A
So
come
by
and
check
us
out
at
the
booth
as
well
we'll
be
hanging
out
there
all
day,
so
yeah,
let's,
let's
time
to
hang
out,
let's
see
Al
pair,
you
were
on
the
call
today.
Sir,
do
you
want
to
give
us
a
quick
update
on
what
your
effort
was
around
the
the
fuzz
testing,
because
the
the
fuzz
testing
that
cncf
is
doing
that
project
is
wrapping
up
now,
so
Albert
has
been
working
on
fixing
some
of
the
issues
that
have
come
up
with
it,
so
yeah
Opera.
D
Yeah,
so
thank
you
Jared,
so
the
first
issue
that
we
had
examined
was
an
out-of-memory
issue.
However,
you
know
when
we
dig
a
little
bit
deeper
into
how
these
tests
are
run.
D
Also,
this
is
documented
by
you
know
the
OSS
fuzz,
you
know,
folks,
it's
generally
out
of
memory
errors
or
timeouts
may
not
be
treated
as
actual.
You
know,
issues
by
the
rep
owners
or
by
the
repo
maintainers
sort
of
thing,
and
you
know
we
had.
We
are
for
now
closed
that
issue.
This
doesn't
necessarily
mean
that
we
will
not
work
on
it
in
later,
but
there
was
another
issue.
D
You
know
that
prevented
these
fuzzing
tests
to
increase
coverage,
because
you
know
these
fuzzing
tests
work
by
penetrating
through
API
layers.
You
know-
and
you
know
there
was
some
panicking
issue
because
of
a
near
pointer-
the
reference
that
prevented
you
know
for
these
fuzzing
tests
to
delve
into
deeper
our
cross-plane
API
patch
API.
So
this
was,
you
know,
brought
up
to
our
attention.
You
know
from
colleagues
from
other
Logics
and
we
have
identified
root
cause,
so
the
root
cause
infectives
I
will
also
maybe
leave
a
link
for
that
in
the
document.
D
It's
not
in
the
cross
plane.
It's
in
Upstream
libraries
that
the
controller
excuse
me
that
the
API
Machinery
Library
depends
on,
but
right
now
on
the
cross
plane
side,
we
have
what
I
would
call
as
a
workaround,
so
that
you
know
the
fuzzing
tests
we
can.
We
can
increase
the
coverage
of
the
fuzzing
tests.
D
A
Alberta
thanks
a
lot
for
doing
that,
and
you
know
digging
into
you
know
those
those
set
of
issues
that
were
found
during
the
first
testing.
I
think
this
is,
we
can
look
at
this.
Also,
as
the
start
of
you
know,
a
a
bigger
security.
You
know
stance
and
and
dive
along
with
the
cncf,
because
you
know
we
are,
we
are
complaining
to.
You
know,
make
the
proposal
and
push
on.
A
You
know
maturing
cross-plane
to
the
point
where
it
could
be
fully
graduated
with
the
cncf,
so
there
will
be
a
security,
a
general
security
audit,
that's
associated
with
that
as
well.
So
this
first,
you
know
step
into
fuzz
testing.
A
You
know
is
a
a
early
step
in
that
direction
and
there
will
be
much
more
to
come
with
the
security
posture
and
stance
for
for
the
cross
claim
so
really
happy
to
get
this
done,
and
you
know
grateful
to
the
cncf
for
helping
in
the
Ada
logic
folks
for
helping
drive
this
and
share
the
their
findings
with
us
and
then
out
there
for
for
fixing
stuff.
So
thanks
a
lot
for
all
that
stuff
all
right,
so
we
also
wanted
to
talk
about
Pete's
proposal
for
the
website.
A
Docs
restructure,
so
Pete
I'm
going
to
go
ahead
and
let
you
share
the
your
screen
real,
quick
to
drive
on
that
one
buddy.
Thank
you.
Jerry.
E
I'll
walk
through
this.
Just
briefly,
there's
been
some
questions
in
the
slack
I.
Think
there's
been
some
challenges
with
new
folks
like
myself,
coming
into
cross
plain
and
trying
to
understand
exactly
where
we
want
to
get
started,
how
to
walk
through
it.
E
You
know
I,
think
that
the
learning
curve
continues
to
be
a
little
steep
and
as
I'm
working
on
some
pretty
big
content
contributions
coming
up,
I
was
looking
at
the
overall
structure
and
how
we
can
improve
that.
How
we
can
make
that
simpler
so
on
the
left
is
the
existing
docs
just
for
reference,
but
really
how
I'm
thinking
about
reorganizing.
This
is
trying
to
break
it
down
into
some
larger,
big
bucket
sections
that
seem
a
little
bit
more
obvious.
E
Sometimes
it's
just
every
name
and
then
introducing
a
new,
a
few
new
Concepts.
So
one
of
them
is
this
idea
of
Quick
Start
guides,
so
whatever
providers
want
to
show
their
quick
start,
you
know
whether
that
comes
from
the
community
from
the
provider
developer.
Somebody
else
entirely,
you
know
how
do
we
get
off
the
ground?
How
do
we
install
that
provider?
How
do
we
show
that
that
provider
works
and
generally
tests
credentials
so
we're
not
trying
to
overwhelm
that
reader
with
a
ton
of
content
all
of
the
capabilities
across
plane?
E
It's
it's
just
a
single
toe
in
the
water
insulin,
management
I
think
is
pretty
obvious,
but
centralizing
that
I
think.
Today
we
have
two
different,
install
cross-plane
sections,
for
example,
that
alone
is
pretty
confusing
to
folks
breaking
down
those
components
which
is
really
the
concepts
section
today.
E
Packages
breaking
that
in
to
its
own
section,
building
publishing
in
the
spec
the
guides
are
more
or
less
going
to
stay
the
same
so
nothing's
there
and
then
consolidating
a
lot
of
some
of
the
miscellaneous
things
that
exist
in
different
places
into
a
combined
additional
information
section
and
then
finally,
reference
implementation.
So
this
is
something
that
could
be
a
baseline
for
a
production,
implementation
of
a
provider
or
set
of
providers,
and
this
is
specifically
different
from
the
quick
start,
because
it
is
the
full
thing
right.
E
This
is
where
we're
going
to
include
Composites
we're
going
to
include
packages
and
really
show
off
how
do
I
take
a
provider
and
build
something
usable
out
of
it,
not
just
turn
it
on,
and
so,
if
you
think
about
the
top
there
at
the
quick
start,
is
I'm
new
to
cross
playing.
How
do
I
use
it
for
the
first
time
and
just
see
it
do
something
show
me
I
want
to
see
the
light
go
from
red
to
Green
the
reference
implementation
is
okay.
E
I've
got
my
hands
on
how
cross-plane
Works
what
it
does
I'm
going
to
go
run
at
production.
Can
somebody
give
me
an
example
to
follow
of
how
to
build
all
of
those
components
and
glue
them
together
and
then
I
won't
go
into
details
here,
but
also
just
a
mapping
of
all
the
existing
pages
and
where
they
would
end
up
Landing
with
this
reorg?
A
Yeah,
you
know
Pete
one
of
the
biggest
things
for
me
that
I,
you
know
there's
like
with
the
history
of
of
the
docs
that
we've
got
right
now,
one
of
the
things
that
we
were
attempting
to
do
I
believe
at
the
current
iteration
of
the
flow
is,
you
know,
kind
of
demonstrate
and
help
people
understand
like
the
power
and
magic
of
crossblane
in
the
sense
of
you
know,
empowering
or
enabling
application
developers
to
do.
A
You
know
self-service
with
you
know:
abstractions
like
infrastructure,
apis
right,
it's
like
so
they
get
this
experience
of
you
know.
Cool
I
can
use
this
thing
and
oh
wow
I've
had
infrastructure
where
to
come
from
I
Didn't
Define
it
like
oh,
my
platform
team
like
they.
They
Define
this
for
me
and
then
now
I
can
create
it
myself.
And
it's
like
this
nice.
A
You
know
magical
experience
there,
I
I,
wonder
like
so
we
kind
of
lead
forward
with
you
know
compositions,
or
you
know
that
experience
there
and
so
I
wonder
if
there's
I
wonder
if
there's
like
some
explanation,
some
pros
some
some
description
that,
like
can
really
kind
of
quickly
get
people
to
understand
that
or
kind
of
put
that
forward
of
like
this
is
the
magic.
This
is
the
power,
like
you
know,
kind
of
a
little
bit
more
easy
to
understand
and
digest.
You
know
some.
Definitely
some
documentation,
expertise
there
to
really
make
it
digestible.
A
You
know
that
we
can
also
get
into
this
to
help
out
with
like
helping
people
get
cross-plane
with
some
simple,
simple
explanation
to
it:
yeah.
E
E
You
know
I,
think
there's
two
kinds
of
cross-play
magic
moments,
I
think
there's
one
as
an
operator
where
you
see
kubernetes
Drive,
something
that's
not
in
kubernetes
right.
Oh
my
God,
my
cluster
just
created
an
S3
bucket.
That's
amazing!
You
know
I
I,
see
on
one
side
and
I,
see
the
S3
AWS
console
on
the
other
and
I
see
something
just
show
up
in
the
console
out
of
nowhere.
I
think
that's
the
first
kind
of
simplest
magic
moment.
E
I
think
the
other
Magic
Moment
is
on
the
other
side
of
the
Cross
Point
consumption,
where
what
you're
talking
about
is
like
how
do
I?
How
do
I
show
an
API?
How
do
I
actually
take
something
complete
and
see?
Crossplay,
do
its
whole
thing
right
and
I
think
there
is
actually
another
opportunity
for
a
quick
start
in
a
different
sense
and
what
I
mean
by
that
is,
instead
of
causing
requiring
you
to
walk
through
every
single
step.
To
get
to
that
say,
creating
a
claim
or
calling
a
claim.
E
We
give
you
99
and
you
you
fill
in
one
box,
I'm
oversimplifying,
but
you
know
we
want
you
to
have
this
little
simple
amount
of
work
to
do
to
help
solidify
that
understanding
to
create
that
more
advanced
consumption
magic
moment.
So
I
think
there's
an
opportunity
for
a
quick
start
there
in
the
in
the
near
future.
A
Yeah,
okay,
yeah,
okay,
I
I,
see
what
you're
saying
there
yeah
I
think
like
if
I
got,
if
I
followed
you
along
there,
you
know
I,
think
that,
like
being
able
to
put
that
you
know
at
the
at
the
front
at
the
you
know,
leading
with
that
type
of
thing
on
my
other
quick
start
type
of
functionality
right,
like
that's
kind
of
important
thing.
A
E
I
think
one
of
the
challenges
today
is
we
get
a
person
to
that
magic
moment,
mainly
through
the
the
current
getting
started
section,
but
it's
still
a
heavy
lift.
E
You
know
we
kind
of
require
you
to
build
every
one
of
the
blocks
to
get
to
the
point
where
you
see
the
magic
and
without
that
understanding
of
what
those
components
are
it
I
think
there's
a
lot
of
toil
there
and
so
I
think
my
proposal,
kind
of
independent
of
this
is
we
would
need
a
quick
start
that,
like
I
said,
accelerates
the
time
to
a
composite
claim.
E
Magic
Moment,
you
know
so
we
give
like
I
said
we
give
you
98
of
that,
but
already
together
you
know
whether
that's
through
a
package
or
something
like
that
and
we
just
say,
run
this
install
it.
And
now,
if
you
do
these
couple
of
things
and
I
will
teach
you
about
those
things
along
the
way
you
will
see
how
to
consume
an
API
and
I.
Think
if
you
compare
that
to
the
quick
starts,
we
have
for
AWS,
gcp
and
Azure
again
like
we're.
E
Assuming
you
know
what
kubernetes
is,
and
we
assume
you
know
what
AWS
is
and
we're
kind
of
walking
you
through.
What
is
a
provider?
How
do
I
install
it?
How
do
I
create
a
managed
resource?
And
again
it's
not
this,
isn't
for
production?
It's
not
to
make
you
an
expert.
It's
to
to
Really,
expose
you
to
that
the
magic
there
and
I
think
you
do
the
same
thing
like
I,
said
at
the
higher
level
of
Composites
and
claims.
A
I
definitely
like
that
mentality
of
accelerating,
like
the
the
time
it
takes
to
get
to
that
magic
moment
and
reducing
toil
to
get
there
as
well
like
that.
That
General
approach
does
sound
like
that,
could
have
a
lot
of
benefit.
E
I'll
I'll
file
an
issue
after
this
and
send
it
to
myself
to
work
on
that.
I
mean
it'll,
probably
be
a
a
couple
of
weeks
before
I
can
even
start
working
on
it,
but
I
think
it's
important
to
to
track
that
idea
of
that.
Other
claim
quick
start
and
what
that's
going
to
look
like.
F
G
F
Like,
on
the
other
hand,
as
well
having
a
bit
more
documentation
on
how
it
actually
works
behind
the
scenes,
if,
if
someone
wants
to
contribute
and
like
not
only
fight
for
the
user's
perspective,
but
also
for
the
contributors
perspective,
because
like
at
least
for
me,
it
was
a
little
bit
of
a
lot
of
reading
codes
to
be
able
to
understand
what
actually
happens
behind
the
scenes.
So
I
think
it
would
be
very
useful
to
have.
E
Yeah-
and
that
is
one
of
my
goals-
I
mean
Yuri-
can
vouch
for
it,
as
I
was
asking
about
a
a
whole
bunch
of
extremely
simple
questions
yesterday
to
try
to
make
sure
we
can
get
to
that
core
piece
and
work
up.
E
You
know:
I
was
looking
at
some
authentication
stuff
related
to
AWS
yesterday,
and
so
I
was
like.
Okay,
I
want
to
understand
we're
creating
a
controller
config
right,
that's
a
totally
different
thing.
What
is
that?
Why
does
that
matter?
How
does
that
link
to
these
things
that
controller
can
fit
like
I
get
a
controller
config
it's
consumed
by
the
provider,
who
consumes
a
provider
config
but
like?
Why
do
I
care?
What
is
this
actually
doing?
E
A
Yeah
and
Roberto-
that's
that
also
reminds
me
too,
of
like
there's
some
things
that
I
want
to
do
soon
around
like
contributor
enablement,
and
you
know
like
giving
giving
new
contributors
more
resources
to
get
to
productivity
faster.
You
know
to
be
able
to
grow
the
number
of
people
that
we
have.
You
know
writing
code
and
being
productive
in
the
ecosystem,
so
yeah
that
that's
that's
kind
of
something
that
would
be
very
helpful
with
that
of.
A
G
User
and
I'm,
just
listening
in
it,
sounds
like
you
guys,
actually
have
largely
discussed
what
I
was
going
to
talk
about
or
suggest,
but
a
map
of
you
know
how
these
components
rely
on
each
other
would
be
super
helpful.
That
was
definitely
a
struggle
for
me
is
figuring
out,
hey,
you
need
this
like
provider
definition
and
then
you
need.
G
G
That
sort
of
mapping
would
be
super
helpful
from
a
user
perspective
of
how
how
these
sort
of
things
rely
on
each
other
and
what
order
they
come
out.
I
wasn't
able
to
find.
E
Yeah
and
I
think
there's
some
work.
There
I
submitted
a
a
PR,
that's
in
review
to
to
contribute.
These
quick
starts
that
we
wrote
on
the
upbound
side,
but
again
the
goal
for
me
coming
in
as
a
pretty
new
person
to
not
just
cross-playing
but
also
kubernetes
and
even
to
some
extent
AWS
is
really
to
say,
like
hey
get
a
secret,
you
know
install
cross-plane,
you
can
see
how
it
created
these
new
API
endpoints.
This
is
what
it's
doing.
E
This
is
what
changes
you
know
we
install
a
provider
so
that
cross
plane
knows
how
to
talk
to
AWS.
You
can
see,
look,
look.
It
creates
a
bunch
of
new
kubernetes
endpoints
because
of
the
provider
really
trying
to
show
each
one
of
those
components
and
how
they
link
through
and
and
how
that
matters
right.
So
here
saying:
look
the
spec,
credential
secret
ref
name
needs
to
map
to
the
name
of
the
kubernetes
secret.
This
is
how
it's
looking
that
stuff
up.
E
That's
exactly
that
kind
of
thread
that
I'm
trying
to
pull
through
the
whole
thing.
So,
if
I'm
going
to
build
a
system,
every
every
one
of
those
blocks,
I'm
gonna,
go
back
to
you
know
college
database
class,
which
the
last
time
I
touched
a
database.
You
know
like
there's,
there's
always
like
a
primary
and
a
foreign
key
for
every
one
of
these
things.
So
how
do
I?
How.
E
100
with
you
and
I
think,
that's
absolutely
the
goal
that
I'm
going
for
for
these
things.
You
know
even
this,
like
the
managed
resource.
Look
that
API
version
comes
from
the
provider
crd
right.
The
name
is
the
name
of
your
bucket
that
we're
going
to
go
and
create
in
AWS
and
the
region
right.
So
again,
this
like
random,
API,
endpoint
like
no.
No,
it
actually
has
a
practical,
tangible
thing
to
think
about
and
Implement.
A
Very
cool,
very,
very
cool
I,
like
the
different
you
know,
layers
we're
kind
of
talking
about
here
of
you
know
what
what
layer
of
you
know
peeling
back
the
onion.
Do
people
need
to
understand
to
be
productive
in
their
current
task
and
it's
you
know.
This
is
really
nice
that
we
have.
You
know
we've
we've
we've
done
a
great
job
as
as
Engineers
on
the
project
of
writing
documentation,
but
having
Pete,
like
you
know,
is
obviously
a
step
up.
You.
E
So
if
anybody
has
any
other
comments,
you
know
please
go
ahead
and
comment
on
the
issue.
33.98
I
Know
Dan
had
some
stuff
that
I
have
to
follow
up
with
him
about,
but
really
hoping
to
try
to
wrap
this
up
this
week.
Unless
folks
would
like
me
to
wait
longer
on
that
for
a
feedback.
A
And
I
know
that
Nick
does
because
Nick's
on
PTO
and
he
does
have
some
opinions
so
yeah,
please,
please
do
wait
for
Nick's
Nick's
feedback
next
early
next
week.
I
Before
we
move
on
just
one
comment
on
the
documentation
side
of
things
so
like
technical
details
and
all
that
is
good,
but
one
thing
I
will
notice,
is
it
doesn't
really
like
address
the?
Why,
of
course,
plane
right
like?
Why
do
you
want
to
move
away
from,
like
you
know,
terraform
or
cdk,
or
anything
like
that
right
like?
I
Why
would
you
know
it's
like
an
Enterprise
companies
want
to
use
this
instead
of
that,
like
you
know,
addressing
that
kind
of
you
know,
questions
I
mean
like
that
QA
section,
or
something
like
that,
that'll
be
great
too.
E
Yeah,
that's
that's
good
to
know.
I
I
think
there's
absolutely
a
need
for
some
of
that
on
the
crosswind
site.
E
I'm
gonna
draw
a
thin
line
between
say,
user
documentation,
and
you
know
what
we
would
classically
call
like
more
like
product
marketing,
but
I
I
totally
agree
with
you
there's
a
you
know
why.
Why
should
I
care
component?
That
I
think
we
can
do
a
lot
more
work
on.
A
There's
an
investment
being
made
as
well
right
now
in,
like
the
main
crossband.io
site
to
you,
know
talk
about
the
control,
plane,
architecture
or
like
what
its
benefits
are.
Why
you
would
want
to
do
this
and
kind
of
help
understand
the
value
a
little
bit
more
easily
as
well
for
people.
That's.
I
A
Right
on
all
right,
Pete,
thanks
for
sharing
all
that
with
us
and
kind
of
give
helping
Drive
some
of
the
documentation
improvements
there.
So,
let's
keep
on
moving
through
the
agenda,
we're
at
38
after
here.
So
just
20
minutes
left
we're
two-thirds
through
so
I
want
to
jump
into
an
announcement
and
a
discussion
area
now.
So
you
know
around
the
fighter
ecosystem
so
and
I.
A
Think
a
big
announcement
from
this
week
is
that
upbound
has
created
a
Marketplace
for
for
cross-plane
and
and
the
community
here,
and
it's
it's
GA
it's
available
now.
So
I
think
the
biggest
thing
to
take
from
this
is
that
it
is,
you
know,
an
attempt
at
a
place
for
the
entire
Community
to
Rally
together
and
to
be
able
to
collaborate
and
share
and
discover
all
sorts
of
you
know
the
great
things
providers
configurations,
Packages,
Etc,
that
we're
all
building
together,
I
think
there's
a
whole
lot
of
things
as
a
community.
A
We're
building
and
I
think
this
is
a
really
an
attempt
at
a
really
good
place
to
get
together
and
find
a
mall
and
be
productive
together.
Right
so
I
think
that
you
know
folks
have
definitely
been
using
DACA
crds.dev
for
a
while
to
you
know
for
crd
documentation
and
you
know
managed
resource
documentation
Etc.
So
if
you've
liked
that,
then
this
kind
of
levels
it
up
a
few
levels
and
that's
the
intent
of
you
know
having
references
and
just
you
know,
having
a
place
for
people,
the
entire
Community
to
be
productive
together.
A
So
it's
open
source,
it's
free!
It's
it's!
You
know
available
for
anybody
to
access
and
then
so
in
that
Marketplace
you'd
also
discover
a
few
new
providers
that
upbound
has
released
as
well.
So
in
these
providers
you
know
I
really
want
to
think
about
what
the
goal
of
outbounds
official
providers
are
and
really
it's
it's
to
help
improve
cross-bane
adoption.
You
know
high
quality
High
coverage
providers.
They
can
really
accelerate
adoption
for
folks
that
want
to
start
using
cross
plane
and
remove
some
friction
in
barriers
right.
A
So
you
know
growing
in
in
and
expanding
the
ecosystem
here
in
our
community
is
is
really
a
big
big
goal
of
this,
so
those
are
open
source
free
to
use
as
well.
These
providers-
and
you
know,
contributions
and
collaborations
there
are
more
and
more
more
than
welcome,
I
think
it's
also
important
to
take
a
to
kind
of
have
a
bit
of
a
reminder
as
well
about
you
know
the
Crossbay
project,
the
project
and
governance.
Here
you
know,
providers
are
are
in.
The
charter
are
out
of
scope
of
the
charter.
A
You
know
the
Crosman
project
does
not
necessarily
endorse
or
or
recommend,
any
specific
providers,
but
that
does
not
mean
that
you
know
we
can't
be
facilitating
conversations
about
converging
on
providers
together
rallying
around
them
together
and
all
you
know
all
come
together
as
a
community
on
all
of
these.
So
you
know
a
note
that
the
cross-plane
contribute
org.
That's
you
know
absolutely
they
is
still
a
place
that
remains
a
place.
A
So
if
you
want
a
neutral
place
to
collaborate
on
writers-
and
you
know
to
be
working
together
with
the
community,
that's
definitely
a
place
to
to
do
that,
and
you
know
nothing
is,
is
indifferent
about
that
than
it
has
been.
But
you
know
we
can.
We
could
be
having
conversations
about.
This
is
an
opportunity
potentially
to
be
consolidating
on
some
of
our
provider
efforts.
Everyone
is
happy
or
is
you
know,
open
and
encouraged
to
be
building
providers
as
they
want
amongst
the
community,
but
you
know
we
could
have.
A
We
do
have
an
opportunity
as
well
to
potentially
start
converging.
Bassam
is
also
here
from
the
steering
committee.
Did
some
did
you?
Is
there
anything
you
want
to
add
to
that?
So
far,.
J
No,
mostly
trying
to
I
think
we
should
have
a
discussion
about
which
providers
make
sense
to
converge
kind
of
what
maybe
try
to
get
to
a
bit
of
a
road
map
around
this.
So
I.
Think
opening
up
for
discussion
is
probably
the
right
next
afternoon,
I'd
love
to
hear
her
feedback
from
folks
on.
J
You
know
pains
around
providers
which
ones
need
to
start
either
getting
on
a
convergence
path
or
staying,
or
you
know
how.
How
does
the
community
feel
around
us.
F
Yeah
I'd
like
to
understand
a
little
bit
better
like
the
now.
You
guys
have
the
I
found
like
official
AWS
gcp
and
whatever
providers,
how
do
those
compare
to
the
community
cross
plane
and
enter
a
jet
ones
or
for
those
similar
things
like
for
AWS,
for
gcp
and
and
any
others
that
you
may
provide
in
the
future.
J
I'm
happy
to
take
that
one,
maybe
more
and
then
others
could
speak
on
also,
but
at
really
high
level
I
think
we,
we
built
the
three
official
providers
using
using
object,
which
is
a
successor
to
chair,
Jet
and
effectively
they're
the
most
mature
I
think
and
have
the
most
coverage
in
terms
of
the
three
major
Cloud
providers-
and
you
know
we,
we
upbound
here
in
this
case
our
not
maintaining
these
providers,
they're
open
source,
they're,
Apache
too
licensed-
and
we
also
you
know,
have
them
deployed
in
customer
sites
and
we
provide
Enterprise
slas
and
support
on
them.
J
So
you
know
we
wanted
to
make
sure
that
there
is
providers
that
have
that
are,
you
know,
are
high
quality
enterprise,
Enterprise
grade
providers
that
are
around
crossbones
and
you
know,
there's
as
I
said,
this
is
a
successor
to
church
it.
We
think
the
same.
J
It's
a
good
time
right
now
to
start
deprecating,
the
terrajet
project
and
and
moving
on
from
from
Terror
Jets,
and
you
know
working
with
the
community
and
trying
to
if
you
have
a
terror
jet
provider,
we
can
work
with
you
on
trying
to
convert
it
to
objects
and
and
also
figure
out
where
it
lives,
and
you
know
from
a
maintenance
standpoint
and
whether
it
should
become
something
that
can
turn
into
an
official
provider
Etc
so
so
feel
free
to
reach
out,
but
really
high
level.
I
feel
like
everybody
benefits.
J
If
in
this
community
benefits,
if
there
are
less
providers,
less
options
for
each
vendor
again
from
a
cross-plane
community
standpoint,
people
there
there
could.
There
could
be
good
reasons
to
have
multiple
providers.
J
There
could
be
different
code
generation
approaches
and
those
are
good
reasons,
but
if
there
is
an
opportunity
to
converge
things
that
look
similar
I,
think
everybody
benefits
so
things
starting
with
convert,
starting
with
deprecating,
Terror
Jet
and
trying
to
go
just
object
is
a
good
first
step,
there's
probably
an
opportunity
to
do
more
in
terms
of
convergence,
but
we
need
all
your
help
here.
F
J
Yeah,
the
the
ones
that
are
Terror
jet
based
I
think
that
that
seems
like
a
a
really
easy
one,
because
you
know
object-based
providers
are
just
so
much
better
and
they're
also
open
source.
So
it
feels
like
that's
an
easy
one
to
go
to.
Obviously
we
need
to
work
with
maintainers
and
folks
that
are
have
terror,
jet
providers
or
started
them.
We
should
work
with
them
and
trying
to
do
that.
J
J
Things
like
the
provided,
AWS,
that's
in
crossbank
and
trip
has
been
there
for
a
while,
it's
probably
the
first
provider
on
the
crosswind,
and
you
know
it's
in
production
deployments
as
well,
so
that
might
be
a
might
be
a
longer
path
to
get
attempt
any
kind
of
convergence
and
but
ideally
again,
if
if
we
can
arrive
at
less
providers,
maybe
ideally
one
for
each
vendor
or
infrastructure,
I
think
everybody
benefits.
J
A
F
G
Yeah
the
question
or
comment
I
had
as
far
as
providers
go
is
I
think
there
might
be
a
need
for
a
mapping
of
provider
versions
to
cross-plane
version
compatibility
or
some
sort
of
file
or
something
defined
within
the
provider
repositories
that
provides
that
information
in
particular
I,
don't
know
it's
totally
possible,
as
a
user
I'm
doing
something
wrong,
but
I've
run
into
using
cross
plane,
190
in
particular,
trying
to
go
to
provider
AWS
version,
one
I
want
to
say
it's
one,
eight
sorry
I
thought
I
had
it
up
here.
G
It
does
not
work
backing
down
a
version
to
version
3.0
30.0
works.
Fine,
you
know
same
same
thing,
just
change
inversions,
so
I
I
think
it's
a
versioning
issue,
but
I
could
be
wrong
so
anyway,
that
could
resolve.
Maybe
some
confusion.
J
I
think
that's
a
great
idea
and
maybe
maybe
a
feature
in
the
model
in
the
marketplace
as
well
to
show
compatibility
of
providers
with
yes.
B
That
that
would
be
yeah
yeah,
just
just
so,
you
know,
like
you
know,
the
package
metadata
does
include,
like
you
know,
like
providers
are
able
to
declare
like
you
know,
hey
I
require
minimum
cross,
paying
1.9
or
something
so
if,
for
example,
0.31
didn't
work
for
1.9,
that's
usually
probably
a
bug
that
that
we
need
to
fix,
rather
than
like.
You
know
something
intentional
that
we
could
prevent.
A
Yeah
good
good
call
out
on
that
cross-plane
metadata
for
each
package
does
have
that.
You
know
at
least
the
author
of
the
package's
intent
captured
there
for
compatibility
with
crossplay
versions.
So
that's
good
call
out
on
that
awesome
sphere.
It
looks
like
your
hand,
is
also
raised.
H
Hello
well
thanks
for
answering
well,
the
question
I
had
for
joining
this
meeting
about
the
future
of
terracet.
But
do
you
also
have
well
a
link
to
the
new
setup
and
or
we
are,
how
do
we
proceed
to
it's.
J
In
it's
in
the
it's,
it's
in
the
agenda
dark.
You
can
see
a
link
to
objects
and
that's
that's
the
a
good
place
to
start.
I
A
Yeah
in
Milwaukee
is
there
any
particular
you
know
guides
or
docs
there
that
you
would
recommend
for
people
to
get
the
lay
of
the
land
quickly.
A
B
The
page
that
you
just
opened
is
the
best
place
to
start
the
the
works
like
you
know,
in
the
process
of
updating
the
guys
and,
like
you
know,
there's
a
single
PR
that
updates
creation
of
a
new
provider
with
option.
That
includes,
like
you,
know,
the
new
code,
generator
that
generates
crd
documentation,
which
did
not
exist
in
terrajet.
B
G
A
Oh
yeah
I
was
just
gonna
say
that
yeah,
like
you
know,
but
more
figure.
You
would
say
that
like
issues
and
discussions
in
this
repo
is,
is
a
good
place
as
well
to
to
be
surfacing
things
yeah.
B
A
A
A
Awesome,
sorry
for
cutting
you
off.
Do
you
have
more
that
you
wanted
to
mention
on
this
yeah.
H
H
So
I
was
wondering
how
to
proceed
and
is
there
there
are
any
plans
for
Terry
yet,
but
well
that
is
discontinued.
So
I
guess
my
best
bet
is
to
roll
a
new
profiler
based
on
up
yet.
A
Yeah
and
then
the
Investments
made
in
into
object
there.
It's
you
know,
since
it
is
a
successor
of
their
jet,
I
I
have
at
least
some
some
optimism
that
you
know
so
maybe
some
of
the
you
know
overrides
or
configuration
options
that
you're
looking
for
that
weren't
supported
interroject
may
very
well
be
supported
in
up
jet,
so
you
know
I
would
look
at
it
also
as
like.
A
Hopefully-
and
motha
can
speak
more
to
this-
hopefully
not
you
know
rolling
a
new
provider
based
on
object,
but
you
know
being
able
to
kind
of
update
the
build
process
and
the
dependencies
ETC
of
of
the
provider.
You've
been
working
on
to
the
news.
You
know
up
jet
as
it's
it's
generation
Pipeline,
and
what
is
that
is
that
a
reasonable
suggestion.
B
Yeah
yeah
I'm
I'm
working
on
a
on
a
short
guide
that
will,
like
you
know,
guide
you
to
update
to
object.
But
essentially
it's
like
a
lot
of
the
signal
like
functions.
Signatures
and
stuff
are
similar.
So
if
you
have
an
up
and
running
like
you
know
in
use
provider,
I
would
suggest,
like
you
know,
just
switching
the
dependency
to
object
and,
like
you
know,
fix
the
compilation,
completion
errors,
but
still
you
know,
go
through
the
new
Provider
Guide
because
it
also
contains,
like
you
know,
new
features.
B
For
example,
if
you
only
update
the
dependency,
you
won't
get
the
documentation
generation.
But
if
you
go
through
here,
like
you
know,
just
making
sure
you
got
all
the
steps
done
in
in
that
provider
that
you're
working
on
you
will
get
to
use
options
with
all
the
features
that
it
provides.
H
A
All
right,
yeah
thanks
thanks
for
the
questions
there
Okay,
so
we've
got
about
five
minutes
left
in
the
agenda
here.
There
are
a
couple
more
items
on
the
list,
so
you
know
I'd
say
that
this
is
you
know.
Definitely
you
know
an
an
ongoing
conversation
for
convergence
on
providers.
Right
I
think
you
know
we've
kind
of
introduced
the
idea
here.
We've
got
some
conversations,
some
discussion,
some
feedback
going,
so
we
could
definitely
continue
this
going.
You
know
especially
whisper
maintainers
providers
as
well.
A
I
think,
that's
you
know
listening
and
weighing
in,
and
you
know,
figure
out
a
good
path
together
is
definitely
something
that
we
can
spend
time
on
foreign.
A
Interest
then
real,
quick
Bob.
You
added
this
issue
here.
C
Yeah
I
mean
this:
this
may
be
an
issue
that's
specific
to
to
provider
kubernetes,
but
I've
run
into
it.
Now
that
we're
we're
basically
we're
trying
to
deploy
entire
cross-playing
clusters,
which
will
then
be
used
to
deploy
other
things
right,
and
we
have
custom
resources
that
we
need
provider
kubernetes
to
be
able
to
manage,
and
so,
therefore
we
need
to
create
a
cluster
role
for
provider
kubernetes
to
be
able
to
manage
those
resources,
but
because
there's
no
role
aggregation
and
the
name
of
the
provider.
Kubernetes
service
account
changes
with
every
release
of
Provider
kubernetes.
C
It's
really
hard
to
make
that
declarative
and
have
something
automated
that
will
always
assign
our
custom
cluster
role
to
whatever
the
current
provider
kubernetes
service
account
is
so
I
mean
this
is
you
know
this
I
think
is
gonna
I
was
planning
to
talk
with
Nick
next
week
on
this,
and
you
know
I
just
wanted
to
raise
it
and
say
hey.
This
is
something
that's
kind
of
blocking
us
from.
C
You
know
deploying
cross
playing
automatically
so
anyway,.
B
Yeah
so
I
think
the
like,
probably
kubernetes,
targets
like
in
in
theory.
It
targets
an
external
API
outside
of
the
cluster,
so
it
usually
requires
a
coupe
config,
so
I
think,
like
you
know,
it
could
be
like
in
a
debate
that
it
works
or,
like
you
know,
as
a
workaround
for
you,
that
you
can
produce
a
cube
config
for
an
arbitrary
service,
account
that
you
have
in
your
cluster
and
then
provide
that
as
provider
config
secret
to
the
provider
kubernetes
so
that
it
doesn't
have
to
rely
on.
B
C
A
That
makes
sense
as
well
to
me
thanks
for
explaining
that
perfect,
okay,
great
and
then
Bob
did
you
also
add
this
one
as
well:
yeah.
C
C
You
know
I
think
it
would
be
good
to
have
a
document
or
or
obviously,
first
a
discussion,
and
maybe
some
some
you
know,
collect
some
information
around.
What
does
high
availability
cross-plane
look
like?
How
does
it
work?
What
do
you
do
to
configure
it?
C
C
You
know,
we've
been
I've
been
answering
some
of
the
or
just
having
some
of
those
discussions
and
I.
Just
think
it's
you
know,
I,
don't
have
all
the
answers
to
that.
So
it'd
be
interesting
to
you
know,
have
those
discussions
and
maybe
get
a
document
put
together
on
what
it
looks
like
one
of
the
things
that
I
thought
was
interesting
was
the,
and
this
is
a
I
think
a
kubernetes
issue
more
than
a
cross-plane
issue.
C
But
when
you
set
leader
election
and
set
your
replicas
to
more
than
one,
you
are
getting
extra
controllers
running
but
they're
not
doing
anything.
C
You
know
it's
active
standby,
it's
not
active
active,
so
you're
not
gaining
any
capacity.
You're,
just
gaining
backup,
basically
and
I,
can
definitely
see
where
at
some
point
we
may
need
to
gain
capacity,
and
it
would
be
nice
to
be
running
in
an
active,
active
configuration.
A
Yeah
Bob
yeah,
that's
a
really
good
suggestion
and
that's
something
that
came
up
with
the
AWS
folks
that
the
their
Symposium
they
had
run
is
that
you
know
there
is
the
desire
within
the
Greater
Community
here
for
having
more
guidance
on
these
architectures
that
you
know,
help
across
may
cross,
blade
more.
You
know
scaled
or
secure
or
reliable
whatever
it
may
be
right.
So
yeah,
that's
definitely
an
interesting
topic,
and
you
know
one
that
you
know
some
contributions
to
the
docs.
I
Hip-Hop,
if
you're
interested
I
can
send
you
the
Google
Doc,
and
we
have
with
our
customers
and
I.
I
We,
you
know
we're
talking
to
our
customers
on,
you
know,
Dr
and
all
that
is
definitely
a
topic
that
comes
up
really
often
so
yeah
I
will
send
you
the
link
after
the
after
the
school.
Oh.
J
Hey
Malibu,
can
you
can
you
show
that
doc,
maybe
more
broadly,
maybe
you
can
become
part
of
the
documentation
on
on
crosswind
at
Pete's,
working
on
yeah.
I
Sure
yeah
sure
cool.
A
Yeah
awesome,
but
thanks
for
sharing
that
dude
cool,
okay,
great
so
I
think
that's
the
entire
agenda
Doc
and
we
have
10
seconds
left
by
my
clock.
So
I
think
this
might
be
a
good
time
to
go
ahead
and
adjourn.
Then,
thanks
for
all
the
discussion
today,
it's
really
good
to
see
everyone
turn
out
the
meeting
here
and
great
questions.
A
You
know
looking
forward
to
you
know
as
we're
talking
about
accelerating
the
community
efforts
here
with
a
lot
of
engineering
resources
that
are
going
into
it
from
across
a
lot
of
different
companies
too,
so
very
excited
about
the
the
near
future
and
the
long
future.
Here
with
everybody.