►
From YouTube: Panel: State of the Operators: Hubs, Frameworks, SDKs, a... Diane Mueller, Sonam Saxena & Andre Tost
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
Panel: State of the Operators: Hubs, Frameworks, SDKs, and Beyond - Diane Mueller, Red Hat, Sonam Saxena, Google, & Andre Tost, IBM
Kubernetes developers have embraced the Operator Pattern and there are now many production-ready operators available via a number of hubs and marketplaces. Operators allow developers to create some powerful tools that can be used on any infrastructure where Kubernetes is installed. Operators take all of that knowledge about an application’s lifecycle that a DevOps team practices manually and systematize it. Operators can be used to provision multiple applications in a consistent manner while adhering to best practices for that particular installation. To discuss the future of the operator pattern and why are operators have become so valuable to organizations working with containers, we've assembled some key project leads, engineers, and builders of operators to help demystify the concepts, share lessons learned and best practices for building operators and grow the operator community.
https://sched.co/UadO
A
Let's
start
alright
well
welcome,
and
yes,
there
were
a
lot
of
operator
talks
this
week
so
far
and
there's
probably
a
few
more
tucked
into
the
end
of
today.
So
I'm
really
pleased
that
you
all
came
so
keep
coming
in
if
you
have
an
empty
seat
next
to
you,
raise
your
hand
and
meet
somebody
new,
so
please
keep
coming
in
the
back.
A
Please
do
and
we'll
talk
a
little
bit
about
that
and
other
hubs
today,
as
well.
So
kubernetes
developers
and
service
providers
and
hosting
providers
have
really
embraced
the
operator
pattern
and
since
core
OS
was
2016,
you
know
sort
of
put
out
the
first
blog
post
about
the
operator
pattern.
There's
sort
of
been
an
explosion
of
people,
embracing
that
pattern
and
I
say
pattern
specifically,
because
there
are
lots
of
ways
to
do
it.
A
Sort
of
like
when
you're
sewing
lots
of
ways
to
make
different
dresses,
and
one
of
the
things
that
we're
trying
really
to
tease
out
in
this
conversation
is
what
is
the
future
of
operators.
Where
is
it
going
and
what
are
all
these
frameworks
kind
of
thing?
And
so
there
are
many
production-ready
operators
now
in
operator
hub
in
other
places
and
github
repos.
A
Thank
you
and
I'm
sure
people
will
have
their
interpretations
as
well,
but
they
really
allow
developers
to
create
some
really
powerful
tools
that
can
be
used
on
any
infrastructure
on
any
kubernetes
infrastructure
wherever
kubernetes
is
installed,
and
so,
though,
I
may
be
at
Red,
Hat,
there's
a
reason
why
I'm
wearing
a
neutral,
colored
hat
today,
I
really
have
a
great
enthusiasm
for
the
ability
to
discover
and
find
operators
of
any
ilk
that
can
be
run
on
any
kubernetes
platform.
So
I
don't
want
to
talk
that
much
today.
I
want
to
ask
some
questions.
A
We
have
a
really
interesting.
What
I
tried
to
do
today
was
gather
people
who
have
built
operators,
people
who
are
helping
to
build
the
operator
frameworks,
at
least
one
of
them
or
two
of
them,
and
to
get
some
a
little
history
going
here.
So
I'm
going
to
ask
each
of
the
panelists
to
introduce
yourself,
starting
with
Andre
here
and
a
little
bit
about
your
background
and
your
role
in
building
operators
and
where
you're
at
and
Andre
toast
comes
from
IBM.
B
A
B
We
do
this
in
a
similar
kind
of
fashion,
working
very
closely
with
operator
framework
team
very
active
in
that
community.
We
try
to
do
all
the
operator
work
in
the
open.
Where
we
stand
right
now
is
we
have
a
handful
of
them
built
we
use,
so
we
have
some
early
adopters.
We
have
many
more
to
come.
I
keep
saying
that
next
year
will
be
will
be,
bringing
a
wave
of
new
operators
into
the
community,
basically
representing
like
set
the
entire
software
portfolio
that
we
have.
C
So
operators
has
been
a
big
part
of
the
work
that
I've
been
doing
over
the
past
year
or
so
in
understanding
how
we
can
utilize
operators
to
make
the
experience
of
deploying
applications
built
using
our
software
easier
for
our
our
users
and
our
customers.
I
also
spend
a
lot
of
time.
Writing
operators
myself.
C
A
C
D
Should
work
hi,
my
name
is
so:
I
have
a
product
manager
at
Google
cloud
and
my
team
works
on
a
product
which
essentially
intends
to
make
the
entire
surface
of
GCP
kubernetes
like
so
as
I
was
saying
just
a
few
seconds
ago
to
this
group
of
people.
You
can
basically
install
config
actor
inside
any
communities
cluster
and
then
all
you
will
need
to
do
is
quit
will
apply
a
config
dot
llamó
and
now
you
can
manage
all
your
DCP
resources
as
if
they're
Native
communities,
resources.
E
E
Work
at
red
hat
in
the
operator,
sdk
team
I
came
to
Red
Hat
from
crunchy
data,
where
I
wrote.
The
Postgres
operator
started
that
three
years
ago,
kind
of
at
their
very
early
stages
of
when
operators
came
about.
So
it's
definitely
interesting
to
see
the
the
level
of
interest
in
the
amount
of
tooling.
That's
that's
coming
about
right
now,
yeah.
A
A
The
original
definition
I
got
off
a
core
OS
website,
which
is
still
up
and
running,
is
that
an
operator
is
an
application-specific
controller
that
extends
the
kubernetes
api
to
create,
configure
and
manage
instances
of
complex
stateful
applications
on
behalf
of
a
kubernetes
user
and
the
reason
I
really
like
having
jeff
here
is
that
he's
been
doing
operators
since
the
pattern
sort
of
first
emerged
and
I'm
wondering
starting
with
jeff.
If,
if
you
think
that
definition
still
holds,
you
know,
there's
a
lot
of
debate
if
anyone's
on
any
of
those
threads
I
apologize
but
go
ahead.
A
E
Think
so
I
know
whenever
I
started
this
it
was
my
focus
was
around.
You
know
how
do
I
manage
a
database
on
a
platform
like
kubernetes,
so
it
was
kind
of
infrastructure
or
foundational
frame
things.
People
would
need
to
build
higher
level
applications,
and
you
know
to
two
years
heads
down
building
that
that
was
kind
of
my
my
thought
process
is
that,
well
that
these
operators
concept
is
really
for
big
infrastructural
type
things,
and
you
know
application
developers
might
need
I.
Think
now.
E
A
Good.
Look
at
that
consensus,
I'm
a
consensus
builder,
not
an
instigator,
so
the
other.
The
other
thing
that
we
hear
a
lot
about
operators
is
and
operator,
life
cycles
and
management
is
their
ability
to
help
with
the
day
to
stuff
so
going
beyond
the
simple
install
of
your
application
or
your
service
or
your
database
or
whatever
it
is
that
you've
built
and
I'm
wondering.
How
do
you
see
for
me
and
I'm
opinionated?
That
is,
that
is
one
of
the
bigger
benefits
of
the
operator
panel
pattern
and
I
know.
A
D
The
way
we
look
at
you
know
at
krm
in
general
is
that
almost
every
industry
went
through
a
stage
of
evolution.
So,
if
you
think
about
the
energy
industry,
you
know
we
had
steam
as
a
competing
standard
for
transmitting
energy,
but
now
we
all
agree
that
Alexa
T
is
the
right
method
and
then
there
are
subsequent
industries
lower
from
that
that
are
now
adopting
it.
So
you
are
seeing
this
entire.
You
know
electric
car
revolution
that
is
happening
right
now.
D
The
same
is
true
for
for
communities
once
you
agree
that
krm
is
the
right
language
to
use
across
your
hybrid
cloud
environment.
That
makes
a
big
difference.
So,
as
a
DevOps
practitioner
now
you
can
define
everything
in
one
place
and
you
just
keep
it'll
apply
and
things
are
managed.
There's
separation
of
concern
from
operation
websites,
perspectivism
somebody
goes
in
and
you
know
frat
finger.
Did
it
spanner?
D
It
magically
comes
back
up
again,
like
how
cool
is
that
so
from
a
DevOps
engineer
perspective,
if
you
are,
if
you
are
considering
like
you
know,
thinking
about
where
DevOps
is
going
to
be
in
a
couple
of
years,
we
believe
that
being
all
in
on
on
community
is
being
all
in
on.
Krm
is
the
right
thing
to
do
so.
A
In
terms
of
one
other
question
somebody
asked
me
to
ask
is-
and
this
maybe
I'll
start
with
light
Bend
and
everybody
else's,
because
most
of
you
have
written
an
operator
in
whatever
framework
or
whatever
style
of
thing,
but
why
are
operators
so
hard
to
build?
I
know
you've
had
some
experience
at
like
than
building
yours
or
are
they
hard
or
it
is
it
worth
it?
From
light
bends
perspective
to
a
built-in
operator,
I.
C
Think
so
operators
are
our
pattern.
They're,
not
it's
not
like
a
a
framework
that
guides
your
hand,
so
you
you've
got
a
there.
There
is
the
there's
first
I
think
a
mentor,
a
mentor
like
that.
You
need
to
make
to
understand
what
the
pattern
is
at
an
abstract
level
and
then
to
connect
that
with
how
that's
actually
going
to
be
useful
for
you.
A
This
I
mean
we.
We
now
need
to
write
an
SDK
for
Scala
or
Orion
yep.
Why
not?
Why?
Not
all
right
I
got
a
request
for
a
Java
one
recently,
so
I'm
like
okay
great
so
there
there
are.
There
are
some
Java
ones
too
so
and
there's
a
PI.
But
someone
sent
me
a
link
to
a
Python
kit
just
yesterday.
So
there's
there's
a
lot
of
work.
A
That's
going
on
creating
this
scaffolding
that
you
need
to
create
an
operator
from
whatever
language
or
tooling
that
you,
like
so
I,
know
Andre,
Andre,
I'll,
admit
it
I'm
at
Red,
Hat
he's
at
IBM.
We
now
work
together,
sort
of
at
least
on
slack
and
iBM-
has
a
long
history
using
helm
and
thousands
of
helm
charts
all
across
that
and
I'm
wondering,
as
you
see
the
operators
evolving,
how
the
helm
side
of
things
and
the
operators
side
of
things
come
together
and
how
those
patterns
work
and
where
you
see
that
that
going
yeah.
B
B
How
do
we
lay
down
the
resources
into
communities
and
and
when
we
chose
helm
and-
and
so
we
kind
of
you
know,
created
all
these
charts
and
we
now
have
like
you
said
we
have
I,
don't
know
if
it's
thousands,
but
we
have
hundreds
of
them
and
that
served
us
very
well,
but
I
I
feel
like
so
first
one.
It
challenges
that
there
seems
to
be
this
kind
of
religious
debate
about
all
this
right.
There.
B
Some
people
are
saying
helm
is
dead
and
helm
is
bad,
it's
not
secure
and,
and
you
need
to
do,
operators
but
then
I,
I'd
echo.
What
he
just
said,
just
saying
you
have
an
operator
means
nothing
right,
because
that
doesn't
tell
you
that's
a
good
operate
or
if
it
does
anything
useful,
because
it's
really
just
a
pattern
right
so
I
feel
like
there's
a
bit
of
this
hype.
Now,
where
everyone
throws
this
name
ranting
like
I,
did
an
operator
like?
B
So
there's
a
bit
of
this
religious
debate
to
that,
but
but
I
feel
like
we've
hit
the
ceiling,
if
you
will
with
helm
in
that,
it's
really
just
doing
the
initial
install
for
us,
maybe
upgrade,
and
even
though
upgrade
even
with
helm,
has
its
challenges
right.
So
so
now
we
can
do
install
right,
I
think
the
real
value
that
we
need
to
get
because
a
question
that
we
get
asked
all
the
time
is
like.
Why
should
I
run
my
software
on
kubernetes?
B
So
I
need
something
to
manage
that
and
that's
where
operators
play
a
role
so
I
feel
the
value
to
me
for
operators
comes
in
kind
of
the
later
stages
and
that's
something
that
helm
doesn't
do
for
us
right,
and
so
that's
why
we've
decided
to
kind
of
transition
towards
that
operator
model
away
from
helm.
If
you
will,
even
though
that's
another
discussion
that
we've
had
is,
if
I
now
create
an
operator
that
install
software,
then
I
have
okay
now
I
have
instead
of
a
hound
chart
now
I
have
an
operator.
B
A
B
Capability
model:
okay,
if
you're
not
familiar
with
it,
you
should
look
it
up
is
because
I
I
think
it's
an
easy
guideline
to
basically
say:
okay,
I
have
an
operator.
What
does
it
do
and
and
there's
like
the
first
level
says
I
install
second
level
is
I.
Do
upgrades
third
level
is
I,
do
backup,
restore
fourth
and
fifth,
are
more
going
into
operational
monitoring
logging,
auto
scaling
those
kinds
of
things
and
and
I
keep
saying.
B
A
B
And
it
depends
on
who
you
talk
to
right,
I
mean
you
just
ask
the
question
about
what
language
do
you
write
things
in?
We
have
we
have
teams
where
that
it's
it's
the
Installer
team.
That's
writing
this
logic
right
and
they're
good
at
scripting,
because
that's
how
they've
been
writing
install
logic
right,
they're,
not
necessarily
developers,
so
they
would
say,
go
I,
don't
know
goal
I
haven't
done
that
before
right.
B
C
B
C
What
operators
gives
in
those
first
few
levels
is,
is
I,
guess
a
common
language,
so
this
organization
is,
is
using
operators
just
to
deploy
software,
and
this
organization
is
using
not
registered
to
to.
You
know,
implement
some
common
patterns
specific
to
the
organization.
Then,
when
you
move
from
one
to
the
other,
you've
got
this
common
language
that
things
are
expressed
in
which
I
think
is
is
a
lot
of
value
or
as
far
as
the
you
know,
the
more
higher-level
end-users
implementing
operators
concerned.
A
D
Like
to
pick
up
on
something
that
undressing,
which
is
quality
of
the
operator,
I,
think
that
is
going
to
be
super
critical
as
customers
start
to
use
it,
they
start
to
get
to
day
to
day
three.
They
start
to
actually
use
those
operator
inside
the
operator
hub
or
wherever
they
are
using
it.
If,
if
they
find
that
you
are
still
doing
things
the
old
way-
and
there
are
no
advantages
of
doing
it,
then
there
will
be
questions
that
will
be
asked.
A
Think
the
other
thing
that
at
coop
con
this
week,
I
think
I've
been
in
like
a
dozen
hallway
conversations
about
different
toolings
and
different
frameworks
and
and
different
approaches
to
things
and
I
know.
I
have
operator
hub
do
up
here
because
that's
where
all
of
my
references
are
and
I
don't
do
slides
so
I
know
where
things
are,
but
there
are
lots
of
different
approaches
to
building
operators
and
different
frameworks
and
going
forward.
Do
you
see
some
sort
of
unit
unification
or
standardization
in
the
tooling?
E
Since
I'm
in
the
operator,
SDK
team
I
definitely
would
say,
there's
a
lot
of
alignment
going
on
with
different
projects
like
cube
builder
controller
runtime
I
think
that's
completely
necessary
to
get
a
stable
foundation,
and
once
you
have
a
stable
foundation
that
that
works
people
agree
on
what
excites
me
is
that
level
above
it
that
uses
that
foundation
I.
Think
then
you
can
really
see
some
exciting,
tooling
going
on
and
frameworks
that
really.
B
E
Know
drive
operator,
development
and
adoption
you
know
way
bigger
than
it
is
today
in
supporting
multiple
languages,
things
like
dsls
and
things
so
that
people
aren't
having
the
right.
You
know
miles
are
going
code
at
a
very
low
level
themselves,
but
you
know
they
can
base
new
tooling
on
these
foundational
things.
So
I
totally
think
they.
The
lining
of
some
of
these
tools
and
frameworks
is
there's
a
must
to
push
this
to
the
next
level.
A
Just
all
nods
here,
yes,
and
for
those
of
you
who
don't
know
the
operator
framework,
which
is
in
github,
slash
operator
framework.
If
you
go
over
to
the
T
of
the
CNCs
TOC
on
December
3rd,
which
is
my
birthday,
everybody
is
there'll,
be
a
TOC
meeting
and
we
are
working
on
getting
the
operator
framework
itself
donated
to
the
CNC
F,
so
that
it's
out
in
the
community
and
we
can
all
collaborate
on
it
and
extend
it
as
we
need
to
not
that
that
is
the
undo,
Len
be-all
of
frameworks.
But
it's
it's
coming.
A
So
just
so
you
know
it
would
have
come
faster,
but
dates
and
times
being
what
they
are.
So
one
of
the
things
that
I
wanted
to
ask,
because
we
have
a
number
of
software
vendors
on
here-
is
our
operators
just
prevent
software
vendors?
So
you
know,
and
how
do
you
see,
enterprises
taking
advantage
of
operators
themselves
behind
their
firewalls
anybody
III.
C
Touched
on
this
before
we,
we
have
customers
that
are
writing
operators
behind
their
firewalls
and
and
and
and
using
them,
and
it's
really
about
encoding
their
own
specific
specific
to
their
industry,
verticals
specific
to
their
domain
requirements
in
in
what
needs
to
be
done
to
deploy
and
maintain
an
a
an
application,
so
I
also
so
I
think
as
a
lot
of
operators.
Today
are
you
have
an
existing
piece
of
software
and
the
you?
You
then
have
an
operator
that
that
is
able
to
expose
its
configuration
deployment
maintenance
through
through
the
kubernetes
api
I,
think
invidual.
C
A
The
the
other
thing
that
I'm
seeing
anecdotally,
because
it's
hard
to
see
people's
github
repos
when
they're
behind
firewalls
is
this
idea
of
composite
operators
and
trying
to
create
baseline
operators,
say
for
Kafka
or
whatever
it
is,
but
being
able
to
so
that
they're
not
reinventing
the
wheel
and
I
think
one
of
the
ways
that
we
can
help
enterprises
is
so
that
they
don't
write
things
from
scratch.
They
can
take
these
and
compose
them.
So
that's
that's
I!
A
A
B
Like
we
see
this
everywhere
right,
there's
usually
a
you
know,
a
single
piece
of
software.
You
can
deploy
that
and
manage
that
using
an
operator
that
that
again,
you
know
that
the
value
that
added
value
starts
happening
when
you
put
that
in
the
context
of
a
larger
application,
so
there's
always
multiple
pieces
that
need
to
be
orchestrated
together
and
so
I
think
everywhere
we
look.
We
find
this
pattern
of
operators
of
operators,
you
know
whatever
you
might
want
to
call
that
and
I
think
there
is.
B
So
that's
one
of
the
things
I'd
love
to
see
the
SDK
team
kind
of
tackle
and
have
more
support
for
how
do
I
assemble
these
things
right?
How
do
I?
How
do
I
a
granade
operators
into
new
operators
and
there's
some?
Is
there
some
tooling
that
we
can
that
we
can
get
to
make
that
happen?
So
I
think
there's
a.
E
Gonna
say
operator,
you
know,
composition,
discovery
binding,
you
know
between
operator
is
a
really
big
topic.
I
think
it's
I
think
that's
gonna
be
kind
of
the
next
frontier
of
where
you
start
seeing.
You
know
interoperability
and
people
actually
trying
to
orchestrate
operators
together
to
do
larger
grain
tasks.
I've
heard
a
lot
about
that
at
this
conference
and
there's
a
lot
of
interest
there
and
I
think
there's
that's
really
exciting
to
see
those
kinds
of
talks
going
on.
Yes,.
B
A
And
moving
those
conversations
forward
as
rapidly
as
we
can
is
really
one
of
my
my
goals.
I
think
the
other
piece
of
the
puzzle
for
things
like
operator
hub
and
whether
it's
helm
hub
or
anything
else
is
automating
the
testing
and
scanning
and
the
building
process
for
these
operators
and
we've
heard
a
little
bit
I
think
later.
Today
we
have
another
meeting,
not
hallway
meeting
that
another
session
talking
about
just
testing
operators
to
make
sure
that
they're
they
have
the
metadata
tags
and
that
where
do
you
see
that
going
in
the
not-too-distant
future?
A
D
Well,
what
we
are
seeing
is
that
there
is
definitely
need
for
that.
So
when
we
have
conversations
with
customers,
they
definitely
want
to
go
out
simulate
via
date
test
even
before
something
is
rolled
out
and
and
in
the
in
the
context
of
DevOps
right
like
as
a
if
you
think
about
the
day
in
the
life
of
a
DevOps
engineer,
if
you're
trying
to
push
out
code.
A
D
D
A
So,
if
anybody's
interested
in
the
testing
side
of
this
stuff
see
me
afterwards,
because
we're
gonna
be
working
on
working
through
some
of
that
and
building
those
guardrails,
so
I'm
told
we
only
have
five
minutes
left.
So
what
I'd
like
to
do
is
like
give
each
of
you
a
chance
to
talk
a
little
bit
about
where
you
see
operators
going
in
the
future.
So
what
is
what
is
the
next
thing
you
need
besides
more
SDKs.
E
I
guess,
from
my
perspective,
it's
just
making
the
technology
as
consumable
as
you
can
so
that
you
know
typical
developer
can
work
with
this
build
applications
based
on
it,
I
think
it's
still
fairly
complex,
I
think
when
you
go
to
write
a
reconcile
function
or
if
you
go
to
try
to
perform
Indian
testing
on
these
things
today
and
deploy
it.
It's
still
pretty
complex.
So
to
me,
I
think
the
future
is
making
this
really
consumable
by
the
mainstream
developers
in
a
variety
of
languages
and
I.
Think
that's!
E
That's
the
next
big
challenge,
I
think
in
order
to
pull
that
off
alignment
between
some
of
these
open-source
tools
and
communities
is
really
going
to
go
a
long
way
and
then
I
think
just
having
people
work
on
these
foundational
elements
and
building
up
it
goes
a
really
long
way.
But
to
me
that's
the
future
of
this
is
getting
it
into
mainstream,
so
people,
you
know
it
can
really
consume
this
and
it's
easy
to
use.
That's
what
I
think
is
the
next
big
challenge
once
the
foundational
pieces
are
there.
I
would.
D
Have
to
agree
with
that
and
I
think.
A
lot
of
this
is
great
I.
Have
these
set
of
resources
that
behave
community
is
like,
and
that
is
great,
but
how
do
you
extend
it
to
everything
that
I
own?
So
if
I
have,
if
I
have
legacy
application,
sort
of
hybrid
applications
or
if
I
have
other
things
that
I
still
have
to
go
and
do
manually
or
in
some
other
console?
E
C
Think,
with
the
will,
there's
still
new
new
patterns
and
patterns
maturing
every
day
in
urban
areas,
so,
for
example,
how
the
status
sub
resource
works
with
with
conditions,
and
things
like
that.
If
you
create
a
new
operator
today
that
those
conditions
aren't
automatically
in
in
your
in
your
there's,
no
way
that
I'm,
aware
of
in
any
of
the
SDKs
to
to
be
able
to
automatically
add
support
for
that,
and
then
you
know
the
next
level
from
that.
C
If
we
start
incurring
those
patterns
into
this,
Tico's
has
been
able
to
bubble
up
these
conditions
through
statuses
of
operators
that
aren't
each
other
and
and
and
a
lot
of
that
higher
level.
Those
higher
level
things
that
that
make
that
will
make
it
really
easy
to
write
operators
to
write
good
operators
that
are
really
is
four,
so
I
think
it's
it's
a
lot
more
higher-level
patterns
in
the
SDKs.
A
C
B
Would
say
it
needs
to
live
up
to
its
height.
There
is
a
lot
of
hype
around
operators,
I
think
there's
not
nearly
enough
like
when
you
go
out
on
the
internet,
there's
not
enough
examples,
tutorials
references,
best
practices,
standards
and
then
to
Lee
I
mean
everything
these
guys
said.
I
feel
like.
We
still
need
to
mature
this,
some
more
so
that
people
have
it
easy
to
build
these
things.
I
still
believe
that
the
knowledge
about
all
this
and
how
to
build
good
operators
is
very
it's
very
isolated.
B
Also
within
enterprises
there's,
you
know,
there's
pockets
of
people
that
are
interested
in
this
technology
and
obviously
here
everyone
is
that's
why
you're
here,
but
I
think
it's
it's
still
not
quite
at
them
at
critical
mass.
Yet
for
the
rest
of
the
world
and
I
think
that's
what
we
need
to
push
for
cool.
A
Well,
I
think
that
is
the
end
of
our
time
for
today,
I'm
going
to
thank
all
of
our
panelists
for
coming
and
Alexander
for
hosting,
and
if
you
guys
are
interested,
if
you
how
many
of
you
have
built
an
operator
all
right,
take
your
hand
down
if
that
operator
is
not
an
operator
hub.
Yet
obviously,
okay
so
come
see
me
afterwards
and
let
me
know
what
it
is.