►
From YouTube: TAG General Meeting - 2022-07-20
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
A
A
B
D
C
B
A
D
D
I
think
it's
not
necessarily
A
lot
of
people
most
likely
showing
up
today,
yeah
Josh
I
mean
I,
think
we
have
like
which
Peak
summer
I
think
it's
the
great
vacation,
not
but
great,
not
great
resignation.
It's
a
good
thing,
though,
is
that
going
forward.
I'll
have
more
time
to
invest
here
as
well,
so
it
was
also
kind
of
kept
a
bit
busy
with
other
tasks
and
I.
Think
yeah.
D
Let's
just
go
over
a
quick
account
and
look
into
some
updates
that
that
we
are
working
on
and
one
thing
I
also
asked
by
the
way.
If
more
people
are
interested
to
join
Violet
UC
mailing
list,
we
should
prepare
something,
but
we
would
have
a
wider
reach
there,
but
realistically
I
think
August
is
not
going
to
be
a
great
month
for
getting
people
involved
in
anything
like
everybody,
after,
like
the
last
two
years,
is
just
enjoying
vacation
whenever
they
can,
which
is
something
totally
understandable.
D
Nevertheless,
okay,
let
me
just
put
up
the
agenda.
I
am
just
having
too
many
tabs
open
right
now,.
C
Yeah
I
they,
the
I,
guess
on
the
general
topic.
I
mean
I've,
been
thinking
about
this
for
a
while,
because
I've
been
you
know,
trying
to
contribute
and
yeah
I
mean
we
should
all
think
about
how
we
can
help
people
contribute
to
cncf.
I
I
feel
like
there's
it's
hard
to
get
either
you're
on
the
talk
and,
obviously
that's
a
big
contribution,
but
other
than
that.
How
can
people
ask
me
this?
Like
and
I?
Don't
know,
so
that's
just
something
putting
out
there.
How
can
people
that
want
to
contribute
effectively
contribute.
D
Yeah
yeah
I
think
it's
still
something
we
need
to
to
figure
out.
I
think
it's.
What
will
definitely
help
like
we
kept
now
a
chance
to
like
Lindsay
wrote
to
submit
for
I,
think
half
a
day
event
for
Detroit
I
think
that's
definitely
going
to
help.
C
D
Yeah
knowingly,
obviously
seeing
cubecon
is
like
massively
booked
but
I
think
it's
still
an
opportunity.
So
let
me
see
what
I
have
to
email.
B
Another
thing
which
I
found
out
yesterday
was
that
we
as
a
tech
would
have
the
possibility
to
to
have
a
boost
and
keep
calm
and
I'm,
not
sure
if
this
makes
sense,
but
I
think
it
would
be.
It
would
be
nice
to
have
some
kind
of
office
hours
or
whatever
so
have
half
of
both
like
like
some
projects.
D
D
If
we
run
demos
of
certain
projects
so
and
if
we're
allowed
to
do
that,
to
be
fair,
because
usually
only
bigger
projects
get
a
booth.
C
B
But
then
I
think
it
would.
If
we
get
one,
it
would
also
make
sense
to
present
some
sandbox
projects
or
whatever,
who
don't
have
a
pool
to
give
them
some
visibility.
D
D
The
request
form
should
be
available
early
next
week.
Text
created
with
an
incubating
projects
have
the
opportunity
to
take
advantage
of
in-person
meetings.
This
benefit
allows
tags
and
projects
to
book.
A
four-hour
meeting
to
discuss
project
related
topics
features
issues
amongst
the
maintainers,
so
a
four
hour
meeting
at
kubecon
would
actually
be
pretty
massive.
D
Think
we
do
I
think
we
do
like
we
could.
We
could
like,
and
it's
not
just
us
I
mean
I-
would
almost
handling
it.
Like
a
maybe
mini
conference.
Even
like
that,
we
have
like
one
presentation,
then
people
can
do
like
birds
of
a
feather
kind
of
talks,
whether
it's
about
multi-tenancy,
whether
it's
operator
best
practices
or
pros
and
cons.
So
I
think
we
can
fill
this
out
four
hours
with
some
very
some
very
interesting
content.
D
So
what
I
will
do?
Thomas
I'll
reply
back
to
Lindsay
that
we
are
interested,
but
in
case
anything
moves
later,
as
I
will
be
on
vacation
to
also
Loop
you
into
that
conversation
that
we
are
can
ensure
that
we
actually
get
registered
for
that
four
hours
and
I
think
we
can
learn
a
lot
about
the
four
hours
event
to
like
run
a
bigger
type
of
event,
going
forward.
D
B
I
think
there's
a
thing
we
should
clarify
for
how
long
we,
how
many
slots
a
pools
could
have,
because
so
in
Europe
that
sometimes
to
to
Project
Share,
won't
move
and
I
think
what
it
takes
is
map.
This
map
also
sends
to
have
one
pools
with
two
or
three
tags.
D
C
B
No
I
think
we
would
apply
as
a
pack
as
as
some
other
picks
would
also
apply,
and
they
would
they
would
try
to
find
out
how,
when
which
Tech
uses
the
boost.
For
for
how
long.
C
I
mean
I'm
happy
to
help.
If
you
all
want
to
include
me,
I
mean
that
I
think
that
all
of
these
things
like,
if
we
couldn't
there
are
more
people
that
would
help
I
think
if
we
could
make
this
more
open,
I,
meaning
even
just
getting
people's
thoughts
on
what
sessions
they
might
present,
what
they
suggest
for
the
booth.
I
don't
know,
I
just
would
love
to
have
more.
It's
so
hard.
Maybe
that's
you
know
we
don't
have
that,
but
maybe
just
open
up
a
GitHub
issue.
You
know
how.
C
D
Yeah
and
I
think
Josh
what
would
also
help
because
we
obviously
got
the
chance
to
submit
the
session
so
Alex
and
I
submitted
the
session
where
I
can
share
the
abstract
and
it's
it's
more
as
Cooperative
delivery.
D
It
will
be
fun
to
get
this
session
actually
done
because
it's
like
harvest
the
entire
Cooperative
delivery
life
cycle.
I
think
we
could
easily
bring
along
another
speaker
as
well,
but
it
might
create
some
interest
so
I'm
sharing
the
I'll
share
the
outline
with
you
as
well.
I'll
post
it
to
you
later.
That's
pretty
much
like
the
101
we
submitted,
but
this
is
like
Cooperative
delivery
by
example.
Almost.
C
I
I'm
Cur
I
gave
a
talk
a
couple
weeks
ago
at
open
source.
Summit
kubernetes
is
your
Cloud
control
plane
and
it
was
about
like
using
the
same
mechanisms,
get
Ops
kind
of
mechanisms
for
all
for
infrastructure,
for
middleware,
for
integration
and
for
apps
I'm
wondering
how,
if
there's
some
overlap,
I,
don't
know
what
you.
D
Guys,
yeah
I
think
yeah.
We
should
at
least
have
a
conversation
if
you
just
submitted
it
so
that
it
is
submitted
I.
Think
if
we
have
another
speaker
there,
that
that
would
not
be
an
issue
honestly,
so
I'm
sharing
it
with
you.
C
C
Comment
so
then,
if
you
have
some
comments,
yes,
sweet,
okay,
I'll
check
this.
D
D
That
that
would
be
interesting
in
many
ways
very
much
interested
in
that
topic.
I
think
we
should
talk
more
like
how
to
do
I.
Think
that's
what
cooperatively
I
think
everybody's
as
I
mentioned
everybody's
struggling
with
Cooperative
delivery,
but
how
to
roll
out
infrastructure
together
with
applications
and
do
this
in
an
orchestrated
way
and
which
roles
operators
place
I
think
important.
D
What
did
you
actually
use?
What
did
you
recommend
people
using
like
a
full-fledged
kubernetes
cluster
for
management
or
something
trimmed
down
like
Tinkerbell
or
bad
idea,
or
something
like
that.
C
C
It
was
about
like
look,
let's
all,
let's,
let's,
instead
of
using,
you
know
cloud
formation
and
terraform,
and
you
know
10
different
mechanisms
for
managing
control,
but
it's
all
Center
on
the
kubernetes
API,
so
I
it
was
just
showing
that
you
could
use
the
API
to
deploy
up.
You
know
using
operators
you
could
deploy.
You
know
an
app
using
Argo
or
using
an
opera
using
a
helm-based
operator.
Now,
basically,
that
was
yeah
krm
kubernetes
resource
model
for
everything.
D
C
Yeah
I
agree.
That
would
be
a
perfect
like
so
the
two
areas
and
that's
multi-tenancy,
so
that's.
In
other
words,
the
reason
people
are
doing
that
is
to,
and
red
hat
is
doing,
that.
There's
a
project
called
kcp,
which
is
also
in
the
Upstream
there's
a
thing
in
the
kcp
prototype
in
the
kubernetes
slack.
C
It
doesn't
stand
for
anything,
but
it
really
stands
for
kubernetes
as
your
control,
plane,
I.
Think
and
it
like
you
just
said
it
doesn't,
have
a
scheduler.
So
it
delegates
the
work
out
to
to
child
clusters
like
v-cluster
does
the
same
thing.
It
doesn't
have
a
scheduler,
so
the
actual
scheduling
of
PODS
happens
in
a
host
cluster
or
V
cluster.
D
C
I'm
not
sure
how
that
works.
I,
yes,
I
know
there's
somehow
there
are
crds
and
these
operators
provides
you
I,
don't
know
how
it
do.
They
have
to
run
them
in
the
children's
summer.
I
don't
know.
That's
a
little
I
haven't
gotten
that
deep.
Yet,
but
yeah
I
hear
you,
you
might
have
to
run
something
in
the
control
plane.
I
need
to
get
the
v-cluster
folks
on
here.
D
Yes,
it
might
actually
be
an
interesting
way
how
to
run
a
control
plane
cluster
for
full
stack
infrastructure
management.
That
might
actually
also
be
a
good
topic
to
have
them
properly
present
with
their
ideas
are
because
I
see
this
pattern.
Kind
of
emerged
like
having
a
control,
plane,
cluster
and
I
overall,
like
the
idea,
but
I
don't
want
to
run
a
super
expensive
cluster
that
has
no
workloads
or
even
prohibit
like
running
workloads
on
it.
They
maybe
could
do
it.
D
I
could
just
restrict
the
cluster
down,
so
nobody
can
use
a
cube
control
apply
against
it,
but
at
the
same
time,
I
can't
just
run
the
API
server
by
itself.
D
C
C
Yeah-
and
this
was
in
the
in
the
conversations
about
here's-
my
blog-
that
I
the
conversation
about
multi-penancy,
which
I
mean
what
you're
saying
is,
is
I,
think
that's
why
I
am
picking
up
on
the
multi-penancy
like
it
seems
like
that
stuff
is
emerging
and
people
are
starting
to
think
like
namespaces
are
not
the
right
are
not
the
right
way
to
break
up
tenants,
or
at
least
their
clusters
have
some
big
advantages.
That's
why
yeah
I
would
love
to
pick
up
this
article
and
talk
about
that
stuff
in
there.
You
know.
D
So
the
draft
text
is
the
actual
article
right.
Oh
this
thing
that
I
just
posted
in
the
other
one
that
was
in
the
meeting
in
the
agenda.
C
C
C
Yes,
okay,
that's
not
the
one
I'm
talking
about
that
is
what
I'm
I'm
talking
about
that
too.
This
is
where
we're
working
as
a
tag
to
put
out
some
opinions.
Nobody's
really
touched
this
for
a
few
weeks
now,
but
Alex
LED,
an
initial
brainstorm,
I've
dumped
in
a
bunch
of
stuff,
I
I,
do
have
some
time
clearing
up
so
I,
maybe
can
hammer
out
some
pros.
You
know
for
us
to
to
build
from
here.
D
So
maybe
not
like
necessarily
get
too
much
into
even
more
like
what
we
want
to
touch
the
topics
while
actually
presenting
Solutions
so
that
we
can
really
get
people
interested
share
on
Twitter
and
other,
usually
Twitter
LinkedIn,
so
that
just
get
more
people
and
like
even
sharing
it
with
a
wider
audience.
C
B
C
A
C
Well,
that's
super
cool
all
right!
Well,
I
didn't
know
that
was
there.
C
Yeah
I
mean
that
this
kind
of,
like
Cooperative
delivery,
I,
think
there's
an
opportunity
like
maybe
something
similar
like
put
a
stake
in
the
ground.
Even
if
it's
like
this
is
not
saying
either
any
of
these
cluster
options.
Are
you
know,
they're,
not
choosing
any
winners?
It's
just
hey.
Let's
start
the
discussion,
maybe
Cooperative
delivery.
We
should
do
something
like
that
too.
You
know
what
what
we
all
envision,
because
I
think
Eloise
I
think
you
and
Alex
might
have
more.
You
know
you
have
some
strong
opinions.
C
D
Yeah,
we
have
also
some
models
like
talk
to
Max
from
liquid,
replied,
they're,
doing
a
lot
of
multi-cluster
and
multi-tenancy
work.
C
D
D
I
mean
the
idea
is
like
how
do
I
ship
more
or
less
full
stack
from
infrastructure
to
app,
including
pretty
much
everything,
and
how
to
do
this
in
an
orchestrated
way,
and
that
was
where,
like
thousand
Thomas
was
involved.
We
will
discuss
lots
of
the
problems
that
that
arise
there
and
let
a
lot
of
the
open
questions.
D
The
good
thing
A
lot
happened
in
between
like
projects
came
along
or
well,
they
were
technically
there,
but
got
more
popular
like
backstage
cross-plane
and
others
that
play
a
a
vital
role
in
this
area.
Right
now,
and
then
we
had
like
some
people
that
joined
that
kind
of
like
were
convincing
everybody.
That
Cooperative
delivery
is
a
good
name.
Yeah.
D
I
think
we
don't
I,
think
that
needs
maybe
not
around
I,
think
we
conceptually
understand
each
other,
but
we
have
I
think
this
is
still
an
issue
to
like.
Do
that,
one
minute
elevator
pitch
or
Cooperative
delivery
that
that
we
would
need
to
do
I
still
think
that
multi-tenancy
perfectly
fits
in
there
and
if
Cooperative
deliberately
delivery
is
only
working
on
this
topic.
It's
fine
and
I
was
like
listing.
D
Projects
for
different
tasks
might
be
also
like
a
nice
like,
as
you
like,
refer
to
like
listing
projects
what
they
can
be
used
for,
like
maybe
putting
this
into
another
blog
post.
Okay,
this
can
you
can
use
for
infrastructure
provisioning.
This
is
what
you
can
use
for
application,
provisioning
and
providing
some
options,
but
you
can
use
as
a
service
catalog.
C
So
I
hear
a
theme
from
us
that
we
want
to
publish
you
know
thought-provoking
articles.
So
maybe
that's
like
that
seems
like
a
thread
here
and
we
should
just
make
sure
that
blog
site
that
you
set
up
Thomas
is
as
we
want
and
then
start
adding
some
stuff
there,
as
as
we
can
I
mean.
D
Well,
I
think
it's
well
thought-provoking
I,
wouldn't
necessarily
say
thought-provoking,
but
I
think
we
get
people
to
participate.
If
we
describe
a
problem
that
they're
having
you
will
get
vendors
always
to
participate,
if
they
believe
they
have
a
solution
and
projects
I
mean
quite
obvious,
but
obviously
also
get
end.
Users
contribute
if
this
is
something
that
they
are
working
with.
D
C
I
mean
I
agree,
I,
don't
get,
I
would
rather
be
published.
More
broadly
I,
don't
I
mean
I'm.
Just
my
blog
has
gotten.
Some
hits.
I
actually
got
some
an
email
from
somebody
from
Loft
who
corrected
something
in
there.
So
he
said
they
were
getting
redirect
over
there,
but
yeah,
please,
let's,
let's
get
it
somewhere
else.
If
you
know
the
goal
it
even
mentions
in
there.
Please
come
to
this
issue
and
yeah.
D
D
Look
at
some
more
visibility
and
I
think
this
is
kind
of
great.
If
you
work
on
these,
these
topics,
like
smaller
blog
posts,.
D
B
D
C
No
I
mean
like
it
seems
like
most
of
these
things.
One
person
writes
up
the
core,
then
gets
a
lot
of
feedback,
maybe
significantly
changes
it,
and
then
we
go
from
there.
So
I,
don't
I
mean
the
main
thing
is
you
know
somebody
pick
one
up
each
person
and
then
you
know
if
they
want
to
talk
in.
You
know
they
could
come
to
these
meetings
every
couple
weeks
if
they
really
want
live
feedback
or.
D
Yes,
what
I've
very
often
seen
is
like
you
start
something,
because
you
have
an
idea
and
you're
kind
of
slightly
too
broad
on
what
you're
working
on.
Then
it
cools
down
entirely
and
then
you
might
narrow
it
again.
Some
somebody
else
has
the
time
to
like
push
it
like
over
that
critical
manner.
So
there's
enough
content
and
then
it
kind
of
keeps
going
so
think
about
like
in
this
natural
cycle.
Right
now
we're
just
apps.
We
know
it
is
important.
That's
why
I
wouldn't
give
it
up
honestly,
but.
D
Just
right
now
like
focusing
on
multi-tenancy,
there's
always
something
there
and
try
to
push
for
this.
It
seems
to
be
more
of
a
topic,
as
you
mentioned,
and
then
make
this.
What
Cooperative
delivery
is
working
on
right
now,
but
for
me
it's
totally
fine.
C
C
Foreign,
oh
yeah,
there
was
one
I
was
gonna,
throw
something
on
you
all
and
see
what
you
thought.
Integration
I
I've
been
thinking
more
and
more
that
it's
that
it's
really
just
integration
in
a
different
form
and
you're,
trying
to
get
your
app
out
there
and
connect
it
to
databases
and
connected
to
identities
and
have
it
running
in
the
right
cluster.
It's
about
in,
like
it's
slightly
different
than
you
know,
integrated
micro
services,
but
it's
integrating
capabilities
and
and
I'm
always
reminded
of
Martin
Fowler.
C
C
Oh
well,
I
don't
know
like
backstage,
can
help
you
provision
15
different
Services,
let's
say,
but
does
it
help
you
actually
integrate
them
all
to
work
together
like.
C
I
mean
it
probably
does
I
mean
I
assume
it's
got
some
functionality,
but
that's.
What's
that's
what
I'm
getting
at
like
it's
we're
trying
to
figure
out.
Okay
I've
got
a
cluster
here,
a
database
here,
a
certificate
management
system.
Here
you
know
a
pod
scheduler
here
how
do
I
as
a
developer,
push
one
button
and
get
that
all
you
know
working
together
as
opposed
to
having
to
talk
to
five
or
ten
different
people
and
coordinate.
You
know
passwords
and
URLs,
and
who
knows
what.
B
Foreign,
that's
one
of
the
issues
which
which
indeed,
the
service
broker
deal
with
yeah.
C
For
sure,
but
that's
kind
of
I
mean
open
service
broker
is
kind
of
dead,
I
think
they
even
deprecated
it
in
kubernetes
the
support
for
it.
C
So,
and
maybe
that
was
that,
but
that's
exactly
what
we
should
write,
we
should
write
like
service
broker.
Was
you
know
an
option
for
this
we've
kind
of,
for
whatever
reason
decided
an
industry
to
move
on
from
that
we've
got
operators.
You
know.
D
Yeah
I
think
that
that's
I
think
I
think
we
should
bring
up
those
topics
and
have
those
discussions
and
even
for
the
time
being,
if
we
just
keep
like
bringing
up
like
how
do
you
handle
this
and
also
not
everything
is
service
provisioning,
especially
talking
multi-tenancy.
So
if
I
want
to
have
a
postgres
database,
I,
usually
don't
want,
like
an
entire
database.
D
Actually,
services,
on
top
of
Action
Service,
like
more
or
less
virtual
type
of
services,
that
I
really
care
about
and
I
don't
care
about
the
underlying
implementation,
because
technically
I
even
don't
care
what
the
thing
is
running.
This
could
be
a
cloud
providers.
The
service
as
well.
C
C
Okay
I
was
putting
that
in
there
so
yeah.
My
takeaway
here
is
I'll.
Try
to
hammer
out
in
the
next
few
weeks.
It's
always
famous
last
words,
but
I
really
have
some
time.
The
next
few
weeks,
I
hope
I
can
hammer
out
an
article
here,
at
least
that's
something
I.
Can
you
know
as
the
Cooperative
delivery
lead
at
least
do
that.
D
C
Like
include
in
there
some
some
breadcrumbs
like
hey,
we
really,
this
is
an
idea.
What
do
you
think
I.
D
Think
a
lot
a
lot
of
what
people
are
doing
in
the
app
delivery
space
is
very
much
custom
made
and
travel
knowledge
these
days
well
with
kubernetes.
Coming
from
this
infrastructure
background,
I
mean
it's
technically
provisioning
services
with
an
infrastructure
Focus,
not
so
much
with
an
application,
Focus
and
MRP
people
who
argue
against
it.
But
if
you
come
like
from
a
pure
application
so
because,
usually
when
application
is
different
and
to
be
fair,
a
lot
of
projects
were
struggling
even.
D
But
this
was
also
by
the
way,
very
interesting
in
sick
apps,
so
I
recently
joined
sick
apps
Mitch
Connors
from
istio
brought
up
looking
into
a
crd
for
Progressive
delivery,
because
there
are
massive
issues
on
the
application
side
and
just
just
paraphrasing
some
some
of
the
topics
there
like
how
can
I
even
Define
application,
health
and
Google
need
this,
because
you
can't
you
can't
Define
it
on
a
pod
level.
You
can
just
find
whether
the
Pod
is
final
or
not,
but
part
health
is
not
application.
D
Health
Portal
smallest
means
it's
ready
to
accept
traffic.
It
doesn't
mean
it's
doing
what
it
is
kind
of
supposed
to
do
like
a
card
service
that
always
returns
zero
items.
But
that's
this
very
reliably
is
still
running,
as
a
part
is
still
fine,
but
as
an
application
is
actually
not
running.
Fine
like
even
defining
this
today
is
not
like
really
possible,
and
then
you
could
think.
Okay,
how
can
you
like
add
health
checks
that
go
beyond
this,
which
then
it's
also
interesting,
because
usually
the
kubernetes
approach
to
an
alternative
part?
D
Is
you
just
like
shut
it
down
and
start
a
new
one
like?
Well?
Let's
throw
it
away
and
try
again,
but
you
can't
do
this
for
an
entire
application.
You
can
just
rip
an
entire
deployment
or,
however,
you
want
just
to
actually
application,
throw
everything
away
and
hope
it's
coming
up
again
so
like
managing
the
application.
Health
I
think
it's
a
very
interesting
concept
as
well.
When
we
talk
about
app
delivery
like
how
do
you
even
know
that
your
kubernetes
application,
after
starting
it
is
healthy.
D
D
It
can't
be
rebuilt
everything
or
wrote
back
everything.
It's
just
one,
failing
service
that
there
must
be
something
more
better
to
do
it
and
the
question
isn't
operated
the
right
answer.
That
would
make
a
really
complex
operator.
C
It
feels
like
we,
we
have
to
better
gather
up
the
application
components
and
then
we
will
be
able
to
aggregate
the
health
of
all
of
the
components.
It
will
be
one
big
red
green
at
the
top,
but
then
break
it
down
into
other
components.
But
right
now
we
don't
have
a
great
way
to
correlate
all
of
the
components
that
participate
in
an
application.
C
D
C
C
Yeah,
it's
always
chicken
and
egg.
The
that
was
actually
one
thing
that
I
was
thinking
about
is:
should
we
like
the
things
like
say,
I
just
need
a
tablespace
or
I
just
need
a
topic
or
I.
Want
you
to
issue
me
a
cert,
you
know:
can
we
do?
Can
we
infer
those
things
or
can
we
suggest
annotations
or
labels
that
you
know
admission
controllers
and
operators
could
queue
off
of
to
you
know
handle
the
integration
that
is
not
there
wondering
yeah
it
doesn't
that
doesn't
address
what
you're
saying,
though,.
D
Yeah
so
I
think
I'm
trying
to
because
I'm
not
starting
to
join
the
sea
gaps
meetings.
Infrequently
it's
a
good
idea,
so
yeah
I
linked
it
so
I
think
it's
an
interesting
concept
and
is
it
an
interesting
problem,
foreign
which
you
start
to
have?
If
you
look
at
something
from
an
application,
Level.
C
D
But
it
might
also
I'm
just
trying
to
find
topics,
also
like
talking
about
engagement,
I,
think
people
are
usually
willing
to
engage
and
to
work
with
you
in
something
if
they
see
that
there
is
actual
value
in
the
day-to-day
work.
If
you
look
at
something
okay,
it's
like
one
of
the
topics
that
we're
addressing.
D
I
feel
like
yeah.
There
is
some
work
to
do.
Then
you
kind
of
get
involved
because
you
want
to
improve
your
own
infrastructure
deployment.
I
think
that's
how
we
can
get
people
working
on
something,
because
otherwise,
if
it's
like
super
abstract,
what
you're
working
on
it,
it
starts
to
get
kind
of
hard.