►
Description
Kubernetes Public Steering Committee Meeting for 20220307
A
B
A
Off
by
noting
that
we
abide
by
the
cncf
code
of
conduct,
so
please
be
excellent
to
each
other.
This
meeting
will
also
be
recorded
and
posted
to
youtube
the
we
have
an
agenda
and
we
have
some
items
on
it.
So
we're
going
to
jump
right
into
those
paris.
I
think
you
have
the
first
one.
C
I
just
wanted
to
give
an
update
that
we
do
have
a
governing
board
meeting
in
april
and
the
process
for
getting
things
on
the
agenda
is
we
have
to
email,
the
ed
priyanka
and
chair
arun
and
get
that
get
our
topics
on
the
agenda
right
now.
I
know
we
are
working
on
a
few
of
those,
but
this
is
just
kind
of
a
reminder.
C
Right
now.
We've
got
topics
that
involve
legal
sustainability
and
some
other
things.
So
our
next
meeting,
I
think
that
will
be
the
meeting
that
will
start
really
finessing
those
so
that
we
can
get
prepared,
but
we
need
to.
I
have
a
meeting
with
arun
and
priyanka
at
the
end
of
the
month,
so
we
just
need
to
get
our
topics
in
by
then.
So
that's
pretty
much
it
so
you'll
hear
more
from
me
in
the
next
coming
days,
so
that
we
can
get
those
squared
away.
A
B
The
paris
are
there,
documents
that
you
are
working
on
or
be
through
it.
I
remember
writing
something
up
last
week.
Yeah
do
you
want
to
get
them
out
in
public
or.
C
A
Okay,
great
there's,
no
other
questions
we'll
move
on
to
the
next
topic,
evaluate
cdns
for
kubernetes
releases
and
other
artifacts.
This
one
is
on
you
dimms.
D
Yeah
sorry
eddie
had
mentioned
in
the
chat
that
he's
he's
jumping
between
meetings,
so
just
give
him
a
shout
before
we
start
this
one
up.
B
Eddie,
are
you
there,
okay,
good,
thank
you
yeah,
so
I
know.
Do
you
want
to
introduce,
or
would
you
like
me
to
introduce?
I
can't
do
that.
E
E
Okay,
hi
everyone,
so
I
opened
the
subject
really
to
artifact
management
in
general.
We
don't
currently
we
the
com,
the
project,
produce
three
type
of
artifacts
binary
system
packages
and
container
images
right
now.
Ownership
is
partially
shared
between
google
and
us
in
the
future.
We
plan
to
migrate
that
to
the
community
infrastructure
and
before
we
do
that,
I
would
like
to
make
sure
those
artifacts
are
destroyed
by
a
third
party,
because
it
costs
us
a
lot
to
distribute
to
be
the
solid
distributor
of
those
artifacts.
E
E
This
is
not
a
new
practice.
Basically,
I
list
a
few
open
open
source
project.
They've
been
doing
this
for
four
years,
I'm
talking
about
the
linux,
the
canon
project,
the
debian
project,
nyx,
os
and
even
alpine
distribution.
So
it's
not
something
new.
We
just
basic.
So
what
I'm
asking
to
steering
is
apply
to
this
program
on
the
behalf
of
the
community
and
we
can
initiate
a
conversation
with
those
cdn
providers.
F
I
just
wanted
us
to
really
make
sure
that
we
set
the
stage
when
talking
to
these
cdn
providers,
so
we
don't
surprise
them
with
a
massive
production
workload
when
they're
generously
offering
open
source
stuff.
B
Yeah
so
I
I
did
open
up
an
issue
against
kubernetes
kubernetes
kubernetes
units,
I
think,
asking
if
we
could
collect
some
stats
on
the
downloads
of
at
least
the
case
binaries,
and
things
like
that,
so
that
that
we
might
need
to
beef
up
some
some
of
things
like
that
to
enumerate
the
costs
where
we
are
facing
costs
and
the
size
of
the
artifacts
and
which
releases
and
are
very
popular.
And
you
know
things
like
that.
I
guess.
E
Just
to
clarify
I'm
not
talking
only
about
the
kk
repo,
I'm
talking
about
every
sub
project
inside
the
key,
the
communities
organization,
because
we
have
some
one
sub
project
that
costly
a
lot
to
dispute.
That.
G
So
yep
my
question
like
I
am,
I'm
all
for
you
know
exploring
these
options.
The
one
thing
is
is
that
both
both
fastly
and
oh
god,
I'm
totally
bringing
the
other
one
right
now,
cloudflare
we'll
use
the
trademark
and
project
name
in
like
they
will
list
on
their
open
source,
page
and
and
elsewhere.
G
B
So
the
plan
what
we
were
trying
to
discuss
there
was
one
of
us,
hopefully
me,
because
I
have
some
context:
we'll
go
talk
to
a
start
with
cloudflare
right
and
then
try
to
get
a
list
of
expectations
from
you
know
between
the
two
of
us.
B
What
would
they
like
to
see
from
us,
and
you
know
what
we
can
offer
and
what
is
the
kind
of
information
that
they
will
ask
us
I'll
point
them
to
gcs
buckets
and
whatnot
to
give
them
an
idea
of
you
know
what
is
involved
in
this
process,
and
you
know
how
much
cost
we
are
paying
and
something
like
that
that
once
I
and
then
I
can
put
together
a
proposal
which
will
go
to
everybody
here
like
it
will
go
to
steering
it
will
go
to
cncf.
B
It
will
go
to
a
sig
release
and
get
sign
offs,
and
then
we
do
anything
after
that.
So
I'm
not
going
to
like
sign
up
for
hey
cloudflare
starting
day
after
tomorrow.
You
know
just
four
things
over
or
anything
like
that
unilaterally,
so
we'll
formulate
a
plan
and
then
we'll
execute
the
plan
go
ahead.
Jordan.
H
That
that
pretty
much
addressed
my
question,
I
saw
stephen
had
tagged
in
sig
release.
I
wanted
to
make
sure
that,
like
I
think
it's
great
for
us
to
sort
of
represent
the
project
to
these
organizations
if
this
is
something
we
want
to
do,
but
getting
the
stats
and
making
sure
with
sig
release
that
like
this
is
what
we
want
to
do,
plus
one
for
getting
that
plan
in
place
and
everyone
kind
of
on
the
same
page
before
we
move
on.
So
that's
like.
D
Yeah
yeah,
I
think
that's
fine,
I
would
say
like
so
I
I
think,
based
on
the
read
through
my
understanding,
is
that
we
have
our.
Maybe
the
cncf
has
pre-existing
relationships
with
fastly.
Does
it
make
sense
to
to
route
through
them
first
or
have
the
conversation
there?
First.
B
Yeah
I'll
talk
to
you
all
and
check
also
before
I
figure
out
which
one
to
go
first
or
maybe
I
could
do
multiple
in
parallel.
Yeah.
E
B
Also
going
into
we'll
have
to
take
into
factor
is
like,
which
is
the
easiest
for
us
to
do
right
like,
for
example,
if
one
of
them
can
work
off
of
the
gcs
buckets,
then
you
know
it's
literally
less
work
on
our
side,
so
we'll
take
technical
considerations.
Also
to
you
know,
and
if
fastly
says,
oh,
you
have
to
jump
through
these
three
five
hoops
and
you
have
to
make
bunch
of
changes
in
image,
promoter
or
whatever,
like
you
know,
just
throwing
it
out
right
then
we
would
say
hey.
B
B
So,
but
the
main
thing
is
like
collecting
as
much
information
from
as
many
of
these
as
possible,
so
that
we
can
come
to
a
you
know,
a
good
plan
and
a
backup
plan
kind
of
thing.
The
other
thing
that
we
were
talking
about
in
the
meeting
was
like
a
small,
reversible
steps
was
one
of
the
philosophy
that
we
were
going
in
for
so
how?
Much
of
this
will
be,
you
know
things
that
we
can
roll
back
and
come
back
to.
B
You
know
the
state
we
are
today
that
will
also
play
some
role
in
the
consideration.
H
H
Switch
over
where
we
sort
of
did
one
step
at
a
time
and
actually
had
to
roll
back
several
times
so
yeah,
avoiding
one
big
bang
type
moves
is
great.
A
I
specifically
have
a
question
around
evaluation
criteria.
It
sounds
like
there's
been
some
thought
put
into
already
from
from
sick
case
and
for
like
what
considerations
are
going
to
go
into
something
like
this
are
those
written
down
somewhere,
because
I
think
those
would
be
a
good
thing
to
have
written
down
prior
to
having
the
conversations
with
you
know,
potential
vendors.
A
I
guess
it
like
I'm
guessing
viewing
this
kind
of
as
a
a
vendor
thing,
even
though
we're
applying
for
a
potentially
zero
like
monetary
cost
program,
but
there
may
be
exchanges
of
other
things
like
you
know,
rights
to
use
the
trademark
and
certain
marketing
materials
and
that
kind
of
stuff.
So
it's
still
like
it
still
feels
like
a
bendery
thing
to
me.
So
I'm
wondering
if
we
have
any
of
the
criteria
written
down
prior
to
having
those
those
conversations
with
cloud
fire,
fastly
or
anyone
else.
B
B
We,
we
haven't
thought
about
those
things.
Yet
you
know
I'm
hoping
that
at
least
cloud
fair
has
you
know,
half
a
dozen
open
source
projects
under
them
and
they
might
already
have
a
set
of
things
they
ask
of
each
project.
So
you
know,
I
think,
we'll
end
up
starting
there.
A
Yeah,
I
think,
that's,
I
think,
that's
good
and
like
the
other
thing
that
just
like
again
clicks
in
with
my
my
mind
around
this
is:
is
there
anything
as
far
as
tie-ins
around
like,
in
particular,
in
what
we
can
offer
question?
A
We
have
like
the
the
infrastructure
credits
program
through
the
cncf,
which
that
may
be
at
least
like
a
good
initial
guideline
of
like
what
things
we
would.
Potentially,
we
would
potentially
even
like
talk
about
on.
A
Yeah,
so
I'm
wondering
like
because
again,
even
if
the,
even
if
the
monetary
cost
is
zero,
if
it's
like
exchanged
through
open
source
like
if
we're
pushing
a
bunch
of
this
cdn
traffic
through
anywhere,
it's
going
to
be
a
not
insignificant
financial
cost
on
whatever
partner
we
we
go
with
or
whatever
partner
would
accept
us
into
that
program
like
this
may
yeah.
I
I
just
see
it
being
very
big
and
lost
wanting
to
understand
what
we're
talking
about
prior
to
necessarily
going
into
those
conversations.
B
Yeah,
though
those
are
definitely
prior
art,
kristoff
the
cloud
credits
program
and
the
you
know,
the
announcements
and
things
like
that
and
the
other
prior
art
also
is
like
you
know
the
google
itself
right,
like
the
three
million
dollars.
So
there
is
some
prior
art
that
we
can
refer
to
when
we
talk
to
them.
B
But
you
know
I'm
hoping
that
I
don't
know
it's.
It's
gonna
be
discovery
phase
at
this
point
right,
so
hippie
hacker
is
not
well
for
you
know
this
since
last
week,
so
I'm
not
I'll
not
be
able
to
talk
to
him.
So
I
I
want
to
initiate
the
conversation
via
an
email
or
filling
up
a
form,
and
then
you
know
see
where
it
goes
from
there.
B
F
Do
we
have
a
place
where
we
wanted
to
start
getting
those
requirements
together,
because
I
know
a
few
things
came
up
for
qualifying
vendors
was:
is
this
where
we
want
to
consider
great
firewall
of
china
right?
That
was
something
on
our
list.
So
do
we
find
a
vendor
that
supports
that
or
yeah.
E
Yeah,
that's
why
I
raised
my
my
hands
just
to
say:
I'm
gonna
open
an
issue
and
trying
to
have
a
basically
basic
list
of
requirements.
We
we
want
to
basically
establish
before
we
start
the
technical
conversation
and
about
period.
I
want
I
put
the
link
of
the
apache
foundation
in
history
and
there
are,
I
think,
five
times
all
size
in
term
of
ingress.
So
I'm
not
really
worried
about
that.
E
E
I
wanted
to
wait
before
do
that,
but
because
last
week
was
perf
review
inside
google,
so
it
was
very
complicated
to
get
people
to
talk
to
people.
So
I
open
the
issue
without
those
information
but
you're
right
crystal,
we
should
basically
collect
the
different
technical
requirements
before
we
start
technical
conversation
with
them.
So
my
my
I
mean
the
issue
is
basically
be
sure
we
start
at
the
conversation.
Even
we
need
to
talk
about
trademarks.
J
H
Don't
wait,
it's
fine
to
be
patient
but
yeah,
throw
it
in
a
queue
and
ping
as
needed,
yeah.
Whatever
random
issues
we
can
open
or
docs
just
link
it
back
to
that
main
one
so
that
there's
a
breadcrumb,
we
can
kind
of
see
everything
that's
going
on
with
us.
A
E
A
And
it's
are
we
looking
at
only
covering
things
that
have
already
moved
over
to
communicator?
A
I
keep
it
covering
things
that
are
still
on
google
infrastructure.
I
would
just
just
to
interject,
so
there's
not
a
lot
of
back
and
forth.
My
thought
would
be.
We
should
cover
what's
in
community
infrastructure
and
that
can
be
an
incentive
for
google
help
us
migrate.
The
things
that
migrate
out
of
google
infrastructure-
I
don't
yeah
that
that's
at
least
my
initial
thought
but
curious.
If,
if
case
infrared
already
have
thoughts
on
that
subject,.
E
So
you
you're
absolutely
right
about
this
basic.
The
my
first,
let's
say
sub
project
I
want
to
try
with
is
cops
right
now.
Cops
distribute
use,
20k,
22
000
months,
just
for
distribution
inside
the
community
infrastructure
per
month,
and
basically
this
is
a
course.
I
would
like
to
basically
delegate
to
a
third
party,
because
it
will
help
us
get
some
room
to
finish
some
other
migration
step.
E
So,
basically,
if
we
can,
if
we
succeed
to
use
a
syrian
provider,
we,
for
example,
co-ops
and
basically
have
in
observation
periods
over
two
to
three
months.
We
can
start
the
conversation
with
everything
inside
google,
I'm
talking
about
system
packages
and
binaries,
so
we
have
an
idea
of
what
is
cost
us
if
we
want
to
use
a
cloud
provider
and,
what's
the
let's
say,
world
back
plan
for
this
initiative,.
D
So
I
mean
I
so
so
one
I.
I
think
that
you
know
if
we,
if
we
don't
have,
I
think
we
should
get
an
issue
open
in
funding
or
something
just
to
track
this,
because
I
think
we've
got
things
kind
of
scattered
across,
like
kate
said,
io
and
kk.
We
should
get
something
in
funding,
because
I
think
this
kind
of
like
goes
both
ways.
The
I
think
some
of
the
technical
stuff
is
is
included
in
one
of
the
or
two
of
the
the
caps
that
are
currently
open.
D
D
I
know
that
there
are
a
few
places
in
sig
storage
that
are
interested
in
valve
promotion
and
we
kind
of
haven't
been
able
to
do
anything
for
them
because
of
like
either
the
system's
not
set
up
or
like
we
don't
have
a
process
to
to
route
them
through.
So
there
there
have
been
a
few
things
on
the
queue
that
just
cannot
be
done
from
the
sig
release
side.
D
I
I
think
that
I
think
that
a
simple
use
case
and
a
harder
use
case
need
to
to
sort
of
run
in
parallel,
sig
releases
effectively
in
a
split
brain
scenario
where,
like
we
run
the
the
release
process
in
google
infrastructure
and
write
the
container
images
to
community
infrastructure,
but
all
of
the
other
artifacts
stay
on
google
infrastructure,
there's
a
huge
cost
behind
those
artifacts
that
we
don't
have
visibility
into,
because
it's
not
on
community
infrastructure.
D
So
we
need
to
find
a
way
to
try
to
do
both
the
being
able
to
yeah
being
able
to
solve,
for
those
artifacts
means
we're
going
to
be
able
to
do
a
lot
more
cooler
things
in
sick
release,
so
maybe
maybe
bias,
maybe
personal
interests.
But
I
would
like
to
see
the
I'd
like
to
see
the
sig
release
stuff
happen
to
you.
A
I'm
just
gonna,
I
was
just
gonna,
say
stevens
that
yeah
like
I
think
it
sounds
like
sig
release
needs
to
collaborate
with
city,
sid,
kate's
infra,
on
this
requirement
stock
to
kind
of
understand
some
of
this
stuff,
because
what
it
sounds
like
is
we're
not.
A
Percent
clear
on
scope-
and
there
may
be
some
reasons
that
we
want
to
expand
or
contract
the
scope
in
either
direction,
but
yeah
collaboration
between
the
two
groups
definitely
sounds
warranted.
D
Yeah
absolutely,
and
that
that's
that's,
that's
already
happening
so,
like
the
the
I
think
you
know,
one
of
the
blockers
has
been.
We
can't
do
anything
new
on
community
infrastructure
until
we
have
the
solve
for
for
like
image,
distribution
right,
but
the
you
know,
but
there
is
there-
has
been
the
need
for
years
essentially
to
move
this
over
to
community
infrastructure,
but
we're
effectively
blocked
so
so
definitely
definitely
happy
to
continue
collaborating
on
on
requirements
really
for
us.
D
It's
it's
the
file,
promotion
capabilities,
it's
the
artifact,
it's
artifact
distribution,
both
on
the
you
know,
both
from
the
the
file
fireball
perspective,
but
also
the
also
the
the
the
dev
and
rpm
packages
and
the
you
know.
The
the
additional
difficulty
from
the
the
the
debian
rpm
packages
is
is
that
effectively
devin
rpm
package
management
is
like
a
managed
service
for
us
today.
If
you
will
right,
googlers
are
handling
that
right.
D
B
Also
steven
we
might-
and
we
might
have
to
end
up
increasing
the
priority
on
signing
of
the
deb
and
rpms.
If
you
know,
if
we
need
to
bring
that
into
the
community
hands
and
then
community
publishes
it
and
then
you
know,
gets
into
the
cdn
from
there.
D
B
Yeah,
so
there
is,
like
you
know,
various
combinations
of
things
that
need
to
line
up
ducks,
duck
scenario
right.
D
Oh,
the
signings
happening
planning's
happening
at
least
for
the
image
stuff.
You
know
the
the
crew
has
been
working
on
a
bunch
of
stuff
on
the
on
the
image
signing
route.
So
we
have
a
lot
of
the
capabilities
to
sign
artifacts
today,
we're
just
not
doing
it
for
the
devs
and
rpms,
and
we
can't
do
it
for
the
dubs
and
rpm2.
A
A
Yeah
yeah,
just
one
just
want
to
keep
us
focused
on
the
the
particular
steering
considerations
here
or
no.
You
have
your
if
anybody
else
want
to
want
to
speak
on
this
before
we
move
on.
I
I
think,
if
I
can
articulate
our
current
position,
I
think
steering
is
supportive
of
sick
case
infra,
exploring
this
with
the
with,
with
the
understanding
that
yeah,
we
need
to
to
take
a
look
at
requirements.
A
We
need
to
take
a
look
at
the
dependencies
on
different
release
aspects
in
determining
scope
and
then
once
we
have
those
pieces,
then
we
can
go
and
talk
to
vendors
and
there
will
definitely
need
to
be
a
flushed
out
plan
that
steering
will
need
to
look
at
and
review
prior
to
us
actually
enacting
anything
everybody
else's
understanding.
A
Yeah
so
there
it
is,
if
there
is
an
issue
already
in
the
steering
repo
on
that,
I
don't
have
a
strong
feeling
of
whether
it
resides
in
steering
or
funding.
Considering
like
funding,
you
know,
I
think,
more,
more
focused
on
funding
we're,
like
our
funding,
asks
to
the
theories.
D
A
A
Okay
sounds
great,
so
dims
or
anyone
else
have
any
other
final
comments
on
this
before
we
move
on.
A
Great,
thank
you
so
much
everyone
for
the
discussion
on
that.
The
next
topic
we
have
here
is
annual
reports.
I
think
you
added
this
one.
Paris.
C
Yep,
but
we
should
do
it
at
the
end
if
we
don't
have
any
other
business.
A
F
Oh,
I
have
one
thing
I've
been.
I
think
I've
talked
to
you
a
few
about
a
few
of
you
about
this,
but
I
think
we
need
to
work
out
with
the
cncf,
a
block
of
coupon
tickets.
That
leads
can
give
out
to
identified
members
that
need
them.
I
know
we
have
the
sponsorship,
I
mean
the
scholarship
for
attendance,
but
sometimes
people
don't
need
like
a
full
flight
and
hotel
for
a
week
and
maybe
just
like
a
ticket
can
help
out.
D
Yeah
we
we
had
a
discussion
about
this
years
back
and
I
had
an
issue
opened
in
at
least
for
the
the
release
team,
I'm
not
sure
where
we
landed
on
that
or
if
we
had
a
resolution
for
it.
So
it's
worth
I
I
think
it's
worth
digging
back
up.
C
Like
that
is
a
very
specific
purpose
for
underrepresented
folks,
and
I
feel
like
getting
those
two
things:
confused
added
to
the
mix
of
the
confusion
and
like
contributors
are
like
you're
telling
me
to
fill
out
a
diversity
thing,
I'm
like
a
white
male.
What
are
you
talking
about
so
yeah?
I
definitely
think
we
need
a
new
process
for
that.
F
And
this
specific
contributor
that
I
was
trying
to
get
a
ticket
for
did
not
get
one
after
filling
out.
He
applied
for
every
scholarship
and
did
not
get
one,
so
they
wound
up
having
to
buy
their
own
ticket.
J
Is
the
I
feel
like
in
the
past,
we've
had
some
back
and
forth
where
the
the
foundation
has
really
felt
like,
and
this
is
maybe
echoed
in
some
of
our
other
conversations-
that
we're
trying
to
take
to
the
governing
board
right
now,
but
this
notion
that
well,
their
employers
should
pay
for
their
ticket.
I
I'm
guessing
that
these
are
people
specifically
where
that's
not
appropriate,
but
can
you
give
some
detail
on
the
particular
ones
you've
seen
eddie.
J
D
So
I
see
that
there
is
still
so
I
linked
in
the
chat,
the
sig
release
tracking
issue
for
it,
as
well
as
the
steering
one.
I
see
the
steering
one
is
still
open
and
it's
actually
assigned
to
me,
so
I
can
take
that
action.
B
So
one
feedback,
so
in
the
past,
when
somebody
said
that
they
don't
have
a
ticket
or
something
like
that,
we
used
to
point
them
to
chris
and
check
and
anichek
would
just
behind
the
scenes
handle
it.
So
that
has
worked
in
the
past
many
times
like
including
full
scholarships,
so
we
won't
get
a
block
of
tickets
from
them.
The
best
way
to
do
this
is
who,
who,
how
are
they?
You
know
what
is
their
community
role
and
then
what
do
they
need?
B
And
if
you
have
a
list
of
those
we
can
without
I,
I
don't
see
any
problems
in
getting
this
thing
in
cngf.
D
Yeah
yeah,
I
think
it's
a,
I
think
it's
a
leads
to
it's.
It's
a
contributors
to
leads
to
steering
to
to
to
lf
it
yeah
and-
and
I
I
I'm
I'm
also
guilty
of
the
like
texting
chris
for
something
like
workflow,
but
we
can
yeah,
we
can.
We
can
get
something
we
can
get
something
a
little
better.
I
think,
as
long
as
people
are
routing,
we
know
that
the
route
is
available.
D
Community
contributors
know
that
the
route
is
available
to
them
via
their
leads,
and
the
leads
are
getting
that
information
up
to
steering.
B
So
for
this
round,
can
we
do
something
really
simple
like
when
you
know
somebody
who
needs
a
ticket
drop,
an
email
to
steering
private
and
we
will
like
bunch
it
up
together
and
take
it
to
cncf
or.
B
Let's
do
it
once
and
then
we'll
see
what
what
to
do
for
kubecon
n,
a
exactly
jordan,
if
only.
B
Who
wants
to
do
the
form?
That's
action.
D
D
Yeah,
who
I
remember
at
some
point,
I
had
access
to
kubernetes
surveymonkey.
I
don't
know
if
that's
still
true,
because
it
was
a
while
back,
but
I
can
poke
at
it.
D
A
Okay,
so
stephen
is
gonna,
take
point
on
this
one
as
far
as
the
the
action
items
and
stuff
coming
out
of
it,
and
I
think
I
think
kind
of
what
has
happened
in
the
past,
at
least
when
I
see
on
this,
like
it
comes
up
usually
right
before
kubecon,
and
then
we're
really
busy
and
then
we're
dealing
with
kubecon
and
then
kubecon
is
over.
And
then
we
promptly
forget
about
it
for
another.
B
Christoph
it
was
mostly
because
of
cncf
dragging
its
feet
on
having
a
separate.
You
know,
path
for
getting
these
requests
and
they're
always
happy
to
do
the
one-off
things
they
just
didn't
want
it
structured
for
some
reason.
B
Yeah,
I
think
it
is
also
because
of
the
scalability
issue,
because
if
they
do
it
do
this
structured
thing
for
us,
then
they
have
to
do
the
structure
thing
for
all
cnc
or
projects
right.
So
that
is
probably
behind
some
of
these
pushbacks
as
well,
but
yeah.
Let's
try
to
do
what
we
can
for
n
a
and
then
sorry
for
eu
and
then
negotiate
for
any.
A
A
Okay,
so
being
that
we
have
about
22
minutes
left
and
we
wanted
to
spend
at
least
15
minutes
on
the
annual
reports.
I'll
say:
is
there
anything
else,
very
quick
that
we
want
to
bring
up
and
deal
with,
or
should
we
move
on
to
the
annual
report
discussion.
A
Okay,
great
so
we'll
move
on
to
annual
reports.
Paris
did
you
want
to
frame
the
discussion?
What
what
you
want
to
talk
about
today.
C
Yeah,
so
we
are
officially
in
the
process
for
getting
this
summary
together,
I
created
the
hack
md
I'll
put
it
in
the
chat
too
one.
Second,
there
we
go,
and
we
pretty
much
need
to
populate
this
at
this
point.
So
if
y'all
have
completed
merged
annual
reports
at
this
point,
this
is
for
you
or
even
if
you're,
seeing
stuff.
That
would
be
a
theme
for
things
that
aren't
merged.
C
That's
still
fine,
too,
but
our
major
responsibility
as
liaisons
is
to
make
sure
that
the
themes
for
your
groups
are
represented
on
here,
as
well
as
anything
else
that
you're
seeing
that
falls
under
these
headings,
and
I
also
need
someone
to
help
me
with
this.
So
I
need
a
partner
in
crime.
For
the
summary,
any
takers.
K
C
All
right!
Well,
since
you
have
really
good
scripting
skills,
maybe
you
can
help
me
with
the
one
part
which
is
to
get
like
all
the
all
the
stuff
under
the
headings
with
like
help
wanted
and
stuff
like
that
merged
in,
so
that
we're
not
doing
manual
work.
So
that's
number
one
that
you
could
help
out
with
sure
sure.
C
C
C
I
added
the
terminology
section
down
here.
This
is
something
that
we
didn't
do
last
year,
but
celeste
and
I
in
the
editorial
process,
did
because
we
went
through
and
saw
that
there's
a
bunch
of
acronyms
and
stuff
like
that
that
people
might
not
know
like
sig.
I
mean
we
know
that,
but
if
there's
anything
else
from
like
acronyms
that
y'all
are
pooling
and
things
that
you
feel
like
need
to
be
explained
better,
please
add
them.
C
We'll
of
course
use
the
one,
the
blurb
that
we
had
last
year,
which
again
is
like
sigs,
and
you
know,
working
groups
and
governance
and
stuff
like
that.
So
add
that
to
the
bottom,
if
you
have
anything,
but
so
you
can
see
free
here.
C
Some
of
the
themes
that
I
was
seeing
already
one
is
the
theme
of
our
independent
contributors
are
still
pretty,
are
still
riding
pretty
high
in
a
lot
of
these
groups,
meaning
like
number
three
and
number
four
spots
for
top
contributors
in
each
group
has
an
independent
label
which
I
think
is
still
pretty
strong,
meaning
like
that's
people
that
aren't
necessarily
supported
by
major
employers
and
or
have
benefits,
etc.
C
And
then
every
group
that
I
have
looked
at
needs
at
least
one
area
that
has
that
needs
help
and
what
that
help
is
like.
We
need
to
define
that
obviously,
and
then
that's
like
the
two
that
I've
seen
and
I
think
we
need
to
ride
at
least
four
at
least
four
themes
and
same
with
the
current
initiatives,
so
they'll
be
populated
from
the
report.
So
that's
nothing
necessarily
from
us
anything
that
has
like.
We
need
to
do
something
like
as
in
steering
and
not
necessarily
a
scripted.
C
You
know
a
back
script
or
whatever
else
it
says,
steering
underneath
the
section
any
questions
about
like
what
y'all
need
to
do
or
give
for
the
summary.
H
C
E
H
A
meta
theme
is
that
everyone's
either
really
busy
or
really
fragmented
right
now,
and
so
even
just
the
pace
at
which
these
are
coming
in
and
feedback
loops
are
happening,
is
yeah.
That's
a
meta
theme.
I
feel
it
myself.
C
And
that's
good,
that's
exactly,
and
we
can
both
describe
these
and
obviously
with
with
more
words
later
on,
and
you
know
get
better
words
not
for.
C
I
think
there
was
actually
a
theme
last
year
as
well.
I
think
so
repeat
the
last
thing
that
you
just
said
something
about
quality.
H
Lower
volume
but
higher
quality,
so
instead
of
hey,
we
did
10
new
alpha
things.
It
was
like
we
graduated
a
couple
things
and
we
got
one
new
alpha
thing
in
flight,
but
we
actually
have
a
design
like
for
all
the
way
to
ga.
For
this
thing,.
C
C
Getting
out
the
report
from
last
year,
so
I
think
that
was
one
of
ours,
which
is
still
good,
because
that
can
be
a
theme
is
that
we
have
a
a
consistent
element
here.
H
H
So
from
what
I've
seen.
That
seems
to
be
paying
off,
maybe
not
100,
but
at
least
directionally.
It
seems
good.
J
J
C
And
so,
as
so
speaking
of
growth
areas,
our
last
year,
growth
areas
were
things
like
we
needed
to
focus
on
reliability.
That's
when
you
know
the
the
proposal
went
out
and
you
know
so.
Maybe
we
could
have
some
kind
of
theme
that
would
follow
up
on
this.
C
The
words
we
used
all
eyes
on
security
and
then
improving
caps.
Those
were
our
growth
areas
last
year
and
the
words
that
we
use.
H
Yeah
the
reliability
aspect,
I
do
feel
like
they're,
not
just
feeling
there
were
a
lot
of
regression
releases
recently
like
over
the
last
three
to
six
months.
122
had
some
high
profile
regressions
that
actually
still
haven't
been
resolved.
H
There
were
at
least
one
or
two
and
123
as
well,
so
I
don't,
I
feel,
like
that's,
probably
an
area
of
growth
like
in
the
design
and
graduation
phases.
I
think
I've
seen
clear
growth,
but
in
the
test,
quality
and
avoiding
regressions.
I
think.
H
H
So,
like
I,
I've
seen
actually
some
pushback
on
the
reliability
proposal
recently
and
like
I,
I
hear
the
concerns
that
are
being
expressed
on
that
in
terms
of
like
we
don't
want
to
disincentivize
people.
We
don't
want
to
discourage
people,
we
don't
want
to
be
punitive
or
punish
contributors,
but
I
think
that
is
addressing
issues
that
are
real
issues,
and
so
I
don't
want
to
see
that
effort
lose
steam.
D
So
I
would
say
you
know
related
to
reliability.
I
think
people
have
not
spent
enough
time
looking
at
it,
and
I
can,
I
can
say,
is
one
of
the
sponsor
say
chairs
that,
like
we
have
not
spent
enough
time
looking
at
that,
so
I
think
raising
visibility
there.
If
it's
going
to
improve
reliability
across
the
project
is
really
important.
D
What
the
right
things
to
raise
visibility
on
and
if
you
like,
what
are
the?
What
are
the
the
lowest
cost
highest
impact
things?
We
can
raise
visibility
on
right
if
it's,
if
it's
a
small
change
that
has
a
possibility
to
to
ripple
out
through
the
community
whether
it's
you
know
whether
it's
like
improving
triage,
for
example
right.
You
know,
if
there's
a
situation
where,
because
we've
talked
about
triage
party
in
the
past
we've
talked
about
you
know,
we've
talked
about
board
management.
We've
talked
about
automating
things
like
that.
D
That's
a
that's
something
that
you
know.
Maybe
maybe
it's
a
few
tweaks
that
causes
you
know
that
causes
a
huge
impact
on
like
on
being
able
to
just
be
productive
right.
I
think
if
we
can
find
more
of
those
things,
we
should
target
them.
I
think
reliability
is
definitely
one
of
something.
H
If
we
want
to
prioritiz
or
if
we
want
to
prioritize
things
that
are
most
user
visible,
so
like
reliability
has
two
aspects,
one
is
on
contributors
and
one
is
on
users
and
from
the
user
perspective,
I
think
things
that
actually
make
it
into
a
release
and
get
marked
as
a
regression
like
that's
a
really
clear
signal
of
something
that,
like
whatever
processes,
we
had
missed
this
thing,
sellover
or
whatever,
like
so
highlighting
that
from
a
user
perspective,
seems
like
a
good
idea,
and
then
I
think
there
are
thinking
from
the
contributor
perspective
burnout
or
like
being
overwhelmed
or
like
not
being
sure.
D
Yeah,
so
I'm
yeah,
so
I
kind
of
leaned
more
towards
the
contributor
maintainer
side
here,
because,
like
we're,
we're
always
going
to
ship,
regardless
of
like
when
it
happens,
how
it
happens
like
it's
like
we're,
gonna
shift
the
thing
right,
the
qual,
the
like
the
the
quality
level
of
it
is,
is
going
to
be
different
depending
on
how
focused
maintainers
are
at
any
one
point
in
time.
Right
like
there
are
going
to
be
things
that
we
we're
going
to
miss.
D
There
are
always
going
to
be
things
that
we're
going
to
miss
but,
like
I
think,
the
process
improvements
like
improving
quality
of
life
for
the
maintainers,
because
we're
always
going
to
ship
like
that,
will
have
a
better
that'll,
have
a
better
impact
on
the
end
user
experience.
H
Yeah,
I
think
I
think
that
the
key
is
making
it
so
that
it
doesn't
require
intense
100
of
the
time
focus
from
maintainers
to
avoid
regressing
like
we
had
places
where
we
have
seen.
Regressions
show
that,
like
we
didn't,
have
systems
in
place
to
protect
us,
and
so,
if
we
had
a
busy
week
or
month
and
lost
focus
like
something
slipped
through
and
so,
rather
than
saying
like,
we
need
to
respond
to
that
by
just
being
focused
for
the
next
release.
D
C
C
B
I
I
think
the
number
of
people
who
can
spend
nine
to
five
all
all
week,
long
on
kubernetes,
it's
kind
of
over.
We
need
to
call
that
out
yep.
I
agree
we
need.
You
know
we
have
to
like
kicking
or
screaming
we'll
go
to
a
thing
where
people
are
only
able
to
work
on
kubernetes
a
few
hours
a
day
a
few
days
a
week,
and
we
have
to
adjust
better
than
what
we
are
already
doing.
C
I
had
like
dm'd
bob
earlier
about,
like
honestly
three
different
archive
things
like
you
know,
one
was
about
like
best
prac
like
they
were
about.
There
was
just
different
all
different
situations
with
three
different
people
in
the
in
the
project
I'm
noticing.
So
my
theme
here
is:
I'm
noticing
that
we're
archiving
more.
Is
that
true
or
false
both
process?
G
No,
I
don't
think
we're
archiving
more
as
as
the
person
that
probably
does
at
least
the
the
archiving
of
the
github
repos.
That's
it's
pretty
like
there
hasn't
been
a
large
influx
in
them.
D
D
Maybe
right
because,
like
I
think,
there's
like
a
chunk
of
things
that
we
were
already
planning
to
do
that
are
kind
of
like
happening
right
or
have
happened
at
this
point
as
opposed
to
like
this
is
a
need,
but
I
think
we're
I
think,
as
we're
we're
doing
annual
reports
as
we're
kind
of
like
checking
in
with
things
and
and
governance
groups,
more
frequently
we're
getting
a
better
signal
on
like
what
is
and
isn't
working
right
if
it's
a
sub-project.
D
J
Yep
this
goes
back
to
the
idea
of
we're
all
busy
and
spread
so
thin
at
some
point.
If
we've
gone
through
the
review
annual
review
process
once
or
twice
and
said,
like
yeah,
we
we
intend
to
get
to
this.
We
intend
to
get
to
this,
but
you
know
what
it's
not,
let's,
let's
go
ahead
and
it's
not
that
it's
not
important,
but
it's
clearly
not
staffed.
It's
not
something.
We've
got
the
bandwidth
for
now.
Let's
take
it
off
the
list
so
that
we're
not
stressing
over
not
delivering
it.
C
It
almost
seems
like
good
sustainability
question
stuff,
though
you
know,
like
that's
the
like
that
that's
almost
kind
of
like
the
answer
to
the
dims
like
we're,
not
a
nine
to
five
yeah.
That's
got
that's
kind
of
where
my
mind
was
going
with
that
as
well,
like
we're
not
a
nine
to
five
and
here's
some
of
the
things
that,
like
you,
can
see
that
we're
doing
like
you
know,
people
are
thinking
about
archiving
more
and
how
to
automate
themselves
out
of
their
contributor
job,
and
you
know
yadda
yadda,.
D
I
think
there's
also
like
a
making
sure
that
the
record
is
visible
about
how
we're
doing
this
stuff,
because
there's
also
like
a
just
because
we
needed
to
archive
it,
doesn't
mean
that
it's
not
necessarily
important
like
we
should
provide
a
route
for
a
group
of
intrepid
contributors
to
potentially
re-raise
something
like
we've
talked
about.
We
had
a
licensing
sub
project,
for
example
in
the
past
right,
and
it
was
kind
of
like
us,
occasionally
poking
steve
winslow
when
he
was
at
the
left
and
going
like.
D
Does
this
like
look,
you
know,
but
it
was.
It
wasn't
like
a
constant
need
right
but
say
we
did
have
folks
who
were
who
cared
about
that
right
or
say
there
were
licensing
issues
popping
up
more
frequently.
There
should
still
be
a
like
here.
Was
the
decision
record
at
point
in
time
x
that
we
decided
to
shut
this
down?
This
is
why,
if
you
wanted
to
pick
it
up,
here
are
some
suggestions
about
what
you
would
do
right
and
how
it
needs
to
be
staffed
to
be
successful.
Moving
forward.
C
D
For
for
kubernetes
kubernetes
in
particular,
you
you
run
into
this
very
interesting
fun,
weird
thing:
where
there's
several
of
the
the
the
sub
directories
that
are
owned,
that
that
are
not
really.
I
think
someone
asked
the
question
recently
about
like
the
build
directory
right
and
we
have
several
folks
who,
on
the
call,
who
are
all
reviewers
approvers
for
the
build
directory,
but
it
doesn't
necessarily
let
like
land
on
you
know.
The
question
got
asked
in
sig
release
and
it
doesn't
land
on
sig
release
and
it
doesn't
really
land
on
sync
testing.
D
It
lands
on
those
reviewers
approvers
right,
so
that
that
you
know
they're
a
lot
of
the
things
that
are
owned
as
a
result
of
like
being
in
a
subdirectory
that
does
not
have
clear
ownership.
You've
got
you've
got
instances
where
there
are
tests
that
run
off
of
that
or
you've
got
instances
where
you've
got
like
if
you
think
about
all
of
the
generated
tests
across
just
kubernetes
kubernetes
that,
like
a
new,
a
new
release,
branch
spins
up
or
something
and
their
new
generated
tests
that
don't
necessarily
get
the
attention
that
they
need
to.
D
You
know
for
us
to
have
that
clear
signal
like
do
we
release
or
not
right,
that's
worth
investigating
right,
and
that
is.
That
is
not
just
that's
not
just
a
release
thing.
That's
not
a
testing
thing
that
is
like
that's
everyone.
D
E
A
It
was
a
great
discussion.
I'm
gonna
separate
just
to
wrap
this
up.
Thank
you,
everybody
for
attending.
If
there's
any
follow-up
items
you
can
reach
out
in
our
slack
channel
hashtag
steering
committee,
or
you
can
hear
committee
repo
on
github
or
lastly,
steering
at
kubernetes
io
thanks
so
much
everyone
see
you
next
month.