►
Description
Kubernetes Public Steering Committee Meeting for 20220912
A
wow.
This
is
a
recorded
steering
meeting.
This
is
a
public
steering
meeting.
This
will
be
recorded
and
available
on
youtube
later.
So
please
be
mindful
of
what
you
say
and
do
please
be
sure
to
adhere
to
the
kubernetes
code
of
conduct
and
in
general,
that
boils
down
to
just
being
awesome
people,
so
we've
got
a
pretty
light
agenda
today.
A
I
know
that
we've
got
a
few
people
who
are
in
around
traveling
to
conferences
prepping
recovering
from
other
ones,
I'm
currently
in
dublin
for
open
source
summit,
as
well
as
some
of
our
some
of
our
other
open
source
compatriots
so
getting
into
getting
into
some
of
the
cncf
updates.
I
think
the
big
one
right
now
is
kubecon
is
coming
up.
Maintainer
updates.
So
for
folks
who
are
sig
chairs
technical
leads,
folks
that
are
intend
to
provide
updates
on
their
governance
group.
A
Please
be
sure
that
you're
working
on
those
updates
for
for
your
presentations.
There
are
also
some
opportunities
to
present
things
that
are
happening
within
the
kubernetes
community.
On
the
on
the
big
stage
for
your
familiar,
maybe
now
around
the
the
cncf
project
updates,
as
well
as
the
kubernetes
project
updates
that
happen
for
the
keynotes.
A
So
if
you're
interested
in,
if
you're
interested
in
providing
some
updates
there,
please
let
us
know,
drop
us
a
note
or
reply
to
the
emails
that
are,
I
believe,
cody
sent
out
a
few
days
ago
christoph.
I
noticed
that
your
hand
was
raised.
B
Yeah,
I
was
just
gonna
suggest.
I
see
some
folks
that
are
dropping
updates
in
here.
Probably
the
best
thing
to
do.
Instead
of
that
is
there's
an
email
that
went
out
to
leeds
at
kubernetes.io
from
cody
replying
to
that
following
the
instructions
there,
that's
the
best
way
to
kind
of
get
that
information
into
the
right
hands.
A
What
christoph
said
all
right,
so
I
see
one
of
them
is
kate's
gtr.io
to
registry.
That
kate
said
I
o,
I
believe,
some
of
them.
C
Is
there
a
time
frame
for
when
we
want
to
have
those
gathered
like
I
assume
there
is,
but
do
we
know
what
that
is.
A
I
would
say
before
the
end
of
the
month
would
be
great
okay.
I
know
that
that
is
a
soft
deadline
for
for
keynote
updates,
at
least
so
I
would
say
to
give
the
co-chairs
time
to
integrate
that
into
their
their
presentations.
I
would
do
before
the
end
of
the
month.
A
Sweet
all
right
anything
else
that
we
want
to
mention
for
cncf
updates,
dims,
maybe
anything
falling
out
of
toc
gb
stuff.
D
No,
nothing
specific.
I
can't
think
of
anything
that
we
want
to
give
a.
B
D
Yeah
right
so
right
now
I
have
four
issues
opened
for
health
checks
of
different
projects,
hcd
for
sure,
and
they
were.
A
D
We
can
try
for
sure
is
yeah.
I
think
one
of
the
things
that
I
I
I
for
the
last
like
two
three
weeks.
I
ended
up
pointing
out
our
annual
reports,
both
the
you
know,
2020
and
21,
to
a
bunch
of
people.
So
I
think
I
don't
know
if
anything
will
come
out
of
it,
but
that
is
a
good
resource,
so
you
know
we
should
try
to
continue
to
do
it
next
year,
too,.
A
Yeah,
I
know
in
addition
to
the
the
other
happenings
around
the
sed.
Maintainership
there's
also
been
a
thread
around
single
cluster
reliability.
It's
a
pretty
big
thread.
There
are
lots
of
folks
involved.
This
is
coming
down
to
like
the
raft
implementation
within
fcd
jordan.
I
think
you're
over
some
of
that.
I
don't
know
if
you
want
to
speak
to
a
bit
of
it.
C
I
mean
merrick
brought
it
up
kind
of
escalated
it
and
actually
several
several
of
the
ncd
maintainers
past
maintainers
jumped
in
and
said
I
I
had
no
idea.
This
was
like
a
thing
and
an
issue.
This
isn't
a
known
issue
to
me.
I
did
not
expect
this,
and
so
I
think
that
helped
get
traction
around
like
actually
fixing
it,
instead
of
just
documenting
it.
So
this
was
in
single
node
cd
clusters.
C
There
was
one
stage
of
committing
rights
that
did
not
go
through
the
same
code
path
as
multi
node
cd
clusters,
and
so,
if
you
had
a
local
disk
crash,
you
could
basically
put
a
single
node
cluster
into
an
unrecoverable
state.
If
the
disk
crashed
at
exactly
the
right
time,
I
don't
know
if
it
was
under
comfortable
or
it
would
lose
something
that
he
didn't
reply
to
the
kind
of
great
anyway.
The
technical
details
are
not
important.
C
The
the
issue
is,
it
was
not
well
communicated,
it's
not
well
understood,
and
so
this
is
an
example
of
a
type
of
sort
of
quality
reliability
issue
and
within
fcd
that
we
would
want
to
have
better
visibility
to
and
better
communication
of
and
better
representation
from
downstreams
who
are
sort
of
building
on
this
as
a
foundation
and
not
realizing
the
current
state.
C
So
it's
communication
in
both
directions
from
etcd
to
the
people.
Building
on
it
saying
this
is
the
current
state
build
on
it
appropriately
and
then
communication,
the
other
direction
to
say
actually,
we've
already
built
skyscrapers
on
this.
It
would
be
awesome
if,
like
this
foundation,
behaved
the
way
everybody
building
on
it
thought
it
did
so
yeah,
so
that
ended
up
happening,
but
it
took
longer
than
I
think
it
should
have
to
kind
of
get
the
two-way
communication
going.
A
C
The
the
fix
is
ongoing,
like
it's
actively
being
worked
on,
and
I
think
there's
agreement
on
taking
some
form
of
it
back
to
the
three
four
and
three
five
versions
that
are
right,
which
is
good.
So
this
particular
issue,
I
think,
is
in
a
good
state.
It's
moving
forward,
I'm
more
interested
in
like
what
about
the
next
issue.
How
how
can
we
close
that
communication
look
more
quickly
and
more
reliable.
C
I
had
a
great
quote:
someone
share
with
me
this
week.
It's
like
that
hill
you're,
trying
to
reach
the
top
of
is
the
horizon.
You
will
not
read
this.
The
question
is
like:
how
are
you
going
to
improve
the
way
you're
running
yeah,
that's
profound
and
depressing,
but.
D
I
I
do
want
to
pick
on
one
more
thread:
there,
jordan,
which
is
has
the
apa
missionary,
talked
about
externalizing
or
putting
an
interface
around
it
to
see?
If
you
know
the
age
old.
C
Question
I
I
have
seen
conversations
happening.
I
don't
remember
if
they
happened
in
an
api
machinery
meeting
or
in
slack
I
I
think
the
concern
is
that
an
interface
that
fragments
doesn't
actually
do
anything
to
help
all
existing
users
who
are
already
on
it,
and
so
that's
the
main
sort
of
conclusion.
Most
of
the
discussions
I've
seen
have
reached.
C
If
we
divide
the
energy
and
don't
have
a
migration
path
from
for
people,
we've
actually
made
all
our
existing
users
lives
worse.
Yeah.
A
Kristoff,
I
noticed
your
hand
is
still
up.
Is
that
from
the
previous
comment
or.
B
No,
so
one
other
thought
that
comes
to
mind
here
like.
I
would
like
to
think
that
we
are
pretty
good
about
this,
but
I,
when
I
think
about
like
foundational
components
and
that
kind
of
stuff,
yes,
we
we
build
on
top
of
ncd,
but
there
is
a
lot
of
downstream
consumers
of
us
that
also
build
on
top
of
us
and
what
one
kind
of
angle
that
at
least
I'm
thinking
about
here
is
you
know,
as
we
have
seen
and
experienced
over
the
last.
B
You
know
couple
years
and
that
kind
of
stuff,
when
we
communicate
about
issues
in
our
product
or
is
there,
is
there
things
that
we
can
learn
and
things
that
we
can
do
better
going
forward?
Not
just
analyzing,
like
you
know,
for
example,
I'm
talking
about
like
doctor,
shim
deprecation
and
that
kind
of
stuff
like
we're,
we're
analyzing
our
own
incidents
and
how
we
can
communicate
better
about
the
things
that
we
do.
We
should
also
kind
of
keep
in
mind
these
other
discussions
that
are
going
on.
B
Look
if
there's
ways
that
again,
we
can
learn
from
others
as
well
about
how
to
communicate
better
about
things,
because
recognizing
that
ncd
is
a
foundational
component
for
us.
But
we
are
also
a
foundational
component
for
so
many
others,
not
just
operators
that
are
building
on
top
of
kubernetes,
but
other
people
that
are
building
other
open
source
projects
and
stuff.
On
top
of
us.
A
There
yeah-
absolutely,
I
think
you
know
in
in
terms
of
the
the
the
communication
that
I've
seen
for
for
the
125
release
and
the
kind
of
the
the
uptick
in
and
like
release
blogs
around,
not
just
feature
sets,
but
like
a
bit
more
focus
around
the
the
you
know
the
the
easy
to
miss
things.
The
the
things
are
foundational
components,
the
things
that
are
infrastructure
related.
You
know
I,
I
think
the
team
has
gotten
better
at
it.
A
Definitely
at
least
not
to
pat
my
own
team,
but
I
I
think
the
you
know
this.
The
definitely
the
cycle-
and
I
mean
really
building
up
over
the
last
few
cycles.
We've
gotten
a
lot
tighter
about
talking
about
non-future
changes
and
things
that
are
kind
of
core
community
concerns.
A
So
I
I
think
we
we
keep
keep
up
the
steam,
and
if
we
see
that
this
is
you
know,
especially
the
ncd
stuff.
We
see
that
this
is
bubbling
into
something
that
is
affecting
you
know,
kubernetes,
operators
and,
and
it's
something
that
we
should.
We
should
absolutely
highlight
when
we,
I
don't
think
we
should
ever
take
for
granted
that
you
know
when
we
say
words,
people
listen
and
that
may
be.
The
exact
kind
of
you
know
effect
that
smaller
projects
within
our
ecosystem
need.
A
They
might
need
that
that
loud
speaker
mega
horn,
that
the
kubernetes
project
is
able
to
provide.
So
if
we
can
figure
out
where
those
where
those
paths
are
like
scd
is
definitely
one
of
them
right,
but
like
building
applications
on
top
of
kubernetes-
and
these
you
know
I
mean
so
much
in
the
cncf
ecosystem
is-
is
either
you
know
based
on
kubernetes
or
some
sort
of
you
know
their
their
governance
models
or,
or
you
know
just
the
way
they
they
do.
D
So
one
thing
I
agree
with
you:
there
stephen
one
thing
I
think
we
probably
haven't
done
yet
is
like
the
install
various
installers
of
kubernetes
right,
some
of
like
even
cluster
api,
has
a
specific
pattern
in
which
they
use
hcd
right.
So
we
need
to
kind
of
like
figure
out
who
who
are
the
popular
installers?
What
kind
of
patterns
do
they
use
and
then
you
know
make
targeted
approach
to
hey
if
you're
using
you
know
three
noted
cd
stuff,
don't
worry
about
it.
You
know.
D
We
know
that
we
have
to
talk
to
the
other
set
of
people
who
are
using
single
loan
or
you
know
something
is
the
default.
Something
is
not
a
default
right,
so
exactly
we
have
to
do
some
more
of
that.
I
think
right
now
we're
just
throwing
it
into
the
void
and
saying
hey,
hcb,
broken
kind
of
thing.
Instead
of
that,
we
should
like
have
a
handshake
with
these
people
to
see.
If
we
can
do
that,
better
roll
things
out
better.
C
Yeah
that
reminds
me
of
some
of
the
sweeps
we've
started
doing
when
we're
making
changes
in
kubernetes.
We
used
to
just
say:
like:
does
this
fall
within
our
deprecation
policy
or
our
change
policy
like
it
does
awesome
ship
it,
and
now
we
will
actually
like
go.
You
know,
search
github
for
callers
of
this
thing
and
go
and
open
like
50
issues
in
repos
and
say,
like
we
noticed
you're
using
this
thing.
C
That
is
not
correct
and
will
change
in
like
two
releases:
here's
what
you
can
do
instead,
like
being
very
proactive
in
identifying
use-
it's
not
perfect,
like
it
doesn't
catch
private
use,
but
we
can
at
least
you
know,
do
what
we
can
to
identify
it,
and
sometimes
that
actually
changed
our
plans
like,
even
though
we
would
be
within
our
rights
to
like
make
a
change.
We
could
really
help
people
and
therefore
help
ourselves
by
doing
this
more
gradually
more
safely.
D
C
D
I
think
we
should
ask
the
people
who
are
installing
hdd
by
default
in
various
places,
to
like
register
themselves
somehow
with
the
cd
folks,
so
the
outgoing
con,
so
the
people
maintaining
a
cd
can
send
out.
You
know:
communications,
better,
targeted
communications,
better
based
on
patterns
of
usage.
Do.
D
A
Okay,
all
right,
so
I
am
not
seeing
any
votes.
Are
there
any
other
cncf
updates
that
people
would
like
to
mention
before
we
might
caught
into
epics.
D
By
torch
joined
by
torch
yes
this
morning,
and
they
had
a
quote
about
cuban,
it
is
there
in
there
too.
So
I'm
sorry.
D
A
All
right,
jordan,
jordan
got
it
awesome
all
right
and
I'm
and
I
you
know,
I
believe
there
will
be
a
bunch
of
announcements
this
week
for
open
source
summit.
I
know
that
there
are
a
few
coming
down
the
pike
for
open,
ssf
and
open
access
day.
That's
something
near
and
dear
to
my
heart,
but
but
I
won't
ruin
any
of
the
surprises.
A
A
If
you
are
a
voter,
you
should
have
received
a
you
should
have
received
information
about
signing
into
electo
and
being
able
to
cast
your
your
vote
in
electo.
You
will
also
find
not
just
your
ballot,
but
also
information,
which
you
can
also
find
in
the
kubernetes
community
repo
for
the
the
voter
profiles,
but
that
is
that
is
also
available
on
alecto.
You
can
read
about
each
of
the
candidates
and
make
very
well
informed
decisions
for
the
usher
in
the
next
next
three
folks
to
join
the
steering
committee.
A
So
please,
if
you're
a
voter,
make
sure
that
you
prioritize
taking
some
time
to
to
do
the
votes.
It's
super
quick!
It's
not
intensive
at
all!
If
you
are
not
a
voter,
please
take
a
look
at
the
voter
requirements
for
for
for
the
next
go
around
and
make
sure
that
you
are
working
towards
being
eligible
to
to
the
next
time
around
dimms.
You
were
actually
one
of
the
election
officers,
so
I
will
stop
yapping
and
let
you
fill
in
some
color.
If
I
missed
anything.
D
I
so
over
the
weekend
we
got
five
requests
for
hey,
I'm
not
not
on
the
list.
Can
you
please
add
me
and
we
kind
of
like
granted
four
of
them,
so
we
are
not
seeing
too
much.
I
think
when
we
took
when
we
looked
at
the
number
of
people.
I
think
it
was
like
80
people
voted
on
thursday
I'll
go
back
and
check.
You
know,
probably
middle
of
the
week
to
see
how
many
awarded
we
might
have
to
like
go.
Tell
more
people
shake
it.
D
Yeah
yeah
shake
the
trees
to
to
do
this
for
sure.
But
you
know
from
from
the
point
of
view
like
the
back
end
and
everything
is
going
fine,
the
we
had.
We
had
new
election
officers
step
up
this
time,
so
there
was
a
little
bit
of
hey.
We
missed
one
thing
here
and
we
missed
one
thing
there,
but
you
know
folks,
like
jordan,
caught
a
few
of
those
too.
So
thanks
jordan.
So
from
the
comms
perspective,
we
are
doing
fine.
It's
just
that.
D
Takes
vote
than
a
minute,
I
think
exactly
yeah.
So
on
the
cncf
side,
I
think
I
missed
a
couple
of
updates.
This
is
not
this
doesn't
come
out.
A
D
All
right,
they
are
doing
like
two
specific
things
around
a
community
safety
and
trust
which
was
literally
driven
by
you
know
some
of
the
some
of
us
going
there
and
saying
hey.
We
need
to
do
better,
so
there's
like
a
crisis
communication
stuff
that
they're
trying
to
do,
and
they
also
do
trying
to
do
some
proactive
things
as
well.
I
am
hoping
that
they'll
announce
it
in
kubecon,
if
not
before,
so
watch
out
for
that,
because
that
that
was
literally
driven
by
you
know,
folks
from
humanities,.
A
Thank
you
folks
who
worked
on
that
yeah,
we're
again
we're
we're
often
the
flag
bearers
for
for
a
lot
of
the
stuff.
So
I
I
think
you
know.
Maybe
if
you
see
something
say
something
is
oversimplifying,
but
a
lot
of
the
things
that
we
do
turn
into
initiatives
like
this.
So
please,
when
you
have
the
opportunity
to
provide
input
on
improvements
for
not
just
our
community
but
the
wider
cncf
and
wider
lf
communities.
A
All
right
sponsor
developer
programs.
I
know
I
know
you
dropped
some
notes
in
here.
Do
you
want
to
speak
to
that.
E
Yep
hi
everyone,
so
this
is
not
a
new
subject.
Pros
say
like
we
brought
this.
I
think
there
was
a
conversation
about
this
during
the
public
meeting
of
may
and
the
this
is
this
week's
this
request
came
through
mini
cube
about
basically
buy
a
developer
program
for
apple,
but
this
is
more
like
my
release
manager
at
coming
and
ask
about
basically
how
we
can
how
cncf
ken
spencer
developer
program
for
mac,
os
x
and
microsoft
windows,
because
this
conversation
came
during
last
six
release
about
how
we
understanding
sure
for
the
operating
system.
A
I
don't
currently
have
any
updates.
I
know
that
that
is
on
my
plate
on
the.
I
know
that,
at
least
from
the
from
the
sig
release
perspective
that
we're
doing
some
investigation
on
the
not
not
on
on
the
mac
os
side,
but
on
the
on
the
devs
and
rpm
side
for
community
packages.
A
I
know
that
there
is
like
mini
cube
is
is
also
one
of
the
projects
that
that
come
to
us
and
and
and
ask
if
it's
possible
to
kind
of
like
be
within
the
official
stream
for
those
packages,
at
least
the
developer
program.
I
I
believe
still
requires
some
investigation,
because
it's
not
clear
exactly
how
one
does
a
group
account
for
it.
E
So
there's
like,
I
think
we
talked
about
the
race
last
secretive,
meeting
defensive
some
limitation
about
basically
group
signatures,
so
we,
this
is
more
like
about
the
financial
aspect
of
artifact
signature.
How?
Basically,
if
we
get
a
consensus
about
how
we
implement
that
with
some
specific
tooling,
how
we
get
access
to
this
account,
like
is
cncf
supposed
to
pay
for
this.
Someone
supposed
to
pay
for
this,
like
okay,.
A
To
to
yeah
to
demonstrate
in
chat
money
is
not
necessarily
the
issue
here.
It
is
they.
I
think
it
is,
and
potentially
an
impediment
of
the
the
functionality
in
general
right
like
how
to
do.
I
think
there's
an
assumption
that
it
is
a
single
user
process
where
I
I
have
to
imagine
that
that
is
that
can't
be
the
case.
A
There
can't
be
enterprises
letting
out
signed
assigned
binaries,
though
on
someone's
laptop
or
something
so
so
it
dims
in
the
chat
is
saying
that
paris
is
trying
to
dig
into
it
for
mac
os
from
the
apple
side.
So
I
will,
I
will
sync
up
with
paris
and
see
what
the
latest
is
and
then
we'll
provide
some
updates
on
that
on
that
issue.
E
Okay,
this
is
just
the
issues
about
apple.
Are
we
open
an
issue
about
microsoft
and
try
to
tie
this
up
in
the
umbrella
issue,
trying
to
basically
figure
out
that
we
can
handle
that
for
microsoft
and
apple?
At
the
same
time,
see
what's
happening?
I
think
the
issue
about
one
individual
sign.
We
can
kind
of
handle
that
with
circuit
manager,
one
password
stuff
like
that
basically
have
like
a
vault
where
we
can
push
those
icon
password
and
share
between
individuals
about
signature.
So.
A
So
so
yeah
eventually
there
will
be-
I
mean
once
once
we
understand
how
it
works,
if
it,
if
it,
you
know,
it
is
tantamount
to
like
this
needs
to
be
a
service
account
system
account
a
bot
user.
What
have
you
then
then
yeah,
the
the
appropriate
people
will
be
authorized
and
have
access
to
that
vault.
I
imagine
that
we'll
probably
have
to
do
multiple
ones.
A
I
I
see
that
what
we
would
consider
what
you
consider
like
the
canonical
artifacts
for
the
project
versus
subproject
artifacts,
those
those
I
I
imagine
would
would
have-
would
be
handled
as
separate
cases.
I
I'm
imagining
with
my
sig
release
hat
on
the.
A
I
think
one
of
the
the
blockers
for
for
doing
things
like
adding
the
mini
cube
packages
to
devin
rpm
streams,
for
example,
is,
is
that
we
don't
have
the
process
nailed
down
for
devs
and
rpms
in
general,
right
and,
and
there
is
there-
is
an
active
danger
in
introducing
more
more
skew
into
that
process.
So
until
we
nail
that
down
and
sig
release
is
looking
at
doing
proof
of
concept
with
open
build
service
right
now
for
devs
and
rpms
to
transfer
transition.
A
That
over
rno
I'm
preaching
to
the
choir,
because
you're
involved
in
some
of
it,
but
the
once
I
think,
once
we
nail
down
the
process
for
the
canonical
artifacts,
the
next
bit
will
be
saying:
okay.
Well,
so
I
have
an
isv
account
from
the
open,
build
service,
folks,
we'll
spin
off
a
poc
sub
project
for
that,
and
I
think,
as
as
we
get
that
flow
down
for
for
our
canonical
artifacts,
then
and
then
it'll
say
like
okay.
A
This
is
something
that
we
can
potentially
open
up
to
subprojects
and
we
can
spin
spin
things
off
of
that.
Assuming
that
open
build
service
is
the
thing
that
we
end
up,
going
with
on
the
on
the
dev
and
rpm
side,
for
the
windows
side
and
for
the
mac
os
side,
I
think
yeah.
I
personally
have
not
started
to
explore
the
the
windows
side.
A
The
artifacts
that
we
are
publishing
for
windows
currently
are
likely
in
the
unsigned
state,
and
I
know,
there's
a
separate
build
process
for
that
as
well,
build
and
sign
process
for
that.
So
it
is
on
our
roadmap.
So
let's
look
like
hand
waves
on
our
roadmap,
but.
C
C
B
Looking
to
jump
in
yeah
the
concerns
that
jordan
raised
a
hundred
percent,
that's
one
of
it.
It's
also
making
sure
that,
specifically,
when
we're
going
into
developer
programs,
so
like
apple's
developer
program,
there's
a
leak,
that's
a
contract
that
is
being
signed
with
apple
and
has
obligations
that
the
entity
that's
signing
it.
If
that
is
the
cncf,
that
the
cncf
has
to
ensure
are
happening
to
meet
their
legal
obligations
to
apple
because
it
is
a
contract.
B
So
the
money
here
isn't
the
issue
at
all.
The
the
stickier
part
of
a
problem
like
this
is
the
legal
side
of
it
and
making
sure
that
all
of
those
parts
make
sure
that
our
obligations
to
apple
are
being
met
to
participate
in
that
developer
program
or
if
we
need
changes
to
the
contract
that
we're
signing
with
apple
like
making
that
that
happen,
and
that
would
be
the
same
for
any
other
developer
program
that
we're
going
into
whether
that's
like
windows,
code
sign
or
that
kind
of
stuff.
B
When
we're
dealing
with
dev
rpms,
that's
something
different,
but
because
we're
only
asserting
our
own
identity
there
we're
not
entering
into
a
contract
or
something
right.
You'd
be
able
to
sign
devon
rpms.
The
one
thing
that
I
would
suggest,
though,
or
no
is,
I
would
not
at
least
from
the
steering
side.
B
I
would
not
lump
these
in
to
umbrella
issues,
umbrella
issues,
big
issues
that
have
multiple
moving
components.
Individual
components
of
them
can
get
lost
fairly
easily.
If
kate's
infra
wants
to
have
an
umbrella
issue.
That
links
out
to
these
specific
asks
from
steering
like
the
the
apple
developer
program-
one
that's
totally
okay,
but
if
there's
other
specific
individual
ass
like
hey,
we
want
to
participate
in
x
or
y
developer
program.
B
I
would
suggest
that
we
open
separate
individual
issues
for
those,
because
they
will
likely
move
at
different
speeds
based
on
the
legal
stuff
and
how
the
legal
stuff
moves
forward.
So
I
would
suggest
keeping
the
apple
one
specific
to
apple.
So
hopefully
we
can
have
a
clear
reference
point
that
other
folks
like
when
we're
engaging
folks
at
the
cnc,
app
or
lf,
or
anything
like
that
that
it's
okay.
This
is
specific
to
our
apple
developer
program
issue
and
any
others
that
folks
are
asking
for
us
to
participate
in.
E
Okay,
so
from
what
everything
god
say
here,
god
said
here,
like
my
understanding
is
someone
I
mean
steering
need
to
reach
out
to
cncf,
about
legal
compliance,
for,
let's
say
mac
os,
for
let's
say
apple
developer,
program
and
microsoft,
developer
program
and
get
back
to
us
before
we
move
forward
with
this.
Oh.
B
So
for
the
apple
developer
program-
one
specifically,
yes,
we've
got
an
issue.
That's
on
our
radar.
We
have
talked
to
a
number
of
folks,
like
paris
has
followed
up
on
it.
Stephen's
been
talking
about
it
like
we've
got
multiple
actions
in
progress
there
for
any
others
like
I
don't
have
any.
I
don't
think
we
have,
or
at
least
I
haven't
seen
any
issues
specific
to
a
microsoft,
developer
program
or
or
like
windows
code
signing
that
we
would
want
to
do
or
participate
in.
B
So,
if
there's
an
ask
there
from
some
group,
whether
it's
kate,
samfra
or
some
other
group,
that's
coming
to
kate
zimfra,
yes,
we
would
need
specific
details
of
what
is
being
asked
for
so
that
we
can
investigate
it.
Obviously
we
have
you
know,
friends
and
colleagues
over
at
microsoft
that
we
can
do
similar
things
to
what
we're
doing
with
our
friends
and
colleagues
at
apple.
A
Okay
and
if
yeah
and
if
the
windows,
the
signing
stuff,
is
starting
to
bubble
up
into
conversation,
please
like,
let's
have
a
let's
continue
to
have
some
discussion
in
sig
release,
because
ultimately,
that
will
be
part
of
our
build
pipeline.
E
A
D
Can
I
have
sure,
30
seconds
so
one
this
just
triggered
something
in
my
mind
that
me
and
anna
who
should
have
brought
up
before,
but
we
haven't,
so
we
did
a
poc
for
cloudflare
for
binaries
and
because
of
the
stuff
that
was
happening
with
cloudflare
and
kiwiforms.
D
We
kind
of
like
put
it
on
eyes.
For
now
we
probably
have
to
go.
Do
a
poc
for
some
other
provider,
because
you
know
we.
D
You
know
on
how
they
conduct
their
business,
etc.
So
you
know
we
thought
we
should
let
you
know
that
that
poc
is
on
ice
for
now.
A
But
yeah
we
should.
We
should
toss
up
some
some
alternatives
there,
but
overall
I
mean
outside
of
that.
The
it
sounds
like
the
registry
stuff
is
moving
forward
nicely
right.
E
My
take
on
this
is
to
completely
drop
cloudflare,
but
the
the
the
problem
we
have
right
now
is
we
want.
We
ultimately
want
to
use
the
cdm
provider
and
we
failed
to
reach
out
too
fastly
about
this.
Like
I
tried,
I
mean
three
times
to
reach
out
to
them
and
I
got
I
got
no
answer
for
them
so
through
the
open
source
program.
So
I'm
trying
to
find
a
way
to
reach
out
to
them
and
say
we
are
not.
We
are
an
open
source
program.
E
D
And
the
contact
that
we
were
talking
to
is
moved
on
from
cloudflare
too,
so
we
really
don't
need
to
do
anything
on
cloudflare
site.
D
I
think
I
pinged
chris
once
he
hasn't
responded
so
I'll
ping
him
again
this
week
to
get
us
a
contact.
You
know,
introduction
through
email
to
you,
know
you
and
a
contact
in
fast
league
who
can
help
okay.
E
Okay,
crazy
stuff:
I
see
sound
grace.
B
Yeah
I'm
wondering
if,
like
another
another
possibility,
like
I
remember
one
previous
point-
that
we
talked
about
this-
we
kind
of
outlined
a
list
of
what
our
requirements
are,
or
we
requested
that
a
list
of
our
requirements
be
outlined
if
we're
kind
of
back
to
a
vendor
selection,
poc
kind
of
stage.
D
If
we
kind
of
know
what
we
want,
it's
just
that:
how
do
we
do
it
using
their
ui
using
their
cli
is
basically
the
question
so
and
what
we
are
trying
to
do
here
is
like
do
the
poc
validate
that
what
we
need
works
and
it
works
the
way
we
would
like,
and
then
we
write
up
a
thing
to
take
it
wherever
else
we
need
to
take
it
to
it
could
be
the
cncf.
D
B
In
particular
like
if
we
are,
if
we
have
requirements
for
a
service,
if
there
is
somebody
at
the
the
the
cncf
like
there's,
I
forget
what
it
was
called,
but
there's
basically
like
the
cloud
credits,
infrastructure
support,
whatever
kind
of
program
that
the
cncf
was
was
doing
even
outside
of
that
particular
thing
going
and
talking
to
the
scenes,
because
there
are
other
what
I.
B
What
I
know
is
that
there
are
other
folks
at
the
cncf
level,
including
folks
that
sit
on
the
gb
who
operate
cdns
at
a
global
scale
beyond
just
google,
which
was
the
our
our
current
infrastructure
provider.
So
if,
because
there's
there's
two
parts
to
any
anything
like
this,
it's
the
cost.
It's
the
cost
to
us,
the
cost
of
the
cncf.
That's
one!
B
That's
one
aspect,
and
then
the
other
is
is
the
technical,
and
if
we
have
outlines
of
what
our
technical
requirements
are
like
how
we
need
to
interact
with
the
cdn,
you
know
what
kind
of
numbers
we're
looking
at
pushing
through
it
all
of
that
kind
of
stuff.
If
we
go
to
folks
and
say
like
it
might
be
one
extra
step
that
we
could
do
as
opposed
to
presupposing
to
a
specific
vendor
that
we've
we're
already
kind
of
like
shooting
for,
say,
hey,
is
there
anybody
out
there
that
is
able
to
support
our
requirements?
B
Come
talk
to
us
because
I
know
I'm
looking
at
the
the
cncf
membership
page
right
now.
I
know
there's
other
companies
that
operate
cdns
and
that
kind
of
stuff
that
we
could
potentially
ask
for,
because
the
the
money
part
and
the
the
technical
part,
I
don't
know
if
the
technical
part
they'll
be
able
to
match
our
requirements,
but
from
a
money
part
they
might
be
willing
to
to
donate
or
support
us
with
some
of
those
services
yeah.
I.
A
Would
say,
like
you
know,
the
any
any
poc
like
outside,
of
just
identifying
the
technical
requirements
to
kristoff's
point
getting
a
list
or
a
list
of
potential
vendors
once
you
have
those
technical
requirements,
it
does
sound
like
we
were
like
okay
flair.
Let's
do
it,
but
getting
that
list
together
because,
like
all
of
us,
have
the
potential
to
reach
arms
into
cncf,
lf,
open,
openssf,
so
on
and
so
forth.
To
have
these
conversations.
D
Oh
sorry,
one
data
point
also
here
is
the
cloud
provider.
Stuff
is
not
yet
real.
D
There
is
on
a
meeting
next
week
or
whenever
I
think
this
week,
where
the
google
person
is
wanting
to
meet
with
priyanka
and
hippy
hacker
and
chrissy
so
I'll,
let
you
know
when
I
hear
you
know
if
something
comes
out
of
that,
arun
was
also
invited
for
that
meeting.
Can
you
say.
C
The
cloud
provider
stuff
just
for
the
notes,
the
crowds
programs.
A
E
Yeah
just
want
to
add,
like
we,
we
have
just
one
technical
requirement
for
the
specifically
for
cdn
and
when
I
did
the
investigation,
the
the
company
is
part
of
the
governance,
but
we're
not
meeting
that
requirement.
So
it's
like
a
technical.
I
don't
know
if
I
need
to
go
in
detail
about
this.
Basically,
google
has
a
partnership
with
specific
cloud,
cdn
providers
and
that's
not
that's
all
technical
requirement.
So
if
we
want
to
basically
pick
a
cdn
provider
this
product,
this
provider
need
to
be
in
partnership
with
google.
A
E
So
we
were
like
okay,
let's
pick
cloudframe
faster
and
see
how
this
is
happening,
also
fastly
as
like
a
partnership
with
linux
foundation
because
they
basically
host
the
kernel
site
in
the
distribute
the
linux
kernel.
So
I
was
hoping
through
cncf.
We
can
reach
out
in
this
foundation
about
this,
but
I
didn't
hear
about
any
anyone.
Science.
A
C
B
A
If
we
know
that
these
are
the
only
providers
that
are
like,
if
if
it
is
true
that
these
are
the
only
providers
that
will
give
us
the
kind
of
it
will
satisfy
the
requirements,
then
I
I
would.
I
would
imagine
that
the
next
step
is.
Do
we
have
contacts
at
these
providers
and
then
we
start
tapping
the
network
so
between
the
so
between,
like
if
our,
if
our
technical,
so
from
the
the
the
casein
for
side.
A
If
our
technical
requirements
are
clear
and
they're
documented
somewhere
that
we
can
hand
off,
and
if
it
is
really
just
that
that
cdn
interconnect
page
then
we've
got
the
list
of
providers
and
then
from
there
steering
to
tap
networks.
E
E
A
All
right,
all
right
so
end
of
meeting
we
are
at
the
end
of
the
meeting.
Are
there
any
topics
of
interest
for
either
the
next
steering
committee
meeting
or
with
the
time
that
we
have
left.
A
Yeah
this
is
this
is
all
either
in
flight
or
done
so
for
in
between
this
meeting
and
next
I
will
clean
up
the
remaining
topics
here.
D
I
had
one
I
think
I
was
supposed
to
transfer
google
sc1
to
steven,
and
I
forgot
to
do
that
like
so
we'll
figure
out
like
you're
in
dublin.
Now
when
you
come
back,
we
will
get
that
started.
Okay,.
A
B
A
I
think
we
had
a
skew
of
one
week
for
us
today
right
so
next
week.
I
have
signed
up
for
the
public
and
private
session
for
september,
but
it
would
be
good
to
get
boson
for
october
if
someone
wants
to
take
both
sessions.
If
someone
wants
to
take
one,
we
should
also
we've
also
got
kubecon
coming
up,
so
I'm
not
sure
if
our
yeah.
D
Yeah
yeah,
that's
what
I
was
going
to
say
that
does
the
date
look
all
right.
You
know
private
announcement
of
to
ac
members,
not
up
for
election.
D
That's
the
30th
and
then
the
there
is
a
october
2nd
candidates
october
3rd
at
the
next
public
steering
meeting.
So
we
we
do
have
something
already
scheduled
for
october
3rd
public
meeting.
B
E
D
A
B
A
B
Well
then,
bozen
for
those
meetings
if
based
on
who's,
going
to
still
be
here
in
october.
I
think
it's
probably
if
stephen
doesn't
want
to
do
back-to-back.
It
might
be
up
to
you
and
me
bob
to
rock
paper
scissors
for
who
takes
october.
Okay,
I
can
definitely
take
the
first
one.
Don't
know
about
the
second
one,
because
I
am
up
for
election
yeah
area,
fair.
B
B
A
Cool
cool,
thank
you
kristoff,
all
right,
so
thank
you.
Everyone
for
the
time.
Thank
you
for
hanging
out
with
us,
see
you
at
the
next
one
bye,
doctors,
all.