►
From YouTube: Community Meeting, September 28, 2023
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
Okay
and
I
think
we're
live
so
welcome
to
the
kcb
community
meeting
just
before
we
get
started.
As
always,
we
have
a
code
of
conduct.
Please
please
follow
it.
Please
be
excellent
to
each
other.
A
With
that
first
point
on
the
agenda,
so
we
a
while
back
actually
I
think
one
or
two
months
already
time
flies.
We
talked
about
moving
the
meeting
notes
to
Google
Docs
and
we
now
have
a
document
for
that,
because
we
also
still
need
to
update
the
website
and
all
of
that
I
chose
to
open,
like
essentially
one
last
GitHub
issue
for
the
community
meeting,
but
I
believe
I
have
transferred
everything
already
over.
A
The
document
should
be
public.
Let
me
share
the
link
on
the
chat
already
and
it
will
then
also
be
published
on
the
website
and
probably
on
the
GitHub
readme,
wherever
we
can
put
it
for
for
visibility.
A
And
then
let
me
quickly
check
the
community
meeting
GitHub
issue.
Yeah,
okay,
looks
good.
Okay,
then
I
will
share
the
agenda
documents.
A
And
that
was
already
my
first
agenda
topic:
the
new
community
meeting
minutes.
We
we
discussed
this
and
before
so
I
think
everyone's
in
agreement.
But
are
there
any
any
comments?
Any
notes
on
the
templates?
The
template
is
also
just
something
that
we
put
together.
If
there
is
information
missing
from
it,
then
we
can
and
should
add
it,
but
it
was
the
first
start
to
to
get
the
documents
running.
A
Okay,
then
I
believe
that
topic
it's
over,
so
in
that
case
I
would
hand
over
to
Stefan
for
a
rebase
update.
B
Yeah
so
I
put
everything
there
I
reported
and
select
in
the
last
days.
So
MJ
did
a
great
start
with
share
picking
couple
of
PRS
I
continued
to
accept,
and
the
result
is
basically
that
we
have
two
branches,
one
in
Cube,
one
and
kcp
postcode
P2,
and
they
are
kind
of
cleaned
up.
So
they
have
nice,
commit
names,
separated
commits
for
different
topics
and
kcp
start
comes
up.
That's
the
most
important
Milestone,
it's
not
noisy,
it
just
stays
there
and
kind
of
works.
I
mean
I.
Just
did
manual
tests
before
next
steps.
B
Obviously
unit
tests
in
kcp,
I
haven't
tried,
so
I
will
be
out
until
Wednesday.
So
if
anybody
has
time
just
try,
I
have
no
idea
what
comes
on
and
after
unit
test,
of
course,
end-to-end
tests
in
kcp
as
well.
Just
as
a
reminder,
we
don't
maintain
the
testing
Cube
at
the
moment.
Maybe
we
should
and
we
come
near,
what
will
this
goal?
But
it's
all
we
don't
just
all
trying
they
don't
work.
The
only
thing
Cube
I
try
to
compile
this
Cube
API
server.
This
was
my
my
test
example
anyway.
B
So
unit
tests,
end-to-end
testing,
kcp
and
Upstream
those
PRS
which
are
Upstream
PRS,
which
are
basically
all
but
they're,
not
carried,
so
they
don't
have
to
carry
label.
It's
a
great
message,
so
a
couple
those
which
have
a
number
they
have
a
PR
there's
a
number
already,
and
it's
one
big
one.
This
is
a
generic
control,
plane,
rebase,
nothing
based
so
split
in
Upstream,
so
splitting
the
API
server
into
more
components.
It's
done
in
my
my
working
power
spr,
it's
not
Upstream!
Yet
I
will
do
that.
B
After
the
long
weekend,
I
guess
I
tried
with
one
there's
this
big
PR
from
ND
when
she
got
merged
in
April
or
so
in
in
Cube,
but
it
got
converted
because
there
was
an
issue
and
I
opened
the
pr
today,
adding
tests,
adding
unit
tests,
fixing
things
in
the
conversion
and
there's
a
long
threat
in
in
secret
machinery,
where
you
can
just
meet
up
last
words
there
in
this
channel
yeah.
This
is
the
report.
Basically,
so
I
don't
think
it's
really
support
of
those
PR.
B
B
So
I
have
to
cleaned
up
a
bit
slinters
complaining
at
the
moment,
but
I
hope
this
will
get
in
soonish.
Some
people
seem
to
be
out
so
Jordan
and
Egypt
is
he's
responding,
but
he
is
on
PTO.
It
seems
so
I'm
not
sure
when
this
will
happen.
This
one
is
basically
in
our
chain
of
commits
in
our
Fork
and
it's
pretty
big
and
it's
ugly.
If
we
have
to
maintain
that
getting
rid
of
that
I
think
it's
something
like
100
times
of
code,
a
thousand
line
of
codes
or
something
like
that.
B
So
if
you
can
also
even
more
okay,
that's
a
lot.
So
if
you
go
back,
I
posted
lines
of
code
of
our
rebase
Branch
once
including
everything
which
was
I,
think
8
000
lines
editor
so
and
half
of
that
removed.
B
If
you
remove
the
Upstream
PR
from
that
calculation,
you
get
way
below
5000
I.
Think
I
haven't
got
any
details,
so
you
can
do
with
a
mask
yourself
to
punch
yourself,
but
it's
getting
smaller
and
smaller
and
I
like
that.
So
yeah,
that's
important.
C
A
A
A
A
A
A
A
So
I
think
this
is
the
first
time
we're
meeting.
Actually,
since
the
proposal
to
join
sandbox
was
accepted
by
the
TOC,
so
I
don't
know
someone
has
a
little
party
hat.
That
would
be
the
time
but
yeah.
How
do
we
want
to
tackle
the
sandbox,
the
sandbox
checklist?
B
C
C
B
D
B
We
had
we
had
contact
to
somebody
a
product
manager
from
from
redheads
in
this
regard,
so
we
had
to
talk
with
her.
You
know
a
few
months
ago,
Sebastian
myself,
and
we
should
contact
her
again,
so
she
promised
that
that
her
people
will
do
all
this
legal
things.
A
Yeah,
that's
great
news
but
yeah
I
guess
the
question
is:
do
we
want
to
go
over
them
in
the
community
meetings
or
do
we
want
to
discuss
it
async
on
the
slack
Channel
and
then
maybe
bring
like
bi-weekly
updates
to
the
checklist
to
this
meeting.
C
I
would
go
for
the
latter
one
and
basically
like
I,
think
two
or
three
Community
meetings
down
the
line
we
will
read.
The
state
was
missing.
I
think
we
just
need
to
engage
earlier,
identify
the
ones
which
could
drag
on
an
example
the
domain
transferring
stuff
legal
stuff
which
needs
strategory
engagement.
We
might
need
to
start
engaging
earlier,
but
most
other
ones
should
be
quite
quick.
B
A
A
A
Looking
at
the
list,
I
guess:
yeah:
okay,
let's,
let's
track
that
in
the
documents
and
then
we
can
also.
We
can
also
Mark
the
things
that
we
believe
are
already
fulfilled
and
where
we
need
to
do
work
and
I
think
once
we
have
that
kind
of
Discovery,
then
we
can
move
on
and
start
working
on
the
things
that
we
that
we
can
Implement
like,
for
example,
I,
believe
there
is
an
item
to
to
add
the
sandbox
logo
to
the
website.
So
that's
things
we
can.
A
We
can
do
or
prepare
just
for
my
quick
understanding
is
the
project
already
part
of
the
sandbox
or
do
we
need
to
fulfill
the
checklist
to
become
part
of
the
sandbox.
B
B
A
A
D
B
A
C
C
C
B
B
C
B
C
B
Okay
yeah,
so
what
about
so?
Do
we
have
to
assign
anything
or
can
we
assign
anything
already.
A
I'm
not
exactly
sure
how
we
would
do
this.
To
be
honest,
maybe
we
maybe
we
look
at
other
onboarding
issues
and.
A
Yeah,
okay,
so
the
whole
understanding
things
here,
I
would
say:
okay,
someone
needs
to
read
those
and
maybe
give
us
it
the
gist
of
it.
We
have
written,
I,
guess,
open
governance.
Someone
would
need
to
check
the
criteria.
A
We
would
need
to
submit
a
project.
Sorry,
a
pull
request,
I'm,
not
sure
to
reach
Reaper,
but
I
think
that's
something
we
can
figure
out.
A
Then
the
maintainer
list,
also
by
your
PR,
also
something
that's
durable,
the
artwork
and
the
pull
request,
something
that
I
believe.
Maybe
we
should
wait
for
so
you
said
that
the
the
artwork
is
owned
by
us
right.
B
A
Then
that
that's
already
doable
as
well
the
domain
transfer,
we
need
to
discuss
those,
that's
it
emails
for
the
maintainers
yeah.
We
can
provide
those,
then
trademarks.
Okay,
let's
skip
that.
For
now.
We
can
add
those
two
to
the
GitHub
org
that
shouldn't
be
a
problem
either
dco
I
believe
we
have
enabled
already
code
of
conduct
I'm,
not
sure,
if
you're
using
the
cncf
template
already.
If
we
need
to
adjust
something.
A
Then
the
footer
we
can
add
that
to
the
website,
maybe
wait
with
merging
the
pr
into
the
web.
I,
don't
know
pretty
well
into
the
process.
I
don't
know,
but
like
I
can
prepare
that
analytics
transferred
I,
don't
believe
we
have
analytics
right
now.
What.
A
A
It
seems
to
be
ends,
then
that
I'm
I'm
I
haven't
looked
into
open
ssf
yet
but
I
guess
this
is
also
something
we
can
do.
Maybe
once
the
maintainers
have
been
on
board
and
have
access
to
the
cncf
infrastructure,
I
don't
know.
A
B
Think
so
yeah
yeah
MJ,
if
you
can
start
a
conversation
with
Karina
with
me
and
Sebastian.
A
If
there's
something
that
you
need
from
Cuba
Matic
more
urgently,
you
can
also
reach
out
to
me,
but
I'm
not
sure
what
that
would.
B
Be
to
be
honest,
but
I
mean
website
stuff
right,
as
you
can
just
do
on
your
own.
If
anybody
has
time
to
look
through
the
documents
like
this
code
of
conducting
whatever
listed
here,
all
this
paperwork
we
have
to
put
into
the
repository
and
I
think
we
have
taken
every
single
cncf.
So
maybe
it's
really
no
work
at
all.
Just
checking.
A
A
Okay,
cool
cool,
then,
let's
take
a
quick
look.
We
already
talked
about
this
topic.
I
just
had
a
thought
this
morning
and
I
just
wanted
to
quickly
talk
about
it
in
our
cncf
project.
I,
guess,
congrats
to
ourselves.
The
kubecon
Europe
cfp
is
open
since
a
couple
of
weeks
again
and
I
was
just
wondering
if
people
were
thinking
about
handing
in
the
kcp
talk.
If
we
wanted
to
coordinate
that
a
bit,
so
we
don't
cannibalize
each
other
or
if
there
are
no
plans
so
far,.
B
A
Okay,
well,
that's
that
that's
a
topic,
that's
much
bigger
than
kcp!
So
I,
don't
think
a
kcp
talk
would
conflict
with
that.
A
Okay,
I'm
I'm
thinking
about
handing
something
in
but
I'm,
not
sure.
If,
if
I
can
find
the
novelty
from
I
believe
two
years
ago
there
were
a
couple
of
talks.
So
let's.
B
A
Last
year,
yeah,
that's
real
spell:
okay,
okay,
one
thing,
sorry
something
about
the
sandbox
checklist.
We
should
maybe
add
to
our
own
checklist
that
we
write
at
least
a
short
blog
post
for
the
website
about
joining
the
sandbox.
A
A
A
A
We
have
three
new
issues
actually
so
before
I
can't
count
this
one
I
can
quickly.
I
can
quickly
say
I
forgot
to
guess
put
it
into
something.
It's
just
like
our
tests
are
rather
flaky
and
that's
not
not
necessarily
because
the
tests
itself
themselves
are
unstable,
but
I,
don't
know
someone
needs
to
take
a
look,
so
I
added
it
to
the
backlog.
B
A
But
yeah
I
don't
know,
then
we
have
bug,
cannot
sync
delete,
namespace,
H,
okay,
this
one
is
older
and
it's
TMC.
So
let's
move
it
to
the
backlog
and
we
might
want
to
revisit
migrating
the
TMC
issues
to
its
own
repo
yeah,
like.
D
We
have
a
similar
issue
in
who
Stellar
I
think
there's
a
really
a
fundamental
question
in
the
kubernetes
world
about
keeping
track
of
why
namespaces
exist.
D
Yeah,
maybe
it's
easier
for
TMC
right
in
coop
in
kubernetes
in
general,
you
know
is,
is
something
once
a
name
space
to
exist
that
already
exists.
It
just
goes
merely
along
its
way
and
uses
that
namespace
it
doesn't
record
right.
We
don't
keep
track.
We
have
something
like
a
owner's
right.
D
A
little
easier
because
you're
really
making
stuff
exclusively
for
one
purpose
right
for
for
coup,
Stellar,
we're
really
saying
we're
kind
of
merging,
what's
being
requested
from
several
sources
with
what's
already
there,
for
maybe
other
reasons,
so
it's
There's,
No,
One
exclusive
owner.
In
our
view,.
A
Yeah
I
think
the
question
there
might
be
if
you,
if
you
want
to
take
ownership
of
things
a
bit
like
forcefully.
So
you
track
on
the
kcp
side.
If
you
own
something
or
if
you
believe
that
you
own
something
or
if
you
want
to
track
on
the
physical
cluster
sites
through
annotations
or
something
that
this
is
owned
by
someone
specific.
D
Yeah
and
in
the
Coop's
teller
world,
the
problem
is
so
it's
not
TMC
and
we
want
to
support
managing
clusters
that
have
stuff
you
know
from
the
kubernetes
world
in
general,
installed
and
general
stuff
for
the
kubernetes
from
the
Cooper's
world
you
know
can
make
or
start
using
namespaces
at
any
time
for
any
reason,
without
recording
the
fact
that
that
stuff
is
cares
about
them.
B
Stefan
I
guess
in
this
case
yeah
about
this
issue:
it's
probably
some
finalizer
something
blocking
it.
So
maybe
it's
a
good
question
to
ask
if
this
is
consider
relevant,
June,
9,
6,
actually.
A
Yep
yep,
that's
and
that's
probably
a
good
question.
A
Yeah
for
sure,
for
sure
just
so
we
know
what
it
is
but
yeah,
let's
see,
should
we
move,
we
can
backlog
after
asking
or
should
we
keep
it.
A
I
will
then
we
have
our
kcp
workload.
Sync
command
does
not
work
with
eks
AKs
gke
clusters.
A
A
Yeah-
and
this
is
also
multi-cluster,
so
we
had
the
TMC
label
and
then
we
can
go
on
maybe
ways
on
that
and
then
there's
one
last
from
you
Mike
about
the
cubesatel
vs
commands
and.
D
I
think
it
so
the
the
last
couple
of
comments
I
think,
really
panicked
it
to
a
larger
issue.
So
let
me
kind
of
explain
the
context
right
we
have
we
want
to
be.
We
have
a
Helm
chart
right,
like
kcp.
Does
right
so
suppose,
one
installs
this
in
a
cube
cluster
and
then
wants
to
access
it
from
within
that
Kube
cluster
and
also
wants
to
access
it
from
something
outside
right.
D
I
mean
suppose
this
is
you
know,
on
eks
or
any
other
Cloud
and
you've
got
clients
that
are
outside
the
cloud
right
and
you've
also
got
clients
that
are
inside
the
same
Coupe
cluster.
You
want
to
support
both
and
the
way
I've
done.
It
so
far
is
go
ahead
and
you
know
do
kcp
start
without
any
particular
consideration
for
this
and
then
take
the
kubeconfig
file.
The
kcp
start
produces
produces
and
edit
the
server
URLs
for
external
clients
to
put
in
the
domain
name
and
Port.
D
You
know
that
they
use
to
come
in
the
front
door
from
outside
the
cloud
and
that
works
until
they
do
Cube
cuddle
WS
tilde,
at
which
point
the
workspace
plugin
edits
the
kubeconfig
file
and
puts
back
in
the
IP
address
or
inside
the
coop
cluster.
B
This
reminds
me
I,
think
there's
something
in
Cube
which
takes
the
incoming
domain
name
like
in
the
Hep
server
knows
the
domain
names
are
Supply
and
users
right,
so
it
can
identify
whether
it's
outside
or
inside,
and
we
could
build
something
which
then
returns
a
different
home
workspace
domain.
D
So
I
noticed
that
kcp
explicitly,
you
know
started
inherited
with
inheriting
all
the
command
line,
arcs
of
cube,
API
server,
except
it
deleted
one
very
relevant
one
which
was
I,
think
the
one
that's
called
bind
address
but
anyway,
so
the
issue
I
wanted
to
raise.
You
know
I,
think
I,
think
you
know
the
important
point
right
is
to
support
the
use
case,
one
way
or
another.
So
if
people
are
agreed,
the
use
case
could
should
be
supported.
Okay,
then
the
next
case
is
okay.
D
C
C
D
No
I
had
not
consider
that
if
you
could
put
it
in
a
suggestion
in
the
you
know,
any
issue
I'll
see
if
I
can
make
that
work.
It's.
D
Over
I
assume
that
the
helm
chart,
though,
doesn't
do
that
yet
right.
C
D
A
Okay,
I
I,
do
think
that
it
makes
sense
to
think
about
this
from
the
client
perspective
or
from
the
plug-in
perspective
overall
as
well,
because
the
host
aliases
our
way
to
get
it
working,
but
maybe
what
Stefan
described
should
still
be
supported
in
a
way
but
yeah
for
now.
I
think
it
makes
sense
if
there's
a
workaround
that
can
help
Mike
I
think
that's
good
and
I
would
move
it
to
backlog.
For
now.