►
From YouTube: Kubernetes SIG CLI 20200617
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,
good
morning,
good
evening,
good
afternoon,
depending
on
where
you
are
today,
is
June
17th,
and
this
is
a
matter
of
our
bi-weekly
60
line
meetings.
My
name
is
Marty
and
I'll,
be
your
host
today
and
I
think
we
have
agenda
pre,
packed
with
a
lot
of
discussions,
at
least
for
the
second
topic,
but
before
we
move
through
that
one
to
announcements,
there's
a
blog
about
call
next
Wednesday
exact
same
time,
exact
same
number.
A
A
The
another
important
news
for
119
release
is
the
date
has
changed
for
their
long.
The
new
code
freeze
date
is
July,
9th
and
I
length
in
the
agenda
to
all
of
the
dates
but
I'm
guessing
the
majority
of
you
are
interested
in
and
the
code
freeze
is.
This
gives
about
two
additional
two
or
three
additional
weeks
between
the
original
code
freeze
date
and
the
current
one
Sean.
Do
you
have
any
other
additional
announcements
that
you
want
to
add?
A
D
So
hi
I
am
you
honest.
I've
worked
closely
with
Alex
and
Paul
regarding
the
feature
we
recently
merged
in
customized
customized
components
for
today's
agenda.
We
would
like
to
have
an
update
regarding
the
kubernetes
enhancement
proposal
that
we
have
submitted
from
what
I've
seen
it
has
been
very
good.
We
did
some
minor
changes
that
were
needed
and,
if
I
pull
correctly,
we
had
said
like
wait
for
about
a
week,
so
that
the
community
could
see
this
get
more
closely
and
possibly
raise
any
issues
with
it.
D
D
A
D
E
Is
my
mic
working
yep?
Okay,
so
it
looks,
it
looks
good
to
me.
I,
don't
actually
have
the
ability
to
approve
caps
in
the
in
its
particular
repo,
but
I'm
excited
to
see
the
work
it
going
and
looking
forward
to
seeing
the
implementations.
A
D
E
No
I'm
pulling
up
the
the
enhancer
now
and
I'm,
just
gonna
tell
G
Tim
it
if,
if
I
haven't
already
and
that
will
get
merged
and
then
the
next
thing
is
to
put
out
a
new
release
of
customize
at
the
end
of
this
week.
So
Friday,
if
you
have
any
changes,
you'd
like
to
make
before
then
go
ahead
and
send
some
pr's
out
I
noticed,
is.
Is
this
Alex
speaking?
E
You
also
have
a
proof
of
concept
for
the
running
of
the
kept
style,
config
functions,
which
is
interesting
too
so
yeah,
nothing,
nothing
remains
to
be
done,
except
whatever
you
might
want
to
get
in
any
PRS
you'd
like
to
get
in
before
Friday,
and
you
know
we're
trying
to
release
on
a
weekly
basis,
or
at
least
bi-weekly.
So
if
it
doesn't
get
in
this
week,
you'll
get
it
next
week.
Okay,
so.
E
B
E
E
I
can't
do
it,
then
maybe
somebody
else
can
take
up
the
work.
I've
got
a
lot
of
other
things
that
come
up
lately,
but
yeah
I'm
working
on
refactoring
and
to
get
rid
of
the
API
machinery
stuff
so
that
we
can
reintegrate
it
in
the
coop
cuddle.
That's
a
separate
issue
right,
that's
separate
from
me!
I'm
thinking
go
for
it
anyway.
Okay,.
D
The
reason
we
asked
about
the
Catalan
decoration
is
because
we've
been
using
the
extensively,
both
customizing
cute
couple
for
both
development
and
production
employments.
So
we
would
really
like
to
to
work
with
like
a
vanilla
to
cuddle
and
get
this
picture
out
of
the
box.
So
for
now
I
guess
we
have
to
fall
back
to
just
to
customize,
build
pipe
I.
Keep
that
lovely
right.
D
A
A
F
A
Cube
cuddle
I,
it's
it's
more
of
a
number.
A
Than
number
of
weeks
or
months,
so
the
number
is
pretty
low
because
it's
mostly
one
or
two
commands,
but
the
problem
is
that
they
are
touching
functionality
which
is
being
shared
across
the
entire
kubernetes.
That's
the
auerbach
utilities.
I
spoke
with
sig
off
Jordan
and
separately
and
yeah.
We
just
I
just
need
to
find
the
time
and
push
this
over
the
hump.
A
A
G
Yeah,
it's
just
it's
just
one
command,
actually,
that
the
one
that
matcha
is
mentioning
the
the
auth
reconcile
using
our
back
the
conversion.
We
actually
deprecated
back
in
one
or
announced
deprecation
back
in
113,
so
we
can,
we
can
take
out
convert
whenever
we
want.
Basically,
we
haven't
done
it
just
because
we
have
this
other
command
blocking
it.
So
that's,
basically
all
that's
remaining
it's!
It's
actually.
F
A
We
won't
be
able
to
add,
convert
in
any
way
back,
and
the
reason
for
that
is
that
converts
relies
on
the
internal
api's,
so
whatever
we
will
do
if
we
basically,
the
plan
was
and
still
is,
that
we
will
keep
convert
as
part
of
the
queer
ladies
repository,
but
it
will
be
built
as
a
plugin
to
keep
cuddle.
So
if
you
want
to
have
convert
functionality,
you'll
have
to
get
it
from
the
main
commodities
repo,
rather
rather
than
as
a
built-in
command
for
cube.
Caudill.
G
F
A
F
F
A
G
G
With
a
recent
coop
control,
you
get
very
very
verbose,
output
and,
of
course,
everyone
has
their
opinions
on
this,
and
so
the
question
is
that
has
been
put
to
us
is:
what
are
we
gonna
do
about
it?
We're
gonna
turn
this
off
by
default
or
give
us
some
flags
to
turn
it
off.
So
actually,
if
Antoine's
here,
do
you
want
to
pipe
in.
H
Yeah
I
don't
have
much
to
add
its
it's
just
to
give
more
details
and
context.
It's
sort
of
replacing
the
last
applied
annotation,
but
every
one
that
anybody
would
touch
is
the
object,
is
gonna
have
its
own.
So
here
is
my
first
and
last
applied
annotation.
Also.
The
way
we
plaintiff
is
it
takes
more
space
on
the
screen,
because
it's
a
one
thing
per
line.
Instead
of
being
this
condensed
JSON
that
we
have
followed
a
stoplight
annotation
people
seem
frustrated
with
seeing
that
much
we've
been
between,
because
it
has
important
information
and.
H
We
also
believe
that
criminals
should
paint
the
objective
wait
saved
by
the
apos
over
in
general
and
that's
about
it.
So
we
would
rather
not
change
the
default
of
showing
it,
but
we
are
completely
open
to
having
an
option
to
hide.
It
and
I
mean
we're
open
to
pretty
much
anything,
because
we
want
people
to
be
happy
so.
A
I
must
admit:
I
was
pointed
to
this
issue.
Just
recently
after
I
opened
a
change
to
the
cube
Caudill
edit,
because
as
long
as
and
I
agree
with
what
Daniel
and
Antoine
mentioned
and
the
thread
for
cube
Carol
get
where
you
explicitly
asked
for
yum
or
JSON
of
the
resource,
we
should
put
a
complete
dump
of
the
resource
as
it
exists
on
the
server.
So
there's
no
talking
about
by
default
at
least
there's
no
talking
about
removing
those
fields
from
there.
I
really
like
I,
think
trend
when
Daniel
a
couple.
A
A
That's
true:
well,
theoretically,
you
could
exclude
them
through
either
using
some
kind
of
a
JQ
or
y
q--
query,
not
sure
if
they
support
negation
negation,
but
maybe
I'm
I
might
be
missing
something
yeah.
So
that's
definitely
that
part
is
still
open
to
discussion,
whereas
for
edit
describe
I'm
pretty
confident
that
people
would
be
okay
with
having
the
managed
fields
completely
hidden
and
in
describe
and
honestly
I
would
not
expect
describe
to
showed
the
managed
fields
because
describe
is
a
very,
very
specific.
A
We
don't
want
to
see
them
most,
probably
I'm
open
to
suggestions
where
this
would
be
an
optional
flag
for
edit
and
not
a
default
or
in
the
other
direction,
where
we
would
have
a
fallback
to
get
the
entire
resource
as
it
exists
on
the
server.
So
you
would
have
to
opt
in
to
seeing
the
manage
fields,
and
maybe
some
others
in
the
future.
H
You
know
so
my
general
argument
for
the
idiot
is
that
I
think
I've
seen
suggestions
in
the
past,
you
say:
do
a
cooker,
ologies
and
change
the
last
applied
annotation.
If
you
have
a
gun
and
yeah,
that's
unfortunate,
I
know
but
I'm
afraid
we're
may
end
up
in
the
same
situation.
But
again
in
this
case,
maybe
we
people
could
specify
an
option
if
they
need
to
edit
this
specifically.
H
A
When,
when
I
was
putting
together,
this
PR
I
started
wondering
whether
our
cubicle,
edit
should
switch
to
the
server
side
apply
instead,
currently
to
cuddle
edit
calculates
the
patch
on
their
own
and
sends
only
instance
only
a
patch
which
ends
up
not
being
reflected
properly
in
the
manage
fields.
If
we
would
be
using
the
server
side
apply,
the
modification
coming
through
cube,
Caro
added,
would
be
properly
reflected
in
the
resource
history,
so
I
think
that's
that's
a
senator.
H
So,
first
of
all,
we
can
do
a
solo
side
apply
from
the
entire
object,
because
we
wouldn't
know
what
fields
you
care
about
and
what
field
you
don't
care
about.
When
you
do
an
edit,
you
have
the
full
object
and
we
don't
want
you
to
own
all
the
objects,
all
the
fields
in
the
object
and,
second
of
all,
we've
considered
the
case.
H
Well,
somebody
is
doing
an
edit
and
changing
an
object,
and
then
you
apply
and
try
to
change
back
this
field
you're
going
to
end
up
with
a
conflict,
and
this
is
a
use
case
we've
seen
well.
Maybe
sies
are
like
trying
to
edit
the
object
to
fix
something
and
later
you're
trying
you're
trying
to
apply,
and
this
is
going
to
stomp
the
change
and
break
the
thing
again.
So
we
want
a
conflict
in
this
case,
I
think
it's
working
as
inside
yeah.
A
A
A
Somehow,
but
I've
noticed
that
a
lot
of
people
came
up
already
with
with
their
own
options
for
removing
them.
There
are
couple
cube,
coddled
plugins
that
does
just
that
and
since
I
have
noticed
that
people
were
complaining
about
export
being
removed.
This
use
case
is
another
support
for
that
these.
Why
we
removed
the
export,
because
there
will
be
use
cases
where
you
do
want
to
see
the
manage
fields
in
your
export,
and
there
will
be
other
cases
where
you
don't
and
an
export
is,
is
usually
very
subjective
to
your
particular
use
case.
A
A
A
Api,
it
is
pretty
well
documented.
It's
part
of
the
metadata
metadata
maida
one
object
meta.
So
if
you
do
OC
sorry,
caveat,
I'll
explain,
for
example,
deployed
metadata
that
manage
fields
you'll
get
it
in
full
information
about
how
the
resource
looks
like
this
weather.
Is
there
or
just
you
know,
type
into
any
resource
and
peek
into
what?
Let's
under
made
a
few
on
object?
Later
great.
Thank
you.
I
think
it's
object
matter,
not
type
meta,
but
I
might
be
wrong.
One
of
the
two.
A
C
Yeah,
it's
very
quickly.
We
released
a
new
version.
We've
been
working
for
a
while
now
on
Jen
think
how
we
can
do
a
better
job
in
helping
user
understand
the
ongoing
sort
of
status
of
you
know
all
the
resources
that
they
have
in
their
cluster,
so
I
have
a
little
screenshot
in
the
in
the
in
the
Google
Doc.
So
at
least
now
there's
a
jiffy
also
on
the
release
notes
for
the
get
up
baby
there
is
that
when
you
do
cook
go
get
W
instead
of
getting
a
plain
table.
C
You'll
now
get
a
little
sort
of
heat
map
of
the
resources
that
you
specified,
whether
it's
my
labels
doctor
by
resource
kind,
whatever
it
is,
you
can
always
revert
it
to
a
table
if
you
want,
but
then
that
table
belie
the
cells
will
change
colors
as
the
resource
statuses
change.
So
we
throw
that
out
there.
It's
we've
been
doing
some
internal
testing,
but
we
wanted
to
get
some
extra
feedback,
so
we
threw
it
out
there
and
see
what
people
think
that's
the
biggest
change.
C
C
In
general,
if
you
guys
have
any
sort
of
insights
into
how
we
can
improve
the
kind
of
situational
awareness
of
users
with
respect
to
their
cluster,
this
kind
of
watch
being
ID
was
sort
of
our
first
sort
of
staff
in
that
direction.
But
in
general,
few
guys
have
any
thoughts
about
situation
where
it's
generally,
how
I
can
no,
sir
than
a
large
scope,
six
thousand
foot
kind
of
perspective.
What's
going
on
Ben
source
resource
status,.
A
Unfortunately,
within
create
that
audience
does
not
work
which
forces
me
to
use
the
long.
Long
name
like
I
need
to
type
fully
keep
cutting
out.
I
started
wondering
whether,
when
I'm
inside
of
cooee,
whether
I
would
want
to
type
just
internal
cubicle
commands,
as
in
just
get
pods,
get
whatever
in
just
a
command
without
the
cubicle,
prefix
or
I
would
want
to
have
the
aliens
being
supportive.
F
A
He
has,
he
has
a
generator.
That
literally
replaces
every
single
cute
command
with
an
alias,
pretty
smart,
alias
it's
a
little
bit
heavy
when
it
loads
initially,
but
it
can
be
pretty
effective.
I
haven't
I
I
I
did
try,
it
I
haven't
used
it
excessively,
but
it's
pretty
smart
idea,
so
maybe
when
people
will
be
coming
up
with
aliases
supporting
aliases
would
be
would
be
reasonable,
because
then
you
know,
workload
with
Kui
would
not
be
any
different
from
the
workflow.
You
have
and
they
see
a
lot
I.
Don't
really
got.
C
It
I
think
I'll
check
double
check,
but
we
definitely
put
some
fixes
in
in
8.9
for
aliases
I'll
double
check
to
see
whether
that
works.
We
also
have.