►
From YouTube: Kubernetes SIG CLI 20210630
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
First
of
all,
the
announcements
there's
a
code
breeze
in
slightly
over
a
week,
specifically
next
thursday.
If
I
remember
correctly,
and
my
calendar
foo
is
working
the
way
it
should
be
working
because
thursday
is
july
8th.
A
So
that
should
be
enough
enough
of
time
for
folks
that
have
an
extended
weekend,
this
upcoming
to
end
the
the
elements
that
the
prs
that
they
need.
There's
another
freeze
for
tests
a
week
later,
so
there
is
a
little
bit
more
time
if
you
want
to
improve
the
testing.
A
If
you
have
any
particular
prs
that
you
are
waiting
me,
eddie,
phil,
sean,
just
please
ping
us
directly
on
slack
or
in
a
private
message
or
whatever
to
get
them
reviewed
as
soon
as
possible.
I
know
that
I
own
a
couple
reviews
if
nikita
will
be
watching
that
I
will
be
looking
at
her
pr
for
sub
resources
tomorrow
morning.
A
So
it's
yes,
it's
top
of
my
list.
Currently
eddie
you
wanna,
you
wanted
to
add
the
the
announcement
about
look
good
in
the
episode
of
cube,
cuddle.
B
Yeah
real
quick
on
the
code
freeze,
so
I
know
sean
correct
me
from
wrong,
but
I'm
pretty
sure
the
googlers
have
off
monday
and
tuesday
next
week
and
I
took
off
the
entire
week
so
yeah
we're
gonna.
We
have
like
the
rest
of
this
week
really
to
like
crush
those.
A
In
I'm
in
I'm
in
all
the
time,
there's
no
pto,
there's
no
day
offs
in
poland,
so
I'll
be
working
like
crazy
until
the
next
thursday
and
after
that
I'm
gonna
pass
out,
but
but
seriously
you
can
always
ping
me
in
a
private
message.
It
will
probably
the
fastest.
I
have
a
couple
of
unread
notifications
in
slack,
so
private
messages
are
probably
the
the
easiest
to
reach
out
to
me.
B
Yeah,
and
so
the
lgtm
is
the
new
cloud
native
tv
that
they're
doing
it's
a
show
about
how
to
get
started,
contributing
to
cncf
projects,
and
so
they
asked
me
to
be
on
it
last
friday
for
kubernetes
and
cube
control
walk
through.
So
if
you
have
any
interest
in
how
to
contribute
to
kubernetes
and
you're
new
here,
that
might
be
something
worth
watching.
A
Okay,
sean
anything
other
that
you
wanted
to
add
to
the
announcement
section.
C
No,
I
I
think
that,
amongst
the
leadership
we
also
have
the
july
6th.
I
think.
D
C
Already
on
it,
for
we
have
to
put
in
a
proposal
for
maintainer
track
presentation,
we
have
kubecon
north
america
coming
in
october
and
it's
in
my
home
city
of
los
angeles
and
yeah,
and
it
sounds
like
we're
going
to
be
able
to
show
up
in
person.
So
it'll
be
really
good
to
see
you
guys
in
person.
D
C
Guys
can
make
it
if
you
feel
comfortable
with
the
protocols
etc,
and
if
we
actually
have
it
in
person,
it
sounds
like
we
are.
A
Yeah
I'll
probably
skip
this
one.
Unfortunately,
it's
a
little
bit
of
time
for
me
to
get
there,
I'm
not
sure
what
what
the
the
world
will
look
like
in
october.
A
So
I'm
my
my
personal
impression
is
that
it
might
be
just
the
us,
mostly
for
the
kubecon,
but
we'll
see
how
things
will
go.
If
I
remember
correctly,
they
are
planning
to
do
it
hybrid
approach,
so
there
will
be
some
folks
joining
virtually
and
some
other
folks
that
will
be
in
in
person,
but
six
eli
will
be
there
one
way
or
the
other
or
both
way.
At
the
same
time,
we
will
figure
it
out.
As
usual.
That's
october
12th,
through
the
15th.
B
A
Now,
I
regret
not
picking
the
kiwi
land
after
san
diego
two
years
ago.
Anyhow.
Do
we
have
anyone
new
on
this
call
that
would
like
to
introduce
themselves
tell
a
little
bit
of
who
they
are,
what
they're
interested
in
contributing
I'm
scrolling
through
the
list.
I
think
I've
seen
most
of
you,
but
if
you're
new
now
is
your
time
to
speak
up.
E
Okay,
hi,
yes,
I'm
new!
Here
I
was
thinking
about
contributing
to
a
sick
release
during
the
past
period.
So
yes,
I'm
attempting
to
know
more
about
sexy
ally
and
maybe
maybe
I
would
think
about
the
six
year
old
before
yes,.
E
A
A
Okay,
I
hear
none,
so
we
can
jump
over
to
the
main
topic
eddie.
You
want
to
take
it
over.
Take
it
away
from
here.
B
Yes,
so
this
was
a
so
if
you're
not
on
the
sig
architecture,
mailing
list,
you
should
join,
but
we
started
a
meeting
to
do
a
bi-weekly,
kep
reading
club
review.
So
we
pick
a
few
caps
and
go
through
them.
This
one
was
up
on
the
agenda
for
last
week
and
we
read
through
it
and
had
a
whole
bunch
of
questions
actually
afterwards
about
certain
pieces
in
here.
B
The
tl,
dr,
is
this
adds
a
cube,
control
logs,
node
command
type
thing,
so
you
can
view
node
logs
and
tim
hawkin
and
dims
had
a
bunch
of
questions
after
we
were
reading
this,
it
doesn't
sound
like
it's
making
this
cycle.
So
that's:
okay,
there's
no
rush
on
this,
but
there's
definitely
some
things
that
we
should
address
and
I
wrote
down
most
of
the
notes
here
on
the
agenda.
A
Yeah,
I'm
I,
it
would
be
good
to
ping
the
author,
either
of
the
pr
the
pr
is
still
in
a
work
in
progress
state
from
what
I
was
checking
before
this
call.
I
remember
looking
at
the
cap.
It
was.
It
seemed
reasonable
to
have
the
implementation,
although,
yes,
even
after
the
cap
emerged,
there
was
quite
a
few
questions
from
different
people.
I
remember
that
tim
also
commented
on
the
cap
for
different
angles:
how
to
approach
this
topic.
A
A
Gonna
make
a
122
if
the
pr
is
in
a
work
in
progress
a
week
before
the
freeze,
I'm
not
sure,
if
that's
gonna
land,
I'm
not
sure
which
part
or
how
far
they
are
with
their
pr,
whether
that's
something
that
they
are,
they
haven't
completed
on
the
cli
end
or
on
the
cubelet
end,
because
it
has
it
needs
to
have
both
sides.
B
A
Remember
correctly,
that
was
not
about
adding
a
sub
command,
or
at
least
when
I
looked
at
it.
Initially,
it
wasn't
about
adding
sub
command,
but
it
was
more
about
adding
another
resource
that
we
would
support.
So,
similarly
to
how
you
can
do
cube
catalog
logs,
I
don't
know
deployment
or
log
spot
and
specify
a
particular
resource
name.
A
It
will
work
similarly
for
knots,
notes
where
you
would
specify
keep
catalog
locks
node,
whatever
the
name
of
your
node
is
that's
at
least
how
how
I
imagine
that
initially,
the
logs
by
default
just
looks
for
node
pod
names.
Yes,
if
you
don't
specify
any
no
resource
name
we
default
to
pod
and
honestly,
I
would
prefer
we-
and
this
is
probably
just
for
historical
reasons.
A
When
we
were
adding
additional
resources
support.
We
are
always
requesting
you
to
explicitly
specify
what
is
the
resource
name,
your
your
you're
expecting
to
work
to
work,
the
locks
command
against,
so
you
can
do
catalogs
job
my
job
or
you
can
do
keep
catalogs
deployment.
My
deployment
and
the
logic
behind
cube
catalogs
is
responsible
to
figure
out
which
pot
and
it
picks
first
part
in
deployment
state
for
set.
Daemon
said
replica
said
job.
B
A
Yeah
it
if
there
are
more
questions
and
answers,
it
would
be
good
to
get
them
answered
before.
Turning
this
on
it
looks
like
the
author
haven't
looked
at
the
vr.
A
Yeah
so
the
author
a
week
ago
said
that
he
has
a
working
progress
pr,
but
it's
looking
very
tight
at
this
point.
So
even
if
the
author
is
aware
that
this
got
stuck,
then
I
I'm
not
sure
it
looks
like
it
wasn't
reviewed.
A
A
Yeah,
definitely
I'm
I'm
in
favor
of
having
this
sorted
out
before
we
actually
jump
into
implementation.
If
there
are
questions.
B
Okay,
I
will
get
the
I'll
get
all
those
comments
over
there.
A
Cool
awesome:
does
anyone
have
any
questions
about
this
particular
feature
or
comments
that
they
want
to
share
with
the
group.
A
Okay,
hearing
none
eddie,
you
have
another
topic,
ignore
delete
events.
B
So
this
one's
hopefully
quick.
This
was
the
issue
that
I
did
on
lgtm
and
I
threw
together
the
quick
solution.
The
problem
was
people
were
seeing
duplicated
events,
but
the
reality
that
I
think
brian
dug
into
was
that
these
are
really
just
the
lead
events
that
are
coming
across
60
minutes
after
they're
garbage
collected,
but
that
wasn't
surfaced
in
any
of
the
events,
so
you
just
kind
of
see
like
60
minute
old
events
pop
up
with
the
same
exact
data
in
it.
So
I
opened
a
pr
it's
a
work
in
progress.
B
I
have
a
change
I
felt
to
push
up,
but
I
added
a
so.
I
defaulted
the
command
to
not
outputting
delete
events,
because
I
feel
like
they're,
not
useful
for
the
user
and
then
I
added
a
flag
for
include
delete
events,
so
you
can
opt
into
those
delete
events
to
get
the
old
behavior,
but
I
consider
this
a
implementation
miss
to
print
those
out
in
the
first
place.
So
what
are
your
thoughts
on
that
before?
I
fix
up
all
the
test
suite.
A
I
would
I
would
my
great
desire
is
to
have
the
cube
cuddle
events
command
pushed
over.
A
If
I
remember
correctly,
there
is
someone
who
did
created
a
brian
created,
a
pr
that
I
still
owe
review,
because
we
merged
the
pr
the
cap
couple
releases
back
and
unfortunately,
the
original
author
did
not
get
a
chance
to
complete
the
move.
A
A
A
If
I
initially
thought
that
maybe
we
should
just
add
additional
filtering
into
get,
but
the
problem
is
we
are,
we
will
be
introducing
flags
that
are
tied
to
only
one
resource
and
cubecadoget
is
meant
to
run
with
every
possible
resource.
A
B
No,
I
think,
you're
right.
That
makes
sense.
We
shouldn't
be
introducing
flags
for
specific
resources.
Like
that
all
right,
I
will
I'll
close
that
pr
and
I'll
put
effort
into
the
open
pr
and
see
how
we
can
get
that
across
the
finish
line.
A
Exactly
if
you
have
the
resources,
please
get
a
review
which
reminds
me
and
I'll
do
I'll
do
that
in
a
bit.
It
would
be
nice.
It's
definitely
not
something
that
will
be
able
to
finish
for
122.,
but
it
will
be
nice
to
get
it
completed
and
over
the
finish
line
for
123..
A
If
we
have
people
looking
at
the
pr
giving
a
try,
even
if
that
will
be,
and
as
an
introduced
as
an
alpha
commandment,
which
is
the
normal
thing,
we
could
get
it
out
there
and
gather
feedback
and
then
work
out
from
there,
which
reminds
me.
I
wanted
to
do
a
big
shout
out
to
three
people,
brian
mark.
If
I
remember
correctly
and
bill
was
the
third
person
I
need
to
double
check
for
the
batch
completions,
they
did
an
amazing
work.
Yeah
eddie,
just
pasted
out
the
the
pr.
A
Yes,
that
was
mark
bill
and
and
brian
they
put
a
lot
of
effort
mark
for
creating
and
bill
and
brian
for
actually
looking
at
the
pr
providing
very
valuable
review.
I
really
liked-
and
I
only
got
there
at
the
end
result,
which
is
very
clean,
very
nice,
very
contained.
A
It
was
like
a
pleasure
to
actually
review
this
vr
at
the
current
state
because
it
was
simple
and
and
very
very
readable,
so
I'm
I'm
very
helpful,
very
happy
that
we
got
it
in
and
I'm
very
thankful
to
all
three
of
those
gentlemen
for
actually
pushing
this
over
and
it's
especially
waiting
for
this
amount
of
time,
because
that's
more
than
six
months
from
what
I'm
looking
at
the
initial
opening
of
the
pr
to
actually
get
it
over
the
hump.
So
thank
you
very
much
and
big
shout
out
to
the
three
of
you.
A
Okay,
do
we
have
any
other
topics
or
stuff
that
people
that
folks
want
to
talk
about?
I
don't
know
any
particular
pr's
that
you
want
to
get
looked
at
or
before
the
code
freeze.
F
As
a
follow-up
on
the
topic
we
discussed
about
the
yaml
spacing
issues
in
customized
natasha,
and
I
and
mike
have
been
working
on
that
problem
and
we
have
some
work
in
progress.
Open,
like
natasha,
has
a
peer
that
fixes
all
the
problems,
and
we
have
a
essentially
vendoring
it
into
into
customized
solution
that
we
are
intending
to
move
forward
with
today,
because
the
maintainer
agreed
to
make
a
fix,
but
did
not
come
through
with
it
in
time.
So
we're
going
to
use
natasha's
is
the
plan.
G
Is
there
a
better
yaml
library
out
there,
students
not
really
it's
not
really
on
the
ball,
I
know.
G
C
F
Yeah,
it
seems
pretty
unfortunate
that
it's
it's
just
the
one
person
maintaining
it
and
it's
such
a
busy,
prominent
individual
who,
who
is
doing
that?
I
think
phil,
would
be
able
to
speak
to
this
more
specifically
than
than
me.
But
my
guess
is
that
the
reason
we're
using
this
one
is
it's
an
implementation
of
ymo
1.2
and
most
of
the
other
ones
are
implementations
of
1.1.
F
So
kml
specifically
wants
1.2
semantics,
and
things
like
comment
preservation
that
most
of
the
other,
don't
don't
support,
so
yeah
phil
could
confirm,
but
that's
my
understanding
of
why
we
are
using
this
one
in
particular,
in
kml,
specifically.
B
Yeah
gustavo
is
he's
the
cto
of
canonical
and
he's
like
super
busy
and
really
smart.
When
I
was
trying
to
upgrade
cobra,
so
we
could
do
the
completions
and
all
that
other
stuff,
I
ran
into
a
ton
of
issues
with
the
yama
library
and
gustavo,
did
put
a
bunch
of
work
into
like
reverting
some
pr's
and
making
things.
But
we
still
have
this
issue
of
just
like
the
spacing
is
going
to
break
everything
into
kubernetes
project.
G
Yeah,
I'm
potentially
just
connecting
with
him
to
add
more
kind
of
contributors
to
that.
That
might
be
a
pass
because
we're
sort
of
in
an
unfortunate
bind
here
where
he's
got
some
strong
opinions
about
how
it
should
evolve
and
little
time
to
approve
or
help
others
contribute.
So
we're
sort
of
in
this
bottleneck
scenario
so.
F
Natasha
and
mike
both
had
prs
up
at
various
points
in
the
past
two
weeks
asking
for
feedback
from
the
maintainers.
I'm
like
you
know
this.
This
seems
to
work.
Can
you
please?
Let
me
know
if
this
is.
This
is
what
you
had
in
mind.
You
would
really
like
to
get
this
merged
for
the
deadline
and
there
was
no
response
on
on
either
pr.
So
we
don't
have
direct
feedback
on
our
implementations.
F
They
pass
the
tests,
they
pass
our
tests
when
we
vendor
them
in,
but
that's
that's
the
situation.
A
Okay,
I
I
noticed
that
katrina
nicely
transitioned
to
stand-ups.
Do
we
have
any
other
sub
projects
that
want
to
present
their
stand-up
or
folks
that
they
want
to
report
what
they've
been
doing
over
the
past
couple
last
weeks.
D
I
want
to
ask
everyone
to
give
some
support
to
eddie's
proposal
about
not
shooting
the
feet,
because
I
don't
want
to
delete
all
my
pvs
all
my
physical
volumes
again
instead
of
all
of
my
physical
volume
claims-
and
this
actually
happened,
unfortunately,
is
what
it
was
on
a
station
cluster,
but
yeah
there
you
go
eddie.
I
support
you
on
that.
A
A
Probably
all
of
us
have
been
at
one
or
more
heavy
stage
of
removing
something.
B
But
I
am
so
so
status
update
on
that.
I'm
planning
to
talk
to
tim
hawkins,
hopefully
this
week
and
him-
and
I
were
tossing
around
some
ideas
that
we
might
throw
out
for
like
the
admin
api
or
something
so
we'll
see
what
hap
I
feel
like
it's
just
stalled
out
now
and
it's
not
going
to
make
the
release
in
any
way.
So
there's
no
reason
to
rush
it.
So
we'll
just
see
where
it
goes.
C
I
saw
that
brendan
burns
had
just
recently
also
given
what
sounded
like
productive
advice
or
feedback.
What
did
you
think.
B
Of
of
that
eddie,
I,
like
I,
like
the
idea.
I
still
think
that
that's
all
orthogonal
solutions,
like
the
locking
thing
is
great.
I
think
we
should
absolutely
implement
it,
but
going
back
to
quote
katrina.
The
issue
is
not
that
there
are
things
that
should
never
be
deleted.
The
issue
is
that
people
have
the
ability
to
make
a
mistake
and
they're
making
it
so
opting
into
like
a
locking
mechanism
doesn't
really
solve
that
issue,
but
I
like
the
idea,
I
still
think
we
should
do
it.
A
A
A
Thank
you
very
much
all
see
you
in
two
weeks
for
another
six
cli
call.
Thank
you
very
much
bye
all
right.
Thank
you
guys.
Bye-Bye.