►
From YouTube: Kubernetes SIG CLI 20210113
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
january,
13th,
so
lucky
13
for
us
to
start
our
six
eli
calls
we
have
agenda
packed.
So
let
me
start
with
the
announcements.
A
121
release
dates
are
up,
as
the
announcement
said.
These
are
not
the
final
final
dates,
but
I'm
expecting
that
they
will
be
close
to
this
so
february.
9Th
is
the
enhancement
freeze,
which
basically
means
we
have
less
than
a
month
to
merge
any
kubernetes
enhancements
proposals
that
we
want
to
land
for
121
and
the
code
freeze
for
121
will
be
on
march
9th.
So
that's
about
two
months
less
than
two
months
from
today
there
is
an
ask
from
the
production
readiness
team.
All
the
caps
need
to
be
migrated
to
the
new
template.
A
So
if
you
authored
a
cap
in
the
past,
if
you
can
update
your
cap
and
then
reach
out
to
me
phil,
eddie
or
sean
for
approvals,
we
we
can
help
you
get
those
merch
and
we
will
help
the
production
readiness
review
team
with
with
their
tooling
eddie
phil
sean.
Is
there
anything
else
you
wanted
to
add.
B
A
This
is
for
all
the
caps,
even
the
ones
that
you
you
did
merge
in
the
past
to
the
ones.
So
I
can't
remember
when
exactly
the
new
template
for
the
cap
happened.
A
I
I'd
like
to
say
that
was
somewhere
somewhere
in
december.
So
if
you
were
authoring
a
cap
in
the
past
a
month
up
to
two,
I
would
say
that
there's
good
chance
that
you
already
have
the
new
template.
Although
I
remember
that
john
mentioned
that
there
was
so
much
slight
changes
over
the
christmas,
the
holiday
shutdown.
A
D
A
But
I'll
I'll
figure
out
with
with
voidtech
and
I'll
I'll
double
confirm.
A
A
So,
okay,
I've
noticed
a
couple
new
names
in
the
participants
list,
so
I'll
I'll
give
folks
chance
to
speak
up
if
you're
new
to
this
call
feel
free
to
introduce
yourself
say
who
you
are:
what
you're
looking
for
in
the
6dli
meetings
and
what
you're
interested
in.
E
Hi,
my
name
is
holly,
I'm
I'm
working
for
red
hand
right
now,
john
effort.
I
think
I'm
on
the
right
hand
and
I've
been
on
last
week's
bus
call
and
to
this
call
I
knew
so.
I
have
a
cap
and
the
pr
submitted
the
list
regarding
a
minor
change
ignore
coupe
drain.
Could
control
drain
ignore
demon
sets?
So
that's
the
one.
Thank
you.
F
Hi,
I'm
anirud
and
basically
my
first
time
on
the
call,
and
I
my
passion,
is
working
on
kubernetes
and
getting
things
deployed
so
getting
cli
up
to
the
mark.
I'm
working
for
a
company
that
actually
uses
a
cli
similar
to
how
we
use
similar
to
our
kubernetes
managed
and
I'm
looking
forward
to
learn
more
about
how
things
work
here
and
may
perhaps
contribute.
G
My
name
is
is
francisco
I
work
for
for
an
insurance
company
in
germany
and
we're
using
another
another
tool
other
than
kubernetes,
but
we're
migrating
to
kubernetes
and
I'm
very
interested
in
starting
contributing
to
kubernetes,
perhaps
starting
with
with
the
common
line
interface,
because
I
think
I
have
some
good
ideas,
maybe
after
learning
how
things
work
a
little
bit
better,
I
can
contribute
a
bit
awesome.
Welcome
francisco.
H
I
I
I
did
have
one
question:
yeah
go
ahead,
so
I'm
pretty,
I
understand
how
things
graduate
in
in
the
kept
process
for
components
that
are
not
cube.
Ctl
does
cube.
Ctl
have
future
flag
feature
flags
that
move
it
through
alpha
beta
table.
A
A
It's
usu,
it
depends
if
you
want
to
deprecate
some
flags,
it's
a
matter
of
putting
a
warning
and
we
follow
the
usual
kubernetes
api
guidelines
for
how
long
this,
this
kind
of
deprecation
has
to
be
out
there,
and
only
after
that
we
will
be
removing
flags.
A
If
you
are
changing
flags,
there
is
a
actually
one
topic
that
will
be
going
through,
because
we
want
to
change
a
default
flag,
that's
something
that
we
hasn't
done.
Yet
we
did
change
a
a
flag
from
a
simple
boolean
to
a
more
complicated,
but
we,
but
what
we
did
in
the
past
is
if
you're
using
server
side,
for
example,
it
used
to
be
a
full
flag.
A
True
volts
only
currently,
there
are
three
possible
values,
so
we
parse
both
the
new
flag
and
the
old
one
for
a
transition
period,
and
we
put
a
warning
that
this
is
the
change
and
you
should
be
using
this
instead
of
the
true
false
value
for
this
flag.
So
that's
how
we
usually
do
all
of
the
changes.
It's
the
gradual.
The
graduation
process
is
pretty
much
similar
to
any
other
kubernetes
api.
It
is
just
through
warnings
that
we
inform
users
whenever
it
starts
and
the
timelines
are
identical.
A
That's
something
that
we'll
be
actually
discussing
a
little
bit
later
today.
There
is
a
there's,
a
topic
of
changing
the
default
flag.
J
Other
yeah
go
ahead.
My
name
is
kingdom.
I
am
application
developer
at
university
of
notre
dame
this
week,
but
next
week
I'm
joining
weave
works
and
so
I'll
be
having
a
lot
more
time
for
things
like
sigs,
and
this
is
my
first
sig
meeting,
I
was
invited
by
eddie.
We,
we
talked
a
little
while
ago
he's
a
friend
of
mine
from
a
long
time,
so.
K
K
A
Okay
looks
like
not,
so
we
can
jump
over
to
the
main
topics.
The
first
one
is
from
natasha
about
okay,
open
api
features
and
customize
take
away.
Take
it
away
from
here
in
natasha.
L
Yeah,
so
as
a
recap,
at
the
last
meeting
I
had
some
slides
that
described
a
proposal
for
customize,
adding
an
open
api
field
to
the
customization
file
to
allow
users
to
specify
their
own
open
api
schema,
and
this
would
support
strategic
merging
for
custom
resources.
L
D
Sounds
great
you're
just
discussing
before
this
meeting
that
we're
gonna
need
to
have
like
a
better
system
for
like
queuing
up
this
sort
of
work,
so
that
you
don't
have
to
wait
until
this
meeting
to
get
like
a
review
and
that
this
meeting
would
be
more
of
an
escalation.
If
there
was
like
contentious
things
that
weren't
getting
resolved.
L
D
The
pr
itself,
so
thanks
for
bringing
it
here,
who's
the
like,
I
could
give
this
a
shot.
Is
there
you
get
on
this
review.
L
So
I
tagged
you
and
jeff,
I'm
not
sure.
If
there's
someone
else,
I
should
tag,
but
if
you
could
take
a
look,
that
would
be
great.
C
C
Worth
I'm
already
familiar
with
it
and
I'm
I'm
fine,
the
cap
is
actually
pretty
short,
it's
just
the
boilerplate
that
makes
it
look
bigger
than
it
actually
is,
but
it's
all
about
reproducible
builds.
You
know
putting
a
an
open
api
spec
right
in
the
customization
file,
as
opposed
to
like
using
an
environment
variable
or
something
so
that
you
can
load
a
new
and
different
version
of
a
spec
or
load.
D
Awesome
I'll
make
sure
to
give
this
a
review
by
end
of
day.
L
L
L
A
Yeah,
I
muted
myself,
sorry
ricardo
you're
up
next,
I
didn't
interrupt
natasha.
N
Okay,
so
I'm
gonna,
be
I'm
gonna,
be
really
quick.
I'm
I'm
I'm
doing
some
low
hanging
fruits
reviews
into
pr's
trying
to
to
reduce
our
cue
for
the
like
the
extra
small
and
the
small
pr
that
I
am
able
to
to
review,
and
I've
been
seeing
in
the
past.
There
is
a
lot
of
like
cleanups
prs,
actually,
most
of
them,
with
the
internalization
on
cuba,
city
or
long
usage
and
so
on.
N
So
I
was
reviewing
this
one
and
then
about
if
there
is
a
place,
because
there
was
a
there
was
an
issue
tracking
all
of
those
in
into
cuba,
city
or
repo.
I
think
eddie
was
taking
taking
control
of
that
one,
but
then
that
that
issue
is
closed.
So
if
there
is
something
like
tracking
the
documentation
documenting
this
and
tracking
this
and
seeing
if
something
breaks
and
honestly,
I
don't
know
if,
if
there
is
like,
what's
the
what
what's
the
process
that
that's
being
followed,
follow
it
here.
N
So
so
I
I
propose
to
bring
this
and
because
I
ask
if
someone
is
going
to
to
like
to
play
this
effort,
but
he
he
suggested
us
to
do
something
like
include
test
pro
process,
proper
process
for
for
that
and
and
documentation
of
those
two,
those
things
and
the
best
pr.
They
were
something
like.
Okay,
that's
a
good
first
issue
for
the
for
someone
who
wants
to
take
care
of
that.
So
I
don't
know
how's
the
process.
This
one.
M
Yeah,
thank
you
for
bringing
this
up
so
that
issue
came
up
from
a
while
ago,
when
brian
and
I
were
in
some
code
base,
and
we
realized
that
a
lot
of
stuff
wasn't
using
the
i18n
wrappers.
So
we
kind
of
just
opened
that
to
track
it,
and
then
I
completely
forgot
and
didn't
follow
up
on.
It
is
this
something
that
you
want
to
take
ownership
over
cargo,
because
it
does
need
an
owner
to
get
over
the
finish
line
for
sure.
N
I
can
step
forward.
Yes,
I
am
like
with
low
bandwidth
with
that,
but
I
can
step
forward
for
sure
eddie.
So
I
can.
I
can
ping
you
on
on
slack
and
ask
what
what
do
you
expect
if
that's
only
like
replacing
the
descriptions
for
the
internalization
function
or,
if
there's
something
more,
that
I
need
to
take
care
when
users
are
like
when,
when
the
the
the
contributors
are
like
changing
something
like
need
to
take
care
of
documentations
or
tests
or
just
replacing
the
internalization
streams
that
it's
enough,
but
I
can
take
care
of
that.
Yeah.
A
I
a
good
starting
point
is
something
what
we
did
in
the
past
is
create
a
some
kind
of
an
uber
issue
where
all
the
prs
related
to
internet
internationalization
are
linked
to.
That
is
more
or
less
tracking
the
progress
of
where
we
are.
A
And
then,
if
anyone
asks
then,
for
example,
ricardo,
if
you're
interested
you
could
go
through
them,
provide
the
the
initial
reviews
and
then
reach
out
to
eddie
myself
or
sean,
and
we
can
do
a
bulk
approval
for
for
the
prs
that
you
went
through
and
you're.
Okay
with.
I
think.
That's
that's!
Basically
what
what
sean
was
also
mentioned.
Sorry
phil
was
mentioning
earlier
with
regards
to
owning
some
the
functional
piece
of
functionality
that
you're
trying
to
push
forward.
A
If
you're
interested
in
in
driving
that
effort,
I
think
it
will
be
pretty
cool.
N
M
M
A
A
A
E
Yeah,
so
you
can
pretty
well.
I
think
the
eddy
also
did
a
run
on
this
and
in
the
by
default,
the
coup
drain
ignore
ignore
demons.
That
is
very
every
in
order
to
move
forward.
Most
most
people
establish,
for
example,
in
in
red
hat
people
are
encountered.
Who
who
picked
me
this
side?
This
thought
is
nobody
ever
use
coup
drain
in
production
without
setting
ignore
demons?
E
You
want
to
do
something
meaningful.
If
you
know
you,
you
know
what
you're
doing
and
you
want
to
do
meaningful.
You
will
definitely
need
to
include
that,
so
why
don't
we
change
to
true
and
save
people,
this
hassle
to
type
those
16
or
20
characters?
So
that's
the
sum.
So
the
change
is
really
simple
and
the
the
the
documentation
is
also
easy
to
change.
E
A
So
I
agree
overall
with
the
change,
because
I
I
don't
remember
invoking
a
single
drain
without
specifying
the
ignore
daemon
sets
to
true,
and
I
see
that
eddie
and
others
are
saying
the
exact
same
thing.
The
problem
with
that
is,
we
would
be
changing
a
default
and
that's
problematic.
A
We
need
to
figure
out
a
way,
because,
with
all
the
changes
that
we
did
in
the
past,
it
was
usually
we
were
either
deprecating
or
adding
flags
which,
which
is
not
a
problem,
or
we
were
entirely
changing
how
the
flag
is
being
passed,
but
we
were
still
able
to
maintain
the
the
the
previous
behavior
here
here.
A
We
are
actually
changing
the
behavior
180
degrees,
so
the
the
thing
is
one
possible
approach
would
be
to
put
a
warning
that
we
are
going
to
change
the
default
for
this
flag
in
the
upcoming
release
and
have
that
warning
for
the
usual
period
of
time,
which
is
a
year
or
four
releases.
Whichever
is
longer.
A
And
and
only
then,
we
would
be
able
to
to
do
the
change.
A
I'm
not
sure
eventually,
we
could
probably
send
out
emails
to
cubedev
asking
for
people
to
for
feedback.
M
About
this
well,
this
this
ties
back
to
what
we
talked
about
last
week
with
brian's
idea
about
a
version.
Skew
warning.
I
think
we
we
talked
about
so
jordan
actually
implemented
this
at
some
point
right,
there's
actually
a
version
skew
for
ap
sure
that
came
out
no,
not
version
skew,
but
a
warning
feature:
yeah
yeah.
A
Yeah,
but
the
api
warning
is
for
deprecated
api.
So
that's,
I
think,
that's
a
separate.
The
version
school
warning
is
something
that
we
we
talked
about
several
times
and
I
I'm
pretty
sure
that
all
of
us
are
in
agreement
that,
yes,
we
want
to
have
it
done.
A
I
might
be
guilty
of
not
reviewing
the
pr
in
time,
but
I'll.
That's
definitely
something
that
we
want
to
have.
This,
though,
is
is
actually
the
warning
itself
is
not
any
different
from
a
deprecation
warning
that
we
have
a
bunch
across
any
of
the
cube
cutout
commands.
A
If
you,
for
example,
look
at
the
cube
cuddle
run
command
when
we
remove
the
the
majority
of
the
functionality,
there
are
still
a
handful
of
flags
that
are
currently
deprecated
and
if
you
try
to
pass
them,
you'll
just
get
a
warning
that
the
flag
is
not
anywhere
applicable.
One
such
example
is
dash
dash
schedule.
A
It
just
doesn't
make
sense
in
the
current.
Keep
caller
run
command.
The
problem
is
this
is
not
deprecating
a
flag.
This
is
changing
its
default
behavior
and
the
question
is
whether
having
such
warning
for
the
next
year
will
be
sufficient
for
people
to
digest
that.
Yes,
this
is
a
change
that
we
are
planning.
G
Shouldn't
we
maybe
deprecate
this
flag
and
then,
if
you
really
need
it
create
a
brand
new
flag,
that
is
the
opposite.
I
think
it
plays
better
with
the
process.
No.
J
I
have
a
question
also
what
so,
I'm
trying
to
think
through
the
possible
negative
effects
of
letting
this
process,
and
I
understand
the
process.
J
E
E
Actually
don't
got
hard
only
the
first
user
who
has
never
used
this
string.
You
actually
want
to
get
an
arrow
without
doing
anything,
but
if
you
already
read
the
menu
you
know
you
are
going
to
affect
it,
you
won't,
you
won't
get
hurt,
so
only
only
harmful
is
somebody
screaming
the
default
without
specifying
ignore
a
demon
set
which
is
going
to
fail
out,
but
may
not
right.
So
that's
a
that's.
The
only
only
truth
to
negative
effects.
A
Yeah,
the
the
script
is
something
that
I'm
most
worried
about,
because
that's
that's
usually
the
place
that
nobody
will
look
into
and
suddenly
when
they
upgrade
in
between,
they
will
notice
that
the
flag
is
not
there
or
there's
a
different
behavior.
A
It's
the
the
scripts
that
are
somewhere
in
the
in
the
data
centers
running
for
you
and
you
as
long
as
they're
working.
You
will
not
see
a
difference.
N
I
I
don't
know
if
this
sometime
came
to
dc.
I
don't
know
if
also,
this
is
something
that
we
want,
but
couldn't
we
start
for
the
future
like
when
those
things
happens,
to
propose
something
like
a
config
file
like
as
a
user
preference
like.
I
want
every
time
that
I
do
some
cube
ctl.
I
want
to
ignore
the
one
set,
but
this
only
happens
in
my
machine,
so
this
is
like
a
config
of
mine
even
keep
config
or
like
qrs
qqqrc,
or
something
like
that.
A
There
is
a
topic
from
doc
about
having
a
global
cubeconfig
and
we
will
be
touching
on
that
one.
A
little
bit
later.
O
There
are
lots
of
places
where
coop
cuddle
is
awfully
verbose
and
I
think
most
people
have
bash
aliases
or
csh
aliases
for
their
common
cases.
So
I
would
personally
err
on
the
side
of
just
documenting
and
suggesting,
if
you're,
really
doing
drains
with
ignored
damage
that
make
an
alias
seems
like
a
safer
and
more
just
more
in
line
with
the
way
kukado
is
generally
used
in
practice.
M
D
And
I
think
there's
a
couple
areas
where
we
had
to
do
similar
stuff
and
I
think
we've
like
it's
been,
have
a
warning
like
we
keep
the
behavior
to
standard
error
and
then
I
think
it
might
be
like
fail
to
command
or
or
something
like
that.
If
the
there's
different
ways
we've
done,
that,
like
one
is
like,
if
we're
removing
functionality,
it's
like
fail.
The
command,
but
still
like
do
the
right
thing
or
something
so
scripts
start.
F
D
Command
is
not
a
preferred
usage,
then
we
can
always
log
a
message
to
standard
error
and
tell
people
to
change
how
they're,
using
it.
A
Okay,
so
we
still
have
a
little
bit
of
time,
I'm
leaning
towards
what
eddie
proposed,
as
in
let's
bring
it
up
for
sick
architecture
and
cube
dev.
So
how
can
you
write
an
email
to
cubedev
and
sick
architecture
describing
the
situation
and
asking
for
feedback
that
we
are
considering
changing
the
default
or
if
there
will
be
no
objections
to
entire
to
deprecate
that
one
instead
and
entirely
remove?
A
If
we
don't
hear
any
objections,
we
can
start
the
deprecation
process.
If
there
will
be
objections
over
the
next
year,
we
can
always
change
from
deprecating
to
actually
changing
the
the
default
value
at
the
time,
and
if
there
will
be
no
objections
within
a
year,
the
flag
is
gone
and
we
don't
have
a
problem.
E
A
I
mean
write
an
email
to
sick
architecture,
mailing
list
and
kubernetes
dev.
Okay.
This
has
sufficient
exposure
to
the
stick
architecture
and
most
of
the
devs
that
are
working
with
kubernetes
on
a
daily
basis.
Okay,.
A
Basically,
either
we'll
be
fine,
we'll
go
that
way.
Yeah.
A
Thank
you
for
bringing
in
brian
the
school
warning.
P
Yeah,
this
is
something
that
that
came
up
in
the
in
the
bug
scrub
last
week
and
a
long
time
ago,
I
started
working
on
this
pr.
That
would
always
display
a
skew
warning
if
it
was
out
of
skew,
it
was
pretty
intrusive,
the
goal
being
to
let
people
know
that
they're
using
an
unsupported
skew
of
the
client
and
server
versions.
P
We
I'm
not,
I'm
not
sure
you
know
we
talked
about
sending
something
out
to
the
community
and
getting
feedback
about
the
warning,
but
you
know
I
guess
I'm
not
sure
about
that.
But
what
came
up
on
last
week's
call
was
you
know?
Maybe
what
we
need
is
just
when,
when
you
do
a
coupe
cuddle
version
to
show
the
warning
that
might
maybe
a
good
first
step
is:
let's
show
the
warning
there
when
you,
when
you
see
the
client
server
version,
we
can
have
a
warning
that
says
hey
by
the
way.
P
Your
your
client
and
server
version
is
out
of
sku,
and
that
will
help
when
we're
getting
issues
reported.
When
people
do
cook
cuddle
version
to
report
the
issue
they'll
know
that
they
have
a
skew
version
problem
without
us.
It
seems
like
over
and
over
we're
having
to
tell
people
hey
your
your
version
is
out
of
the
supported
sku,
so
I
guess
two
things
I
wanted
to
get
feedback
on
the
group
of
whether
that's
a
good
first
step
to
simply
show
the
warning
when
we
do
a
kubeco
version,
and,
furthermore,
do
we
want
to
try.
P
Q
N
Guess
that
maybe
being
annoying
and
printing
a
lot
of
those
messages
will
make
the
users
updated
their
qctl
comment,
instead
of
relying
only
into
the
version.
I
I
see
a
lot
of
issues
in
in
sig
network
and
other
ones
and
like
the
version
skill
is
huge
and
probably
it's
someone
that
just
received
it
closer
to
to
to
make
administration
and
even
don't
know
that
there
is
a
qctl
version
common
so
and
maybe
having
a
user
preference
that
allow
again
the
user
to
ignore,
like
warnings.
A
Right
having
it,
it
would
have
to
be
a
painful.
If
I
remember
correctly.
Currently,
you
can
pass
an
option
to
require
a
hard
version
match
so
that
if
you
have
a
sku,
it
won't
work
at
all.
We
won't
be
going
as
far
but
we'll
be
going
up
to
the
point
where
you
are
reminded,
but
we
don't
break
you.
A
Yeah,
but
the
question
is
whether
the
warning
was
only
in
which
the
direction
is
the
warning.
If
we
do
the
warning
so
that
cube
cuddle
can
be
newer
than
the
server.
Also,
don't
forget
that
at
some
point
in
time,
when
cubica
will
be
in
a
separate
repo,
the
school
will
be
always
we
will.
I
would
I'm
inclined
to
say
yeah.
This
is
the
minimal
version
and
we
don't
have
any
particular
for
mocks.
D
We
want
the
warning,
we're
gonna,
do
the
warning,
it's
valuable
to
it
in
both
directions,
because
there
are
times
where,
like
control
stops
using
an
api,
that's
been
deprecated
and
moves
on
to
a
different
api
right
and
doesn't
check
that
the
old
one
exists.
You
could
imagine
server
side
apply
at
some
point.
The
intent
would
be
to
just
use
that
api
and
remove
the
code.
That
does
the
client
side
apply.
D
The
the
warning
message
seems
like
the
in
the
version
commands.
It
seems
like
a
no-brainer
like
I
don't
see
any
reason
not
to
do
that
in
the
other
commands
it
seems
like
you
know,
for
writing
it
to
standard
error.
It
seems
inappropriate
information
to
give
the
users
exiting
non-zero
is
probably
like
something
that
needs
to
be
discussed
more
like
there's
trade-offs
there.
D
I
I
totally
get
the
arguments
for
it,
but
like
it
also
breaks
something
that
doesn't
break
today
and
will
break
some
users,
so
we
should
that
one
requires
more
discussion.
A
Okay,
so
let's
let's
proceed
with
just
the
version
command
and
we
can
probably
ask
for
something
more
in
the
long
run.
P
That
sounds
good
I'll
I'll
I'll
update
the
pr
and
or
submit
new
pr
with
just
the
version
command,
and
then
we
can
either
create
a
new
issue
or
update
the
issue
to
to
track
the
overall
warning.
A
Sounds
great
all
right
thanks,
brad
doug,
your
topic.
Q
Hello,
everyone,
so
the
first
one
is
actually.
The
two
topics
I
have
here
is
about
the
review.
So
it's
asking
people
to
take
a
few
minutes
of
the
day
and
take
a
look
in
the
two
pr's
they
contain
enhancements.
Q
So
the
first
one
is
about
to
have
a
generic
notation
for
the
crew
control
exact
log
and
attach
so
a
single
notation
for
all
the
commands.
So
you
can
select
the
pod.
Let's
say
you
have
a
a
sorry
container.
You
have
a
pod
of
side
containers,
so
you
want
to
select
one
when
you
do
control
logs
or
cube
control
attach
so
have
a
single
notation
for
all
commands.
Q
For
me,
it
looks
good
it
works,
but
you
know
I'd
like
to
have
more
people
reviewing
and
the
approvers
as
well.
A
I
owe
you
reviews
in
both.
I
remember
that
I
was
behind
at
least
the
default
container
annotation,
because
we
did
something
similar
on
the
openshift
side
or
at
least
in
the
logs
command,
and
I
remember
that
I
suggested
to
have
this
one
generic,
I'm
hoping
to
go
through
cap
reviews
sometime
next
week.
A
It's
it's
pretty
hard
for
me
to
do
any
more
work
this
week
than
I
already
have
planned,
but
I
think
next
week
is
is
feasible,
I'll
ping
you
on
slack
as
soon
as
I'll
be
done
with
those
and
and
we
can
I'm
I'm
as
long
for
the
the
default
container
I'll.
Take
the
ownership
on
that
one
and
I'm
not
sure
if,
if
anyone
is
anyone
else
is
interested,
but
we
can.
A
A
Q
Sure
I
think
I
split
from
the
the
next
capped
that
I'm
going
to
say.
I
think
we
need
to
create
a
cap
for
that,
in
my
opinion,
at
least
right,
so
the
next
one
is
migrate,
the
the
home
dot
cub
directory
to
the
one
that
is
under
dot
config,
which
most
of
the
applications
are
using
these
days,
and
we
have
we
had
some
people
from
red
hat
and
also
jordan.
Q
Reviewing
this
one
doing
some
comments
like
okay,
guys
we're
going
to
replace
the
the
new
directory,
but
what
happens
with
the
tools
that
are
using?
It
is
going
to
break.
Why
don't?
Why?
Don't
you
we
just
say
to
the
customers?
I'm
sorry
to
the
users
like
use,
could
config
environment
variable
you
know,
so
I
added
all
these
points
into
the
capped,
so
the
people
can
see
my
my
points
of
view
and
also
I
added
a
patch
to
to
show
this
movement.
So
if
you
guys
can
take
a
look,
it'll
be
nice.
A
The
cube
conflict
move
will
be
a
bigger
one
and
it
will
be
something
that
honestly,
my
guess
is
that
we
will
be
maintaining
backwards,
compatibility
for
for
an
extended
period
of
time,
a
lot
more
than
the
default
deprecation
time
frames
are
because
I
it's
not
only
us
but,
like
you
said
there
are
other.
There
are
other
tour
tools
that
rely
on
that
on
the
file.
Obviously
it
it
is
possible
to
have
a
link
in
place
that
will
solve
the
situation
and
probably
somewhere
along
the
line.
A
We
will
have
to
have
such
a
link
so
that
the
transition
is
smooth
to
a
lot
of
people
and
we
support
both
location
at
the
same
time,
but
yeah.
I'm
I'm
pretty
confident
that
this
needs
a
broader
exposure
to
the
entire
community.
A
So,
like
I
said,
I
would
definitely
encourage
you
to
send
an
email
to
cubedf
and
6eli
asking
for
feedback
broad
feedback.
I
already.
I
don't
want
to
rush
with
this
one,
because
that's
a
that's
a
big
change
yeah,
but
we
can
start
laying
the
grounds
for
this
gather
the
feedback
and
slowly
prepare
ourselves
for
how
to
go.
A
Okay,
we
are
about
seven
minutes
to
to
the
top
of
the
hour.
Let's
move
on
to
the
stand-ups,
I
see
koi
nick
you
want
to
talk
about
it.
O
Sure
yeah
keep
it
off
of
the
short
and
sweet
here,
we're
doing
some
stuff
to
ready
for
the
the
upcoming
pretty
soon.
Hopefully,
repo
transition
right
now
we're
still
under
the
ibm
github
org
we're
hoping
to
transfer
that
over
pretty
soon
we're
just
doing
checking
all
the
boxes.
At
this
point.
O
So
in
preparation
also,
we've
done
some
polishing
to
the
code
base,
we'll
be
updating
the
readme.
If
you
guys
have
any
suggestions
about
how
we
can
reorient
the
read
me
to
be
more
sort
of
broad
on
all
ears
open
up
issues,
one
of
the
things
we've
also
done
is
we've
done
some
cleanup
just
in
the
way
that
we're
talking
to
the
api
server
up
until
now,
we've
been
basically
invoking
the
cli
for
every
command,
so
you
type
kui
command.
It
just
invokes
the
coop
cuddle
cli,
which
adds
to
quite
a
bit
of
overhead.
O
So
right,
we've
updated
that
to
talk
to
the
api
server
directly,
at
least
for
some
common
cases.
So
previously
it
would
have
been
taking
me
forward,
forge
it
400,
500
milliseconds
for
some
pretty
common
operations
and
now
they're
down
to
maybe
80
milliseconds,
which
is
actually
faster
than
using
couple
cli
directly.
So
for
most
of
the
common
cases
of
viewing
tables,
I'm
drilling
down
to
the
details
of
a
resource
they're
about
two
to
three
times
faster
than
previously,
and
also
about
roughly
that
same
magnitude
faster
than
using
cuddle.
A
Highlights
totally
get
it:
okay,
cool
thanks,
a
lot
nick
that
that
our
good
news
coming
from
the
korea
camp
jeff
you're,
still
in
the
call
do
you
have
any
updates
on
the
preparation
customized
for
being
re-vendored
into
keep
cuddle.
I
remember
there's
a
person
on
slack
that
that
was
pinging
me
he's
interested
in
in
bumping
customize
in
the
main
kubernetes.
Well
I
mean
cuba
cutoff
repo.
C
But
as
we
make
the
final
changes,
we
are
triggering
failures
of
some
of
the
tests,
so
I'm
going
in
and
fixing
them
I'm
aware
of
the
february
9th
deadline
and
for
the
I
don't
know,
if
we're
going
to
need
a
enhancement
proposal
for
this,
because
arguably
it's
just
changing
the
libraries
that
keep
cuddles
linked
to,
but
I'm
also
where
the
code
freeze
is
coming
up
in
march.
C
So
I
think
those
are
doable
we're
working
really
hard
on
it,
brian,
if
you're
interested
in
helping
with
some
things.
Please
go
ahead
and
reach
out
to
me
because
there's
one
bug
in
particular
that
I
think
you'd
be
able
to
contribute
to.
I
can't
remember
the
name
of
it.
I
have
the
number
of
it
off
the
top
of
my
head,
but
it's
something
about
yeah.
C
Anyway,
you
can
reach
out
to
me
in
slack
yeah
we're
very
close
close
as
we've
ever
been,
but
it's
always
these
last
little
bugs
that
we're
trying
to
squash,
because
right
now,
customize
uses
both
api
machinery,
libraries
and
all
the
new
kb
kbm
libraries.
All
of
the
tests
have
to
pass
for
both
sides
of
that
and
maintaining
that
is
kind
of
tricky.
A
I
I'm
not
100
sure
that
we
need
a
cap
for
this
one,
there's
a
bunch
of
issues
that
basically
people
are
asking
and
I've.
I've
seen
numerous
comments
where
people.
A
C
A
I
see
ackman
on
the
call
ahmad.
Do
you
want
to
give
a
short
update
on
the
crew.
B
Hello,
everyone
so
one
second.
B
Yeah,
so
I'm
ahmad,
I'm
a
software
engineer
at
the
lee's
web
yeah,
I'm
just
you
know,
trying
to
join
nowadays.
You
know
kubernetes
and
yeah.
Somehow
I
yeah
I'm
trying
to
help
with
mostly
issues.
Maybe
in
the
future
I
will
be
able
to
create
a
vr.
A
Okay,
thanks
a
lot
ahmed
for
introducing
I
I
thank
you
for
someone
else.
Chris
corrected
me
and
he
said
that
there's
no
crew
update
on
this
week,
okay,
we're
one
minute
left.
I
think
that
that's
pretty
much
all
folks.
Thank
you
very
much
for
the
past
hour.
It
was
as
usual,
wonderful
time
talking
with
you
all
and
see
you
in
two
weeks,
bye
all
right.
Thank.