►
From YouTube: Kubernetes SIG CLI 20210811
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
Good
morning,
good
evening,
good
afternoon,
depending
on
where
you
are
today
is
august
11th-
and
this
is
another
of
our
bi-weekly
6
july
meeting-
my
name
is
maji
and
I'll
be
your
host.
Today,
our
agenda
is
full
packed
with
announcements
with
topics,
so
let's
get
let's
get
to
it
right
away,
but
before
we
do
introductions,
I
know
we
have
new
people
on
the
call.
A
Now
is
your
time
to
introduce
yourself
say
who
you
are,
what
you're
expecting
from
cli,
where
we
can
help
you
with
prs
issues,
any
kind
of
requests
the
floor
is
yours,.
B
D
E
F
My
name
is
ross
peoples.
I
I'm
in
united
states
nebraska
actually-
and
I
just
recently
joined
red
hat-
to
help
out
with
some
of
the
kubernetes
related
developments
and
right
now,
I'm
sort
of
focused
on
the
cli
tools
and
working
with
majie,
and
you
know
just
trying
to
try
to
get
started
and
get
going.
A
Okay,
hearing
none,
let's
move
on
to
the
announcement,
so
if
you
haven't
seen
it
123
kubernetes
release
schedule
is
out.
I
linked
it
in
the
agenda.
If
you
want
to
check
the
dates,
feel
free
to
do
so.
The
release
date
for
the
123
is
at
the
beginning
of
december.
A
More
details
are
in
this
document
from
the
link
additional
one.
The
kubecon
north
america
2021
in
la
the
schedule,
is
also
out
it
was
published
earlier
this
week
we
have
a
session
jeff,
eddie
and
katrina
are,
are
going
to
walk
us
through
where
we've
been
where
we
are
and
where
we're
heading
to
I
linked
their
session
in
in
that
agenda.
A
Kubecon
will
be
both
virtual
and
in
person
for
those
that
can
join
so
the
best
for
of
both
words
for
all
the
people
and
sean
added.
The
contributor
summit
ni
north
america,
which
is
part
of
kubecon,
is
also
happening.
A
Although
there's
one
kaveed
contributor
summit
is
limited
to
only
the
kubernetes
org
members,
so
if
you're
kubernetes
org
member,
you
can
join
the
event,
if
not,
and
unfortunately,.
A
You
won't
be
able
to
join
this
one
due
to
limitation.
H
No,
that
was
that
was
that
was
it
sounds
like
so
far
they're
going
to
do
the
it's
still
going.
H
Angeles,
obviously
there's
some
there's
been
some
doubts
because
of
the
horizon
in
copen,
but
as
of
now,
it's
still
there's
still
an
in-person
component.
G
Yeah,
as
far
as
I
understand
from
yesterday
in
the
leads
meeting,
it
will
be
happening
as
long
as
la
says
it
can.
So
if
la
the
city
of
la
decides,
nope
la
convention
center
is
closed,
then
it
won't
be
otherwise
it
will
so.
The
other
thing
I
wanted
to
ask
real
quick.
So
when
I
put
together
our
6doi
cfp
session,
I
one
of
the
things
I
wanted
to
call
out
in
there
is
how
we
test
cube
control.
G
I
don't
think
we've
done
a
talking
about
like
our
end-to-end
testing
and
like
any
of
that
framework
stuff.
We
do
so
if
anyone
has
any
other
ideas
for
what
would
be
beneficial
for
a
maintainer
track
session.
Please
let
us
know.
H
Go
ahead,
so
this
pr
was
assigned
to
me,
although
it
affects
everybody,
so
I
wanted
more
eyes
on
it.
So
that's
the
only
reason
I
brought
this
up
was
I'm
hoping
to
get
the
rest
of
you
know
six
eli's
expertise
to
ensure
that
this
is
necessary
and
correct.
H
H
About
caching,
so
he
he
adds
an
extra
set
of
mutexes,
the
author
for
the
discovery,
client
and
the
rest
mapper
or
the
the
client
config.
So
he
splits
that
out
instead
of
using
the
same
mutex
for
both
of
those
structures.
H
It's
actually
not
that
big
of
a
pr
and
yeah
all
of
the
tests
passed,
but
I
think
that
there
is
a
lot
of
subtlety
here,
and
so
that's
why
an
extra
set
of
eyes
extra
sets
of
eyes
would
be
would
be
great.
G
A
Think
he
mentioned-
I
just
read
it
somewhere
that
he
was
testing
this
and
he
hasn't
seen
too
much
of
a
change
so
that.
H
Correct
so
he
he
said
that
there's
125
api
calls
when
there's
the
inability
to
to
write
to
a
caching
directory
and
that
this
pr,
he
didn't
actually
see
a
reduction
in
that
he
said
that
would
require
another
pr
for
in-memory.
Caching,.
A
I'm
not
sure
about
the
in-memory
part,
because
we
are,
I
think
the
majority
are
using
the
on
this
one.
I
just
outlined
the
part
where
he
said
that
he
wasn't
able
to
reduce
the
number
of
api
calls,
but
the
general
directions
seems
reasonable.
A
Like
you
said
sean
I'll,
probably
the
devil
is
in
the
details.
So
probably
we'll
need
to
make
sure
that,
because
he
or
he
is
splitting
the
the
lock
separately
for
reading
the
config
and
separately
for
getting
the
discovery
which
theoretic
theoretically
as
I'm
thinking,
should
not
be
a
problem
in
majority
of
the
cases
that
I
can
think
of
that
shouldn't
be
a
problem,
but
yeah
be
careful.
J
H
It
you
know
it
would
be
which
are
difficult
to
test
he's,
claiming
a
deadlock
which
maybe
we
should
you
know
reproduce.
If
that's
the
case.
A
So
I
don't
I'll
definitely
check
it
out,
but
that
will
be
sometime
next
week,
I'm
out
for
the
rest
of
this
week.
So.
H
A
Okay
yeah,
I
yeah
I
like
his
statement
and
he's
very
energetic
about
the
improving
the
performance,
which
is
a
good
thing
with
the
product,
probably
variation
of
crds.
A
The
discovery
operations
are
becoming
more
and
more
expensive
at
some
point
in
time,
I
was
even
considering
the
ability
to
increase
the
time
between
how
long
a
cache
is
valid,
and
only
because
there
are
several
operations
where
we
explicitly
invalidate
the
cache
and
force
the
command
to
read
the
entire
discovery.
Information.
If
I
remember
correctly,
that's
api
versions-
and
I
think
even
api
resources
were
explicitly
invalidated.
A
Maybe
we
could
provide
a
a
command
for
the
invalidation
or
somehow
provide
a
configuration
option
to
tell
how
long
you
are
okay
without
reading
the
discover
information,
because
that
that
might
be
beneficial
for
for
some
use
cases
and
I'm
pretty
confident.
I
H
K
H
This
case,
where
he,
where
the
author
does
didn't,
have
a
writable
cash
directory.
I.
K
H
They
like,
as
I
mentioned
as
as
matching
mentioned,
it's
the
first
time
and
then,
if
you
don't,
have
a
writable
cache
directory,
then
it's
gonna
happen.
Every
time.
A
Yeah
well,
the
author
is
mentioning
the
the
non-writable
case,
especially
in
the
case
where
we
are
working
with
a
container.
So
you
have
a
cube
cutter
inside
of
a
container
and
you
not
necessarily
have
to
have
a
writable
directory
or
you
don't
have
a
home
set
properly
and
it's
picking
some
kind
of
a
temporary
which
might
be
slow
or
whatever.
A
A
Okay,
let's
move
on
to
the
next
one
added
by
eddie
eddie.
You
want
to
talk
about
this
one.
G
Sorry,
okay,
yeah
better.
There
are
a
few
pr's
that
we
need
to
take
a
look
at
in
a
little
bit
this
first,
one
that
you
have
open
here
is
adding.
So
if
you
have
a
path
configured
for
your
cube
config
for
your
api
server,
so
let's
say
that
you
have
a
proxy
set
up
that
like
slash,
kubernetes
will
go
to
your
api
server
that
will
not
work
when
you're
running
cube
control
proxy.
It
kind
of
ignores
that
path.
G
G
We
asked
for
like
more
context,
and
I
know
that
we
have
someone
from
the
operator
framework
now
who
also
ran
into
this.
So
you
want
to
provide
more
light
there.
B
B
Happened
with
the
opportunistic
as
well
when,
like
user,
is
using
our
key
cluster
and
they
found
like
path
and
baseball
tinder
and
went
through
the
same
issue.
For
time
being,
we
fix
the
issue
like
temporary,
where
we
are
throwing
an
error
for
base
part,
so
we
are
looking
for
the
spl
to
get
merged.
A
A
If
there
are
more
people
being
affected
by
it,
then
probably
this
only
it's
only
a
matter.
I
remember
looking
at
the
pr
overall,
it
looked
reasonable,
so
the
shape
of
this
pr
is
acceptable.
A
I
I
would
have
to
double
check
if
I
want
to
do
a
pen,
server
path
or
just
server
path,
but
that
that's
a
minor
minor
thing,
but
I
guess
if,
if
there
are
more
people-
and
it's
just
my
initial
concern-
was
that
this
was
like
solving
one
person
problem
only,
but
if
there
are
more
people
that
definitely
that's
that's
worthy
of
returning
and
and
I'll
add
that
one
to
my
list
for
next
week
as
well
in
that
case
and
I'll
make
sure
to
review
both
of
these
overall.
C
A
A
A
A
And
it
was
very
hard
for
me,
as
the
author
and
other
folks
that
were
maintaining
the
code
to
justify
this
and
and
similarly
for
users
to
figure
out
what
the
output
will
be
and-
and
we've
agreed
over
a
couple
of
discussions
that
it
would
be
better
if
we
try
to
simplify
the
commands
and
push
in
for
instead
of
using
run
push
people
towards
using
create
commands.
A
And-
and
I
still
keep
in
my
at
the
back
of
my
head-
whenever
making
a
decision
on
any
new
flag
that
if
this
is
solving
just
one
case,
I'm
not
sure
if
it's
worth
to
add
one
flag
for
one
person
for
one
single
case,
because
if
there
will
be
a
hundred
people
with
hundred
use
cases
we'll
end
up
easily
with
a
hundred
different
flags
for
simple
command.
A
If,
if
this
is
only
solving
their
one
particular
case,
if
there's
a
brother
people
affected
by
a
particular
problem,
I
don't
see
any
problem
with
that.
So
that's
that's!
That's
where
I'm
usually
trying
to
justify
whether
we
do
or
do
not
want
to
add
new
flags
to
the
commands,
especially
if
you
can
always
either
pipe
the
commands
together
and
add
them
or
directly
modify
through
cube
cuddle
edit
afterwards,
if
it's
possible,
because
there
are
some
cases
where
you
can't
edit
a
resource
after
it
was
submitted.
A
For
example,
once
you
submit
it,
there's
only
a
few
fields
that
you
can
modify
on
it.
So.
G
G
Yeah
yeah
there's
a
lot
of
scarring,
I'm
sure
matcha.
Thank
you
for
sharing.
I
saw
that
rancher
and
the
container
lens
folks
also
ran
into
this.
So
I'd
say
it's
it's
not
just
a
one-person
thing.
I
mean.
A
In
those
cases,
because
that's
that
basically
is
what
we've
discussed
a
couple
weeks
or
even
months
ago,
where
we've
all
basically
agreed
that
if
a
new
feature
will
have
a
sufficient
number
of
supporters
or
people
that
are
affected
by
it,
then
we
will
accept
it
and-
and
we
can
run
forward
with
adding
such
a
feature.
A
If
a
particular
case
is
is
having
minimal
impact,
then
we
either
wait
for
further
information
or
or
we
don't
accept
it,
and
I
was
waiting
for
additional
info
on
this
one.
G
So
this
is
adding
create
a
certificate
authority,
it
looks
fine
to
me
and
that
ricardo
looked
at
it.
I
had
one
question
on
the
thread:
there
is,
if
you
go
to
the
files
changed.
G
A
little
more
okay
that
one
so
this
that
hardcoded
ca.search
string,
that
was
just
a
question
about.
Should
this
be
a
api
type,
because
all
the
other
core
v1
tls
cert
key
above
it
they're
all
api
types.
G
A
That
seems
like
I
mean
it
mean
that
we
already
have
that
constant
for
the
ca.
That's
crt.
C
G
Go
ahead,
the
constant
is
there,
but
the
constant
variable
name
is
completely
different
than
the
use
case
here.
G
C
I
can
explain
this
one
yeah
so
as
as
an
ingress,
maintainer
yeah.
So
usually
you
have
to
create
you.
You
have
like
authentication,
client
authentication
with
into
ingress,
and
you
need
to
put
like
the
ca
as
an
example.
C
You
may
want
that
for
other
things
like
I
want
to
provide
the
ca
for
like
the
full
chain
for
the
ingress
controller
into
the
secret,
like
I
wanna,
have
the
private
key,
the
public
key
and
also
provide
the
full
chain.
So
it's
this
one.
It's
it's
reasonable.
A
I
would
have
to
it
looks
reasonable.
I
haven't
looked
at
the
actual
implementation,
actually,
the
the
flag
name.
I
would
be
okay
with
sca
if
you
go
through
any
of
the
open,
ssl
or
general
of
tooling
the
they
are
using
consistently.
Third
key
nca
pretty
much
everywhere.
I,
if
I
remember
correctly,
opens
cell,
has
a
ca
sub
command
around
that,
so
that
should
work.
I
would
have
to
just
check
if
the
rest
of
the
code
seems
somewhat
reasonable
and
but
overall
I
think
it
should
should
be
fine.
A
Yeah
I'll
I'll
add
that
one
to
my
list
it
should
be
pretty
quick
pr
to
review
yeah.
A
Okey-Doke,
I
guess
that's
all
the
topics.
Does
anyone
have
any
issues?
Topics
concerns
that
I
want
to
discuss
with
the
group.
C
I
I
forgot
to
put
in
the
agenda,
but
I
am
concerned
about
the
cap
of
the
exit,
call
that
I
own
I,
as
everybody
knows,
I
am
I'm
pretty
busy
with
my
my
new
day
job
and
I'm
not
able
to
to
follow
up
with
every
question
or
even
with
the
implementation.
So
I
guess
that
at
least
I
want
to
take
a
look
into
the
into
the
comments
that
might
say
message
did
with
that,
and
maybe
maybe
someone
when
these
gods
accepted
someone
follow
up
with
with
the
implementation.
A
Yeah,
I
think
I
can
sync
with
ross
and
we
can
somehow
help
you
push
this
one
through.
A
No,
that's!
That's!
That's
perfectly
fine
ricardo!
That's
good
that
you're
bringing
this
up
that
you're
not
able,
instead
of
you,
know
just
abandoning
you're
you're
admitting
that
you're,
you
just
don't
have
the
time.
That's
perfectly!
Okay!
A
Thank
you
very
much
for
that
one
and
I'll
sync
with
ross,
and
we
will
try
to
follow
up
on
helping
you
with
with
the
cap
itself
and
then
with
the
implementation.
It's
definitely
not
something
that
has
to
land
within
the
next
release.
That
password
with
the
next
release
is
something
that
I
that
I
did.
I
I
I
was-
and
I
still
have
on
my
list,
which
is
cube,
cuddle
events
that
I
was
hoping
that
the
sick
ux
will
help
us
with
and
after
cap
day
disappeared.
A
There
is
a
already
ready
implementation
and
I'm
hoping
that
it
will
be
something
that
will
be
able
to
work
in
123..
Okay,
we'll
definitely
I'll.
Add
a
note
here.
C
G
C
A
Yeah
definitely
or
instead
of
the
cap
itself.
Maybe
it's
sometimes
useful
to
to
to
point
to
the
issue
behind
the
cap,
because
cap
tend
to
be
closed
and
the
issue
will
be
open
for
a
little
bit
longer.
A
Okey-Doke
anyone
else
have
any
other
topics.
A
Okay,
hearing
none:
let's
move
on
to
the
stand-ups
katrina,
natasha.
L
I
don't
think
I
have
anything
particular
to
report
so
natasha.
If
you've
got
something
go
for
it
just
that
last
week
or
two
last
week
we
had
our
first
customized
bug
scrub
that
katrina
organized
which
went
very
smoothly.
I
was
super
productive
and
jeff
also
is
working
on
deporting
anchors,
which
is
a
very
popular
feature
request
and
other
than
that.
I
think
there's
just
a
few
caps
being
drafted
that
we
might
see
over
the
next
few
weeks.
L
Yeah,
I
guess
something
that
that
reminds
me-
we've
been
talking
about
among
customized
maintainers.
Is
that
some
of
the
time
we
want
like
a
bigger
explanation
of
a
feature
that's
being
proposed
for
customized.
It
doesn't
really
have
sig
wide
relevance.
It's
just
a
very
customized
specific
thing,
but
in
the
past
sometimes
we've
asked
for
a
cap
or
just
an
issue
we're
thinking
about
making
a
sort
of
a
kept
directory
inside
customized
for
things
like
that
that
are
fairly
small
in
scope
and
and
then
having
it.
L
So
that
kind
of
creates
this
middle
ground,
where
it's
easier
to
discuss
line
by
line
a
proposal
that
needs
more
detail
than
it
would
be
on
an
issue.
But
we
don't
have
to
subject
it
to
the
whole
formal
kept
process
for
something
that
that
isn't
really
big
enough
to
warrant
that
if
anyone
has
any
thoughts
on
that,
I'm
gonna
be
working
on
that
next
week.
So
let
me
know
eddie.
A
A
I
I
can't
remember
what
was
it
called,
I
think
when
we
talked
with
laurie,
she
mentioned
something
like
architecture,
design
review.
Something
like
that,
or
maybe
I
was
talking
with
someone
else.
It
was
like
lightweight
type
of
a
caps
thing
where
you
have
a
document
that
just
has
the
the
design
written
down
for
future
people
to
look
at
and
just
like.
Basically,
what
katrina
proposed,
maybe
there's
even
something
that
is
describing
that
I'll
try
to
find
it
katrina
and
I'll.
Send
you
the
link
to
what
I
I
think.
G
Yeah,
I
sent
katrina
all
that
when
we
talked
about
it
before
I
can
dig
it
up.
It's
probably
in
our
conversation,
if
you
scroll
back
katrina
the
the
idea
of
the
embedded
caps.
I
liked,
because
I
saw
other
people
did
it
as
well,
so
there
are
a
few
other
sub
projects
that
had
like
their
own
embedded
caps.
So.
L
And
one
thing
that
I
intend
to
do
in
there
as
part
of
the
template,
because
we
can,
we
can
evolve
it
on
separately
from
the
main
one
is
like
have
a
specific
section
about
implications
for
keep
cuddle
customized
to
make
sure
that
people
are
thinking
about
that.
And
if
there
is
something
big
there
then
suggest
the
template
will
suggest
that
it
actually
be
promoted
to
an
enhancements
kept,
because
anything
that
affects
cube
kettle
customize
is
going
to
be
a
broader
interest.
A
Eddie
found
it
so
yeah,
I'm
definitely
in
full
support
of
following
this
way,
something
that
will
be
lightweight
something
that
will
leave
a
track.
Allow
you
to
proceed
with
your
work
and
you
know,
doesn't
block
you,
because
the
caps
are
pretty
heavy
as
there
are
today,
but
they
are
serving
a
slightly
different
purpose.
A
K
Not
too
much
we're
a
bit
midstream
just
wanted
to
mention
briefly
that
these
studies
that
we
briefly
talked
about
last
time
are
still
underway.
We
couldn't
quite
get
enough
finish
to
present
anything
today,
but
hopefully
next
time,
cora
will
be
ready
to
present
that
if
we've
been
focusing
on
kubeco
versus
open
shift
console,
this
is
not
to
choose
any.
You
know.
Particular
you
know
villains
or
you
know,
or
victors
these
systems,
two
open
source
projects
that
were
nice
to
compare
to
each
other.
K
If
you
guys
have
any
other
ideas
of
participants,
you
know
in
this
battle.
Let
me
know,
but
this
is
our
initial
focus
and
we're
not
focusing
on
kui
per
se,
but
cool
will
be
kind
of
a
vehicle
for
exploring
how
we
can
enhance
both
sort
of
bringing
them
towards
a
common
medium,
but
the
main
focus
is
on
those
two.
So
hopefully,
next
turn
coral
will
be
able
to
show
some
of
the
findings.
A
Yeah
eddie
just
pointed
out
that
lens
would
be
a
good
reference
for
cree
as
well.
K
A
Sorry
nick
you're
done
with
with
koi
stuff
thanks,
yes,
okay,
so
there
is
one
other
topic
that
russ
pointed
me
to
ross
started
looking
into
building
cupcake
on
windows.
I
remember
that
during
workshops
I
think
that
was
in
san
diego,
which
sean
led.
A
We
had
a
one
person
that
that
appeared
with
the
windows
machine
and
at
the
time
we
didn't
have
any
good
docks
around
the
building,
keep
cutting
on
windows
and
if
time
will
be
good
for
us,
we're
hoping
to
address
that
missing
part,
because
ross
has
a
windows
machine
at
hand
that
he
will
hopefully
be
able
to
figure
out
the
missing
bits,
and
we
will
have
to
probably
document
I
started
looking
at
where
we
are
documenting
how
to
build
cupro
and
honestly,
I
had
some
issues
finding
them,
I'm
not
sure.
G
A
I
think
that's
all,
basically,
even
if
we
and
then
eventually,
we
will
have
to
figure
out
how
how
frequently
we
want
to
release
them.
How
do
we
want
to
support
cube
versions?
We
want
to
release
sooner,
I
mean
more
frequently
than
than
cube
is
in
those
cases.
It
will
have
to
support
eight
different
as
in
longer
school
window
and
so
forth,
but
I
would
leave
it
after
the
initial
bit
where
we
would
actually
switch
entirely
to
separate
repo,
and
we
would
need
an
agreement
from
sick
release
and
sick
testing.
A
If
we
jump
that
obstacle,
we
can
discuss
how
we
can
do
it
differently,
because
I'm
guessing
the
the
follow-up
will
be
simpler
than
figuring
out
how
to
build
and
test,
because
the
testing
bit
will
be
probably
the
most
complicated
one.
G
The
yeah-
and
that
brings
back
the
question
that
we
had
when
we
were
talking
with
jordan,
about
the
delete
cap.
Are
we
ever
going
to
make
a
major
version
change
to
cube
control
right
like
yeah?
That
definitely
is
something
we
have
to
talk
through,
because
I
know
I'm
sure
you
have
a
list
of
stuff
that
you'd
like
to
break
and
change,
and
I
I
definitely
have
my
list
of
stuff
so
yeah.
A
A
G
A
That
gives
us
about
two
releases
to
figure
out
and
sync
with
both
of
those
six
about
the
processes
and
the
necessary
changes
that
would
have
to
happen,
and
in
the
worst
case
we
would
just
slip
it
to
yet
another
release.
A
If
something
weird
pops
up,
I
guess
that
all
the
discussion
would
have
to
be
written
down
in
some
sort
of
a
cap,
or
I
don't
know
something
that
will
allow,
give
everyone
a
path
to
follow,
because
I
think
there
are
others
interested
in
following
our
path.
A
A
Or,
although
maybe
or
I
I
don't
know,
probably
both
would
work.
Okay,
I'll.
G
I'll
get
this
on
the
agenda
for
sig
release
and
just
as
like
a
preliminary
like
hey,
you
know
this
is
gonna
happen
at
some
point,
probably
yeah.
I
did
start
thinking
about
this.
Actually,
I
I'm
building
a
prowl
plug-in
right
now
to
transfer
issues
between
repos,
because
I
feel
like
we're
going
to
need
this
for
moving
tube
control
issues
out
of
tree.
You
know
main
repo
kk
into
our
own
repo,
so
I
talked
to
sig
testing
about
that
and
we're
thumbs
up
on
moving
forward.
So,
okay,
cool,
that's
nice.
A
G
J
L
Back
on
the
on
the
topic
about
the
docs
and
what's
going
to
happen
with
them,
I
appreciate
being
looped
into
any
conversations
they're
having
around
that,
because
I
know
like
customize
on
our
roadmap.
We
want
to
do
a
lot
of
enhancements
to
our
docs
in
terms
of
content.
L
We
would
like
to
to
add
a
lot
of
content
that
we
don't
have
right
now
and
fix
it
up
a
lot,
and
I
think
that
would
be
extremely
challenging
to
do
if
everything
was
trying
to
consolidate
onto
the
the
main
kubernetes
website,
just
because
we're
in
a
small
corner
over
there,
and
I
think
the
arrangement
that
we
have
makes
a
lot
of
sense
and
we
could
even
reduce
the
content
on
the
main
website
so
that
we're
not
maintaining,
duplicates
and
and
just
have
mostly
a
pointer
to
the
fully
fully
expanded
documentation,
whether
that's
one
site
for
cute
pedal
and
customizer
or
two
two
sides.
L
I
don't
really
feel
strongly
at
this
point,
but
I
do
think
that
that
separation
makes
a
lot
of
sense.
So
if,
if
we're
considering
reversing
that,
then
then
I'd
like
to
be
part
of
the
conversation,
if
possible,.
G
Yeah
for
sure
I
had
just
assumed
that
we
were
fine
with
having
a
merge,
but
that's
good
to
know.
Is
there
a
specif?
So
I
I
understand
what
you
just
said,
but
is
there
a
reason
that,
like
what
it
would
a
redirect
would
like
customize
dot
io
redirecting
to
the
kubernetes
docs
solve
your
your
issue
there
or
you
really
want
it
separate
for
velocity
or.
L
L
It's
really
volume
of
content
and
like
information
architecture
that
I
think
would
be
extremely
challenging
to
do
well
on
the
kubernetes.website,
because
you
know
that
rightfully
has
a
lot
of
content
on
tons
of
different
topics
and
customized,
for
example,
would
be
just
one
little
sliver
of
content,
whereas
if
we
have
our
own
website,
we
have
all
of
these
possibilities
for
organizing
in
a
way
that
makes
sense
and
guiding
users
through
journeys
and
having
a
lot
of
different
pages
about
it.
L
And
if
you
look
at
where
customize
is
on
the
main
website
today,
you
know
you
drill
down
drill
down,
drill
down
drill
down
and
then,
if
you
have
an
entire
site's
worth
of
content
in
that
tiny
little
portal,
I
just
think
the
user
experience
that
would
be
able
to
to
offer
to
people
who
are
interested
primarily
in
like
really
diving
into
the
features
of
customize.
L
I
think
that
would
be
very
difficult
to
do
if
sig
docs
have
have
a
concrete
idea
of
what
that
would
look
like
I'm
open
to
changing
my
mind.
But
my
my
reaction
is
that
the
amount
of
content
and
the
level
of
detail
that
we'd
like
to
go
into
on
the
customized
website
would
be
very
difficult
to
integrate
into
a
small
corner
of
the
larger
site.
G
Okay,
cool:
they
are
I'll
I'll
dm
you
about
this.
We
will.
We
can
go
to
sig
docs
and
talk
about
it.
I
imagine
that
they're
just
going
to
be
like
we
can
make
whatever
you
want
work.
They
are
pretty
gung-ho
about
getting
everything
under.
You
know:
kubernetes
io,
docs,
but
I'll
I'll
support
either
way.
So
I'll
get
us
on
their
agenda
I'll
dm.
You.
L
A
Okay
hearing
none.
Thank
you
very
much
folks
as
usual,
that
was
a
very
nice
talking
to
you
all
and
see
you
in
next
week
for
sick
box
crop.
If
I
remember
correctly,
thank
y'all
bye.
All
thank
you.
Thank
you.