►
From YouTube: Kubernetes SIG CLI 20220316 - Kustomize Bug Scrub
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
Hello
and
welcome
to
today's
six
cli
meeting
today
is
a
customized
bug
scrub,
so
we're
going
to
go
through
the
issue
triage
board
together,
look
at
issues
that
haven't
been
triaged,
yet
discuss
anything
that
anyone
wants
to
bring
up
about
customized
issues
and
offer
opportunities
as
to
go
through
them,
potentially
for
a
contributor
to
claim
an
issue
if
they're
interested
so
at
any
time
feel
free
to
bring
up
an
issue
that
you're
interested
in
getting
feedback
on,
or
you
want
to
bring
to
our
attention,
including
now
is:
does
anyone
have
anything
that
they
want
to
start
with
today?.
A
A
A
For
patches,
we
responded
that
we
would
need
it
to
be
a
specifically
targeted
operation
like
we
wouldn't
support
it
generically,
and
what
was
it
exactly?
We
had
some
stipulations
about.
How
it
would
have
to
work
basically
would
be
like
a
separate
field.
That
would
say
you
know
this.
This
target
is
actually
json
and
here's
what
to
do
with
it.
A
A
Explained
this,
I
think
we
would
want
to
accept
it
as
a
patch.
First,
that's
been
a
more
popular
request,
trying
to
think
if
the
same
thing
would
apply
to
replacements
or
if
it
would
be.
C
A
I
don't
know
I'm
inclined
to
think
like
this.
Implementation
needs
to
be
related
to
the
other,
one
that
we
have
like
the
one
that
I
like
that
I
was
showing
here
where
jeff
is
explaining
what
we
want.
So
what
what
it
would
look
like
to
patch
structured
data
within
a
string
field
as
a
when
you're
using
one
of
the
patching
mechanisms
and
what
it
would
look
like
to
do
the
same
thing
with
replacements.
Like
do
you
actually
be
related.
A
Yeah
and
we
asked
for
a
mini
cap
to
be
submitted
because,
like
this
is
a
major
feature
for
us,
like
we've,
repeatedly
declined
to
implement
this
and
then,
after
that
long
conversation
between
jeff
and
and
on
the
other
issue,
we
came
to
an
agreement
about
what
it
could
look
like,
but
like
for
contentious
issues,
I
think
we
do
want
to
have
a
capsule
I'll
understand.
What's
going
on
so
yeah,
I
agree.
Yeah.
C
I
think
that's
fair.
We
can
ask
them
to
submit
a
small
cup
and
also
refer
to
the
patches
issue.
A
Yeah
that
that
is
good
here,
I'm
gonna
stop
sharing
my
screen
for
a
minute,
I'm
not
logged
into
github,
which
means
I
can't
comment
just
a
little
bit
of
a
problem
for.
A
Do
you
want
me
to
comment
yeah?
That
would
be
great,
but
I'm
still
gonna
sign
it.
So
I
can
do
it
during
this
meeting
here.
A
All
right
so
you've
got
this
one.
Do
you
want
me
to.
A
A
A
C
I
I'm
not
sure
what
they
mean.
Their
request
is
a
little
vague.
A
A
D
B
A
A
C
Yeah
the
merge
key
for
containers
is
port
and
protocol,
I'm
not
sure
what
it
does
for
sub
fields
and.
B
C
C
C
This
is
maybe
a
dumb
question,
but
I
tried
to
do
a
find
for
container
and
in
the
open
api
file.
Does
it
exist
yeah?
I
can't
find
it
they
might
be
trying
to
do
and.
A
Yeah,
they
are-
and
I
just
used
my
reproduction
case
and
that
totally
works,
and
that
is
exactly
the
problem
you
could
find
name
is
and
not.
What
is
a
container
container.
B
Good
catch
patching
of
resource
names.
C
Is
this
something
that
replacements
could
handle?
It
just
looks
like
they're
trying
to
replace
a
value.
A
C
B
Okay:
let's
go
grab
the
reference.
C
For
oh,
their
last
point
is
that
there's
a
lot
of
duplicate
code
to
add
to
every
definition
of
a
deployment
which
is
like
their
problem
with
using
patches.
I
don't
know
if
replacements
will
be
better
if
they
want
to
avoid
that.
A
A
E
D
A
Because
they
only
have
to
target
one
thing,
so
it's
not
like
we
can
don't
know
customize
sometimes
just
is
really
verbose
just
inherently
because
our
philosophy
is
that
we
want
to
be
very
explicit
about
what
you're
trying
to
do
so.
Sometimes
that
does
mean
like.
If
you
want
to
do
multiple
transformations,
you
have
to
always
have
to
supply
multiple
specifications.
C
C
A
To
barb
as
quacks
whatever,
and
they
don't
like
doing
both
of
these
things
in
all
of
those
various
overlays,
that's
my
reading
of
it.
You
could
be
wrong.
A
No,
but
I'm
guessing
like
the
name,
space
name
is
what's
varying
between
them
between
the
deployments
like
typically,
if
you
have.
C
A
It's
actually
this
description
of
why
they
don't
like
it
negates
the
benefit
of
just
setting
namespace
bar.
It
kind
of
reminds
me
of
the
problem
that
we're
tackling
around
identifying.
A
Field
specs
in
a
better
way.
A
C
B
D
A
But
fieldspec
it's
just
a
dgbk
path,
great,
if
not
present,
yeah
there's
nothing
about
field
delimitation
like
replacing
things.
C
A
B
A
I
guess
replacements
is
the
latest
and
greatest
version
of
it
and
in
terms
of
the
where
we
have
duplicate
functionality,
that's
the
way
of
the
future,
whereas
field
specs
are
well
related
to
that
issue.
That
you're
working
on
designed
for.
A
Like
we
have
the
configurations,
the
open
api
of
all
sorts
of
ways
of
pointing
our
transformers
to
fields
and
we're
going
to
need
to
reconcile
that.
A
I'll
assign
myself
to
respond
to
this
one
and
say
something
around.
We
definitely
wouldn't
do
a
a
variable
generic
variable
feature.
In
fact,
we
had
something
along
the
lines
of
that.
It's
deprecated.
Here's
the
link
to
the
this
features
document.
However,
it
sounds
like
what
you're
trying
to
do
is
inform
customize
that
the
name
space
appears
at
a
specific
path,
and
you
can
do
that,
but
not
for
a
sub
part
of
the
path,
and
that
would
be
a
feature
we
would
consider
would
look
something
like
this.
What
do
you
think.
B
D
C
C
Yeah,
I'm
on
a
stack
overflow
site
that
says,
based
on
the
json
rfc,
there's
no
way
to
select
an
element
by
property.
A
Could
you
link
that
or
the
specification
so
or
maybe
just
not
do
a
quick
response
to
them?
This
seems
like
it's
just
a
question:
they're
not
asking
us
for
a
feature
necessarily
and
well.
The
answer
is
obvious.
Also
if
it
was
a
feature
request,
you're,
not
gonna
change,
the
you're
not
gonna,
do
anything.
That's
not
inside.
A
A
D
C
It
wasn't
yeah,
we
can
accept
this
issue.
We
found
this
on
this
guy's
pr
and
I
told
him
to
file
an
issue.
So
he
did
it's
legitimate.
C
A
C
Yeah,
so
it
works
if
you
do
like
the
key
equals
value
syntax,
but
not
any.
C
I
guess
one
possibility
is
that,
like
there's
a
name
somewhere
in
the
replacement
itself-
and
you
want
to
insert
this
new
thing
at
like
a
specific
like
in
a
specific
order
at
a
specific
index
of
the
array
that
might
be
a
little
bit
niche.
B
C
A
Well,
that's
the
opposite.
I
just
I
said
one
thing
and
I'm
renaming
it
to
the
opposite:
replacement
filter
create
option.
D
A
C
Yeah
all
the
other
ones
do
have
examples.
C
If
we
have
a
transformers
section
on
the
new
site,
we
can
something
about
namespace
transformer.
A
Because
there
is
already
what
was
I
looking
at
before?
I'm
sure
it's
in
here
all
the
transformers
are
in
here
I'm
pretty.
D
A
D
C
C
Maybe
not
but
like
for
replacements,
the
hash
doesn't
get
propagated
correctly.
C
A
It
is
a
real
field,
as
you
can
see
here
is
your
destiny
as
your
file?
What's
there
right
and
how
are
they
doing
this?
They
are.
This
is
just.
C
A
place
it
might
be
just
is
customize
aware
of
this,
because
if
it's
not
listed
in
name
references
customize
might
not
be
looking
at
it.
A
A
A
Get
to
here
this
one
patch
delete
not
working
as
expected.
I
think
this
is
a
duplicate.
A
B
That
is
not
the
one
I've
commented
on,
maybe
there's
a
third
this
one
might
be.
It.
A
C
C
A
All
right
with
that,
I
think
I'll
call
it
a
day.
We
have
only
two
minutes
to
spare
it.
That
would
be
a
pretty
rush
for
getting
through
another
one.
Necessarily
so
thank
you,
everyone
for
coming
today
and,
as
always,
feel
free
to
reach
out
to
us
in
the
customized
channel,
and
please
join
us
at
the
other
bi-weekly
or
at
the
weekly
sig
meetings
that
we
have
next
week
will
be
a
full
sig
meeting.