►
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
A
B
Hello,
everyone
so
I'm
I'm,
pratik,
Mota
and
I'm
from
India
and
currently
I'm
a
golang,
C
plus
plus
developer,
so
I
have
previously
used
the
operator
SDK
for
creating
crds.
So
right
now,
I'm
here
to
learn
the
cube,
Builder
and
also
contributing
it
so
yeah
right
now
really
reading
its
documentation.
So
yeah
looking
forward
to
work
with
you
as.
A
A
community
that
is
great
you,
you
are
very
calm,
and
definitely
your
help
is
more
than
you'll
come.
Please
feel
free.
That
is
good
for
everybody
knows
as
well
feel
free
to
jump
in
any
issue
that
is
opening
the
repository.
We
have
the
label
Gucci
first
issue,
something
like
that.
Yeah
yeah,
that
are
good
ones,
also
feel
free
to
jump
in
the
channel
and
just
help.
A
C
Yeah,
so
mine
is
a
controller
runtime
specific
thing
that
issue
there
is
a
a
proposal
issue
for
creating
a
new
cache
implementation,
controller
runtime
for
gracefully
handling,
our
back
changes
and
so
kind
of
what
that
means
is.
C
It
would
be
a
caching
layer
that
is
highly
configurable
in
a
sense
that
you
can,
you
know
dynamically,
add
and
remove
your
watches
in
and
handle
our
back
changes.
So
currently
what
would
happen
if
say
you
know
you
have
certain
permissions
on
your
controller
and
you
go
and
your
control
is
running
on
cluster
and
then
a
cluster
admin
says:
oh,
hey,
I!
Don't
want
this
controller
to
have
these
permissions
anymore.
C
What
will
happen
is
it
will
likely
just
crash,
and
you
know,
start
logging
a
whole
bunch
of
errors
and
enter
a
blocking.
Loop
and
kind
of
the
goal
with
this
proposal
is
that
we
would.
You
know
now
have
a
caching
clear
that
enables
operator
authors
to
write
their
controllers
in
such
a
way
that
it
can
handle
changes
in
our
back.
So,
for
example,
you
know
say:
I
have
a
controller,
that's
reconciling
the
foo
CR
and
what
it
does
when
it
reconciles.
C
Fucr
is
it
creates
a
secret
and
a
deployment
in
the
namespace
that
the
fusiar
is
in.
You
know,
if
that
if
a
cluster
admin
is
okay
with,
you
know
that
that
controller
having
full
permissions
to
you,
know,
create,
read
git
list,
watch
whatever
secrets
and
deployments
it
wants
across
the
cluster.
That's
awesome.
They
can
allow
that
or
if
they,
you
know,
only
want
it
to
be
able
to
do
that
in
namespaces,
like
Foo
and
Bar
it
will.
You
know
the
the
Opera,
the
controller
can
say:
okay,
I'm
not
allowed
to.
C
C
So
therefore,
it's
more
like
gracefully
handling,
you
know
permission
changes
in
our
back
and
stuff
like
that.
There's
really
a
lot
more
to
it
than
there's
a
lot
that
is
going
on.
Look
like
there's
a
lot
more
information
in
this
proposal
and
in
a
comment
later
down
in
The
Proposal
that
if
you
know,
people
want
to
get
more
more
insight
into
it
and
you
know
happy
to
to
have
that
discussion
on
the
issue.
C
If
anyone
wants
to
discuss
it
here,
you
know
I'm
happy
to
discuss
it
here
as
well
and
just
get
people's
thoughts
and
opinions
and
see
if
this
is
something
that
the
community
feels
would
be
really
nice
to
have
in
controllers
or
not.
C
D
Sorry
for
being
late,
maybe
I
missed
it,
but
I'm
trying
to
understand
I
the
use
case
and
the
the
problem,
because
that's
looks
like
a
very
big
potential
PR.
So
like
what
is
the
use
case.
C
Yeah,
so
so
the
use
case
here
is
you
know,
it's
making.
Operators
and
controllers
become
workloads
that
are
able
to
be
kind
of
that
are
kind
of
scopable
in
their
permissions.
C
So
you
know,
if
you
think,
about
a
lot
of
like
Industries
like
banking,
or
you
know,
even
potentially,
government
and
stuff
like
that,
where
their
industries
that
and
organizations
that
require
like
high
level
of
security
on
their
clusters,
you
know,
operators
in
their
current
state
can
be
kind
of
concerning
for
them,
because
a
lot
of
times
when
you
go
to
install
an
operator,
you
know
they're
they'll
be
asking
for
cluster-wide
access
to
certain
resources
and
if
a
cluster
admin
does
not
want
that,
there's
no
real
way
that
they're
pretty
much
having
to
decide
between
installing
this
operator
or
controller,
or
actually,
you
know,
keeping
their
cluster
more
secure,
and
so
this
kind
of
targets
the
ability
to
make
these
workloads
that
their
permissions
scoopable,
like
other
kubernetes
workloads,.
E
So
so
maybe
maybe
for
what
I
know
currently,
we
support
scope
to
operators
so
like.
What's
what
is
the
challenge
with
the
current
implementation.
E
C
Yeah,
so
currently,
the
challenge
with
the
current
implementation
is
that
you
can
really
only
scope
it
down
to
like
a
certain
number
of
namespaces,
and
then
it's
kind
of,
like
you
have
the
operator
or
controller,
has
to
have
those
dedicated
permissions
in
those
namespaces,
and
if
those
permissions
are
removed,
then
the
controller
will
start
crashing
or
entering
a
blocking
State
and
no
longer
be
able
to
function
properly.
C
So
if
a
cluster
admin,
you
know,
changes
their
mind
on
the
permissions
that
they
want
this
controller
to
have
they
either
have
to
completely
reconfigure
and
restart
the
controller
or
operator
so
that
it
runs
with
those
new
permissions
or
just
be
okay
with
it
crashing
this
proposal
kind
of
makes
it
so
that
those
changes
can
be
a
little
bit
more
Dynamic
by
just
using
so
the
kubernetes
native,
our
back
controls.
So.
E
What's
so
badly,
is
it
crashing
right
if
you
change
the
configuration
like
like
for
me,
that
sounds
like
it's
your
problem,
if
you,
if
you
are
tampering
with
the
solution
right,
it
will
crash
and
then
you'll
have
to
fix
it
like,
because
because
this
kind
of
changes
is
basically
breaking
the
current
API
and
it's
complexing
stuff.
So
how
big
is
the
problem
in
order
for
us
to
introduce
this
huge
change
and-
and
it's
not
only
breaking
the
system,
it's
also
having
someone
invest
the
time.
E
F
C
Yeah,
so
so,
just
for
some
perspective,
I'm
a
maintainer
of
the
operator
SDK
in
Cube
Builder
and
in
The,
Operators
Decay,
there's
been
a
couple
issues
raised
regarding
you
know
like
I
want
to
make
it
so
that
my
operator
can
be
restricted
to
only
watch
or
receive
stuff
like
information
on
secrets
in
a
specific
namespace,
but
I
want
it
to
be
able
to
operate
like
I,
want
that
to
be
able
to
be
configurable
and
the
operator
Not
Crash.
C
If
a
cluster
admin
wants
to,
you
know
enable
more
namespaces
that
it
can
watch
and
so
essentially
asking
kind
of
for
this
functionality,
but
they
kind
of
scoped
it
down
to
you
know
one
specific
use
case,
but
the
goal
of
this
proposal
is
to
make
it
more
Broad
and
make
that
kind
of
functionality
available
for
all
controllers
and.
C
C
You
know,
if
I'm,
if
I
don't
have
the
permissions
to
operate
in
this
namespace
I'm
just
going
to
update
the
CR
status
and
say
I,
don't
have
those
permissions,
you
know,
and
it
really
makes
gives
some
of
the
that
power
towards
the
cluster
admins
to
say:
hey
I
know
that
the
operator
is
asking
for
these
certain
permissions,
but
I
really
don't
feel
comfortable,
allowing
those
permissions
on
my
cluster
for
namespace
XYZ,
but
really
I'm,
okay
with
it
being
in
ABC,
because
I've
allocated,
you
know,
namespaces
ABC
to
being
less
secure
or
what
have
you,
but
hopefully
that
answers
your
questions.
C
A
So,
let's
give
you
the
the
chains
for
the
orders
as
well,
injuring,
maybe
time
more
time
box
go
Forge,
Michael.
G
I
just
wanted
to
add,
like
I've
run
into
this
too
I.
Think
Bryce's
example
is
a
great
one.
It's
just
if
you
have
a
cluster
scoped
operator,
because
you
know,
customers
who
knows
wants
to
be
able
to
reconcile
a
custom
resource,
and
you
know
cluster
wide,
but
the
operator
needs
to
like
read
a
secret,
as
was
mentioned
from
one
specific
name
space.
The
the
kind
of
the
only
way
that
works
today
that
I'm
aware
of
is,
if
you
give
the
reconciler
an
API
reader
like
a
client
reader,
which
just
bypasses
the
cache.
A
C
And
I'll
I'll,
just
also
I,
think
Michael.
You
made
me
think
of
something
that
one
of
the
other
possible
scenarios
with
a
controller
you
know
implemented
in
the
current
pattern.
C
Is
you
know
it
may
hard
code
certain
things
in
there
and
therefore
not
be
configurable
at
all
by
a
cluster
admin
that
maybe
wants
to
change
those
permissions
and
while
that
might
be
a
bad
practice,
it's
something
that
could
exist
out
there
and
something
that
should
be
you
know,
could
be
combated
against
with
something
like
this,
and
also
the
goal
is
to
try
to
focus
on.
C
You
know
what,
however,
this
is
implemented
whenever
it
is
implemented
if
it's
implemented
in
controller
runtime
or
if
it
ends
up
being,
you
know,
I
see
the
value
in
this
and
I.
Do
it
myself
in
my
own
repo
and
maintain
it
as
a
library.
C
The
goal
is
to
make
it
as
easy
to
adopt
as
possible
on
the
operator
author
or
controller
author
side
of
things,
with
the
goal
of
it
being
pretty
much
just
a
plug-in
replacement
for
like
a
drop-in
replacement
for
current
cash
implementations.
That's
the
end
goal
in
terms
of
like
difficulty
for
adopting
it.
Obviously,
difficulty
for
implementing.
C
It
is
a
different
story,
but
I've
worked
on
a
couple
different
proof
of
Concepts
and
have
links
to
proof
of
Concepts
and
examples
like
an
example
operator
implementing
that
proof
of
concept
in
that
proposal,
issue.
F
Yeah
I
won't
take
much
of
time,
but
I
understand
where
Bryce
is
coming
from,
based
on
the
use
case,
I
haven't
taken
a
look
at
the
POC
which
I
had
price,
but
I
just
wanted
to
bring
up
yet
another
PR
which
I've
been
following,
so
that
allows
to
start
and
stop
the
watches
and
the
controller
on
runtime.
Probably
that
could
be
helpful.
The
only
thing
is
again,
though:
it
allows
to
start
and
stop
the
watches.
It
is
something
which
can't
be
done
dynamically.
F
The
user
has
to
still
pass
on
a
canceled
context
through
the
manager,
so
that
I
don't
have
the
open
PR.
It
was
created
by
someone
else,
I
just
attached
the
link
to
it
in
the
chat.
So
probably
that
would
be
a
great
idea
to
get
started
with,
because
there
have
been
discussions
on
how
to
actually
remove
objects
from
the
cache,
because
stopping
and
starting
controller
is
still
fine,
but
removing
objects
from
cash,
and
it's
one
major
task
which
is
being
discussed.
F
So
if
the
POC
which
you
have
created
has
that
implemented
and
that
somehow
can
be
used
in
this
area,
then
that
would
be
really
great.
So
I
just
wanted
to
bring
that
up.
A
A
Folks,
everybody
that
he
is
interested
in
this
subject
go
there
checking
they
should
check
the
information
and
the
the
POC
that
he
made
as
well
in
to
make
all
questions
you
know
challenging
him
there
and
she
brought
more
light
in
the
description.
A
Is
that
okay?
Could
you
agree
on
that
change
movie,
for
it.
G
C
Know,
please
feel
free
to
go
to
that
proposal
issue
and
leave
your
feedback.
C
You
know
happy
to
to
continue
discussion
there
and
it's
really
what
I'm
hoping
for
is
having
you
know
a
nice
big
discussion
there
in
that
issue.
You
know
talking
about
all
the
pros
cons
and
whether
or
not
this
is
something
that
is
is
really
useful.
So
thanks
everyone,
another.
A
Thing
that
you
like
just
it's
your
share
this
like,
if
you
can't
contact
one
time,
it's
not
very
common,
they
do
design
proposals
requests
with
the
designer
proposals,
but
if
you
see
that
you
get
too
much
stock
like
to
share
the
user
games
right
to
properly
communicate
the
information
that
you
want,
maybe
that
can
be
good
strategy.
A
A
H
That's
good
enough:
okay,
yeah!
Thank
you.
This
is
about
moving
the
declarative,
plugin
out
of
view
Builder
into
Cube
Builder,
declarative
pattern
repository
and
basically
make
it
an
external
plugin.
So
I
have
been
looking
into
this
in
in
the
past
two
weeks,
I
kind
of
hit
a
blocking
issue
at
the
moment.
H
It's
it's
basically
related
to.
How
did
the
plugin
would
check
if
the
project
is
using
a
legacy
layout
or
not?
That
check
is
currently
done
in
a
way
that
it
will
basically
just
look
into
the
plugin
chain.
If
go
language
or
gonad
V3
are
there
then
the
pro,
then
the
project
is,
is
using
a
legacy
layout.
Otherwise,
it's
not
and
based
on
that.
It's
going
to
be
doing
a
scaffolding
accordingly.
H
So
this
is
currently
possible
when
using
the
declarative,
plugin
internally
as
a
built-in
plugin
to
Q
Builder,
because
basically
we
inject
config,
but
if
it's
an
external
plugin
the
config
is
just
lost
in
that
call.
We
don't
really
pass
that
information
in
the
plugin
request,
so
making
that
check
is
kind
of
not
feasible,
specifically
in
the
internet
command,
so
that
that's
one
issue
there
and
the
second
issue
is
also
related
to
this.
H
I
was
looking
into
using
maybe
the
project
file
to
to
kind
of
figure
out
whether
the
project
is
built
on
a
legacy
layout
or
not.
The
issue
is
that
the
project
file
I
believe
it's
something
that
gets
only
created
after
all,
the
scaffolding
is
done
in
the
init
command.
So
it's
it's
also
not
an
approach
that
I
think
can
be
taken
here
too.
H
To
fix
this
problem,
and
yes,
I,
guess
that
that's
my
that's
the
issue
that
I
had
here
so
far,
so
I
was
wondering
if
I
could
maybe
get
some
like
feedback
or
work
around
about
this.
A
So,
let's
just
see,
if
I
could
probably
summarize
the
blockers
that
you
you
have
now
right,
the
first
one
you
are
having
problems
because
to
reach
the
project
config
because,
as
I
started
working,
you
are
saying
that
you
don't
have
the
project
config,
yes,
okay,
but
today,
starting
up
plugging
also
will
be
usage
with
it
could
be
Builder
right.
So
that
is
the
idea.
Has
the
interface
engineering
implements
the
same
interface?
You
have
the
contact
right,
so
the
project
config,
is
a
contract.
What
you
can
do
is
like
e,
but
you
could
also.
A
A
B
C
One
idea,
yeah
so
I
think
I
think
with
the
core
problem.
Here
is
an
many
feel
free
to
correct
me
if
I'm
I'm
wrong,
but
so
the
external
plugin
has
a
interface
that
is
supposed
to
contain.
What's
called
like
the
universe
like
the
plug,
like
the
the
universe
of
files
and
I.
C
Think
with
the
problem
that's
being
run
into
here,
is
that
currently,
whenever
an
external
plugin
is
called
that
universe
is
empty
and
really
the
it
should
be
filled
by
all
the
files
that
are
supposed
to
be
created
on
in
the
file
system
by
the
plug-in
chain.
C
So
even
if
this
is
chained
with
like
the
regular
go,
V3,
scaffold
and
stuff
like
that,
and
then
you
chain
in
this
external
plug-in
that
chain
that
Universe
of
files
that
are
created
from
previous
plugins
in
the
chain
isn't
actually
properly
updated,
and
it
is
always
empty
and
I
believe
that's
a
bug
and
is
something
that
we
need
to
think
about.
C
How
do
we
take
that
plug
and
chain
file
system
and
add
it
to
that
universe
so
that
when
an
external
plugin
is
called
in
that
chain,
it's
able
to
read
that
because
I
think
many
made
a
slack
post
about
this
issue
and
we
had
some
comments
back
and
forth,
and
that
was
like
the
biggest
problem
is
like
they
were
trying
to
use
it
in
this,
like
they
were
testing
it
in
this
plug-in
chain.
C
C
They're
like
the
project
file,
was
never
there
when
they
attempted
to
look
for
the
project
file.
Can.
A
I
make
that
question
so
like.
If
you
call
Kobe
builder
in
it,
could
be
Builder
entertaining
scaffold
to
go
V3
first,
then
you
call
the
declarative
afterwards
as
a
Sterno
plugin,
then
the
universe
is
filled
with
the
config.
The
problem
is
like:
if
I
want
to
call
golf3
aims,
they
start
not
plugging
to
do
this
code.
C
C
First,
we'll
check
if
it's
in
a
chain
we'll
check
and
make
sure
in
that
plug-in
chain,
it's
been
run
and
we
can
verify
that
by
the
project
file
or,
however,
in
the
universe.
But
if
there's,
if
it's
used
by
itself,
we're
going
to
expect
that
certain
files
already
exist
on
the
file
system
that
we
can
read
right
or
now.
A
What
do
you
think
about
Chile?
He
tackle
that
it's
like
for
face
tool
for
the
the
facial
plugins,
the
API
implementation.
For
now,
at
least
we
have
the
limitation
where
you
cannot
use
the
Sterno
plugin
should
do
the
goaling
scaffold.
So
do
the
basically
right
you
need
to
have
the
project
it's
called,
then
you
can
use
external
plugin
to
opt-in
for
features
on
top.
C
Do
we
say
that
that's
on
the
external
plug-in,
like
there's
a
few
conversations
and
decisions
that
I
think
we
need
to
make
and
take
away
from
this
I,
don't
know
if
there's
necessarily
a
way
that
we
can
say
right
now.
This
is
the
way
forward.
Let's
do
that.
C
G
A
S
they're
fine,
now,
okay,
I
tried
to
just
directly
here,
I
think
it
yes,
one
thing
so
I
think
King
King.
We
create
a
specific
issue
with
this
scenario,
like
what
you
are
trying
to
do
in
the
the
expected
Behavior
into.
What's
your
thing:
is
that
possible
me
Magic
Kingdom
of
Grace?
That
is
great.
The
other
thing
justify
you'll,
be
unblocking
moving
forward
with
the
work
of
the
startup
plugin.
Could
you
try
to
reach
the
project
file
and
you
take
the
decision
and
she
make
adults
that
you
did?
A
Like
the
personality,
though,
creates
the
project
and
the
thinking
use,
the
plugin
cannot
use,
it
could
be
building
image.
Plugins
engineering
tests,
the
declarative
one,
just
if
I
don't
block
you
and
if
you
don't
find,
if
he,
if
he's
saying
page
I,
think
you
can
consider
it
like
a
other
full
Behavior
go
V3
or
go
before
no
matter
just
before
they
will
be
able
to
to
move
as
it
stands
alone,
as
well
like
just
before
you'll
be
able
to
test
until
we
discuss
and
address,
then
it's
in
the
API
Justice
won't
block
it.
G
H
A
A
A
Okay,
so
that
is
great,
you
are
unblock.
That
is
great.
We
achieve
our
goal
really.
Thank
you
for
your
help.
That
is
amazing
that
you'll
go
to
choose
to
to
do
right.
Jonathan
as
well
you'll
be
very
happy
with
this
implementation
for
sure
anybody
would
like
to
to
put
any
other
input
about
to
choose
one.
A
No
silence
and
you
move
forward
the
next
one.
It's
me
I
put
a
cheese
one
here
right
that
is
very
I
I.
Just
should
you
like
to
share
the
idea?
The
jlgb
try
to
share
the
quality
of
the
projects
with
the
gold
link
right
we
go,
learn,
CI
links,
we
use
the
Dutch
and
copy
Builder.
We
check
everything
that
we
Implement
in
there
could
be
if
you
go
alongside.
A
We
don't
want
to
put
that
in
the
first
cup
for
which,
because
these
are
in
the
past,
because
it's
a
third-part
API,
but
we
have
people
that
you
asked.
Why
did
the
full
discover
which
doesn't
check?
Because
the
first
couple
can
have
it's
just
that
it
doesn't
pass
in
the
the
full
shakes
right,
but
science
is
at
35
GPI
and
we
have
the
goal
to
try
to
minimize
the
the
dependence
to
make
it
easier
to
keep
maintained
and
everything
that
was
not
accepted
accepted
in
the
past.
A
So
one
thing
that
I'd
like
to
propose
is:
if
it's
someone
would
like
to
work
on
she's.
Well,
if
everybody
is
okay
with
that,
the
agreement
that
we
had
in
the
past
was
okay.
We
cannot
put
in
this
scaffold
into
the
frisco
Forge
in
the
process
that
she
had
done
by
the
two,
but
we
could,
in
our
CI,
linked
the
code
that
we
generate
with
the
two.
So
we
could
ensure
that
water
could
be
Builder
produces,
is
actually
a
linkage
enjoyable
passing
in
the
common
in
the
full
sex
right,
don't
need
your
checks.
G
A
Yes,
yes,
okay,
so
I
just
would
you
like
to
to
share
this
one
here?
Has
someone
any
objection
after
we
try
to
to
address
using
the
CI,
as
we
are
greeting
the
best.
A
The
idea
would
be
likely,
for
example,
your
show
for
you
right
when
we
we,
when
you
push
a
request
like
a
cheese
one,
no
matter,
we
usually
have
another
check
here
that
she
would
link
they.
They
detached
data
signals,
for
example.
If
he
will
find
her,
then
we
will
failing
the
side.
So
you
know
that
you
did
the
projects
that
we
are
producing
with
the
Kobe
Builder.
A
It's
not
a
person
in
the
winter,
so
you'll
be
able
to
fix
in
order
to
get
to
the
pr
the
change
message
right,
like
all
that
should
be
produced
with
the
two.
The
final
result,
it
would
depends
in
the
link,
but
we
would
not
add
a
new
make
it
file
Target
like
a
cheese
ones,
so
what
they
did,
the
link
to
Edge
the
the
gold
Lynch
link,
as
we
have
in
our
make
file
right
like
a
two
maintain
the
project.
C
A
It's
not
a
plug,
it's
just
a
CI,
not
chip,
login,
not
nothing.
Actually,
that
someone
can
use
is
only
a
shaky
for
the
weekend.
Shirt
that
the
code
that
you
read
delivery
can
be
linked.
So,
for
example,
one
example
of
the
implementation
in
the
test
is
in
the
CI.
We
could
just
insert
here
a
Target
right
and
then
call
in
the
CI
to
to
do
the
check
but
see
you
do
not
deliver
cheese,
new
targeted
for
into
user,
and
we
would
also
not
provide
any
feature
like
for
for
someone
about
to
Dutch.
A
Yes,
for
the
production
ensure
that
everything
weighing
someone
like,
because,
for
example,
today,
if
I
go
here
right,
I
take
it
to
enjoy
build
to
the
projects.
Then
I
came
in
here
in
the
making
file
enjoy
Edge
a
new
Target
to
we
started
to
use
the
link
to
check
the
result
right.
The
project
will
fails
in
basic
things
in
the
default
checks.
A
So
if
you
have
a
CI
job
the
cheap
period,
we
measure
changing
the
scaffolds.
We
check
the
result
to
see
if
he
always
passing
in
the
goal
inch.
The
result
that
the
tool
provides
would
be
validating
with
GPS
socially
in
the
in
the
golden
sets,
but
it
had
is
not
a
goal
as
the
the
making
file
Targets
in
the
projects
or
create
a
plugin
just
to
do
making
file
targets.
I
Gotcha,
so
this
is
like
we
won't
modify
the
scaffolds
of
the
make
fire,
but
rather
just
we
modify
the
CI
for
the
computer
itself
say,
have
another
job
stage
and
just
try
to
run
like
gold,
learn,
CI,
link
and
check.
If
the
test
data
works
for
or
or
like
respect,
our
CI
policies.
E
So
like,
why
not
just
adding
a
command
to
the
make
file
to
the
generated
make
file?
Because.
A
A
Yes,
we
can
do
that
to
do
the
testes
the
problem
and
show
using
the
in
the
East.
Coast
is
because
of
how
much
more
dependence
do
you
have
more
heart
is
for
we
keep
the
production
maintaining,
and
that
is
a
third
person
right.
So
ideally,
we
would
like
to
avoid
to
heavy
dirty
Parts,
because,
okay,
that
is
I,
agree
with
you
that
it
changes
the
common
usage
right.
But
probably
if
you
look
in
the
internet,
you'll
find
many
other
approaches
that
she
has
the
same
purpose.
A
E
But
like
the
the
way
I
see,
it
is
exactly
as
the
opposite
like
when
we
integrate
the
golang
CI
lintel
to
our
a
tool
chain.
We
basically
are
bringing
to
our
tool,
another
dependency
and
another
responsibility,
rather
than
like
the
way
the
make
file
work
is
that
if
you
don't
call
the
lint
Target,
then
it's
not
downloading
the
linter
now
downloading,
but
that's
how
it
worked
for
any
other
thought
about
the
executable.
But.
A
What
matters
is
we
see,
we
impacted
the
end
users
like,
for
example,
if
you
want
to
change
anything
in
Kobe
Builder,
that
is
a
big
exchange,
but
to
not
impact
is
someone
that
is
consuming
the
API
or
is
consuming
the
the
projects
that
are
built
with
copy
Builder.
That
is
fine,
so
also
in
the
Kobe
Builder.
You
see
like
we
check.
A
A
E
Not
the
only
I'm,
sorry
I,
just
don't
agree
with
you,
because
I
don't
think
it's
the
key
Builders
CLI
responsibility
to
do
the
linking.
If
you
want
a
lintel,
you
should
download
the
lintel.
If
we
want
to
increase
the
quality
of
the
project,
then
we
can
recommend
or
to
the
make
file
an
optional
linter,
but
it
shouldn't
be
the
responsibility
of
the
Q
Builder
CLI.
That's
the
way,
I
see
it.
A
That
is,
then,
we
have
two
points
to
see
that
right,
that
is
not
Championship
responsibilities
like
a
sponge
like
that
they
could
build
their
bring
final
result
for
the
users
that
she
pays
in
the
link.
That
is
good
enough.
Would
you
like,
as
a
maintainers,
be
a
device?
If
you
did
something
disc
golf
course
that
she
has
a
variable?
That
is
a
usage,
for
example,
we
scroll
up
with
something-
and
you
be
advised
in
the
CI
before
we
deliver
Energy
Management,
that's
changed.
Would
you
like
to
have
a
more
covert
better?
G
As
an
if
I'm
understanding,
I
think
I'll
mark
your
talking
about
something
different
and
what
is
suggesting
she's,
basically
just
suggesting
that
the
code
inside,
like
the
queue
Builder
repo,
is
all
linted.
Not
that
we
help
lint
like
not
that
anyone
downloading
okay,
Q,
Builder
binary,
would
lint
or.
A
Anything
like
that,
the
code
that
is
inside
of
the
Kobe
Builder
project
is
linked,
but
it
could
be
built
there
exactly
you
produce
a
code
right,
she's
end
result
is
not
linked
at
all.
So,
for
example,
we
can
go
there
and
do
a
question
in
the
template
we
put
here
because
it's
a
constant,
it's
a
big
string.
We
put
a
variable.
That
is
another
right.
Then
the
only
way
for
we
get.
That
is
in
the
review.
If
you
pay
attention
see
oops,
this
variable
is
not
usage
in
any
place.
A
A
G
A
E
A
Yes,
this
is
the
task
if
he
is
not
a
clearing
the
taxi
here.
I
can
put
it
because.
F
I
can
speak
on
behalf
of
Shirley.
It
is
too
late
for
Charlie
she's
based
on
European
time
zone,
so
I
just
wanted
to
bring
up
one
of
the
polls
created
by
her.
She
is
from
the
open,
Chef
monitoring
team,
and
we
had
prior
discussion
on
this
on
an
issue
on
whether
to
accept
the
monitoring
plugin
as
a
part
of
cube
Builder
phase
1.5.
F
But
then
we
discussed,
we
came
to
a
conclusion
that
it
would
be
better
off
being
an
external
plugin,
so
she
has
created
a
poll
in
the
slack
asking
if
her
users
would
benefit
from
a
monitoring
plugin,
which
has
all
the
best
practices
as
specified
in
one
of
the
operator.
Sdk
docs,
so
I
just
wanted
to
bring
this
up
on
behalf
of
her
and
ask
the
community
to
take
a
look
at
that
message.
Probably
give
your
thoughts
among
like
give
your
thoughts
on
that
check.
F
If
you
guys
are
comfortable
with
it,
whether
it
will
be
useful
for
you
all
or
whether
you
think
it's
not
useful
at
all,
we
need
something
else,
and
then
the
team
can
work
on
accommodating
that
so
yeah.
Please
do
take
a
look
whenever
you
all
have
time
and
feel
free
to
add
your
comments
or
thoughts
in
the
slack.
A
F
She
had
also
made
the
sample
on
that,
but
then
for
the
external
plug-in,
we
just
didn't
want
to
go
according
to
what
was
existing.
So
we
just
wanted
to
get
the
thought
in
general
about
how
about
a
plug-in,
which
has
all
the
best
practices
which
are
here
and
then
we
can
work
on
a
discussion
on
whether
to
have
a
library
or
whether
to
call
it
as
a
scaffolding
on
the
plugin
and
all
the
Implement
details,
implementation
details.
A
A
F
Yes,
it
is
inside
this
particular
folder
monitoring,
but
the
idea
is
to
take
a
look
at
this
and
check
if
all
these
scaffoldings
would
be
useful
to
you
all
how
we
will
scaffold
and
implement
the
implementation.
Details
are
totally
different,
which
we'll
discuss
later.
Okay,.
A
I
I
Remember
like
previously,
we've
been
learning
on
a
page
where
and
thinking,
if
ever
possible
to
practice
this
over
an
external
plugging
and
I'm,
not
sure
if
that
that
works
and
I
didn't
notice
that
we
have
actually
have
this
monitoring
folder
in
our
upward
SDK
go
project,
so
some
kind
of
the
things
I'm
I'm,
actually
not
quite
sure,
is
how
we
would
really
want
to
bring
about
to
the
user
because,
like
for
what
we
have
right
here
is
like
it
is
more
like
the
monitoring
Stacks
focusing
Prometheus.
I
It's
not
like
a
general
monitoring
plugin
that
can
work
for
like
monitoring
practices,
regardless
of
the
specific,
the
specific
stats
that
the
different
teams
will
be
using
now
right,
and
so,
even
if
that
is
focusing
on
Prometheus
because
like
as
we
consider
that
permit
is
very
popular
and
we
can,
we
can
see
that
in
like
I.
I
Remember
that
in
a
number
of
different
kubernetes
Community
about
other
projects
when
they're
trying
to
provide
some
monitoring
stuff
for
their
users,
they're
rather
directly
give
some
mixings
that
directly
give
some
metrics
or
prompt
cure
or,
like
alerts
not
necessary
to
like
go
over
like
guidance
or
documentations
or
programming.
I
I
I,
don't
know
so
that's
some
kind
of
the
things
where
it
comes
with
the
monitoring
plugin
and
was
trying
to
looking
at
that
and
I
I
felt
like
it's
something
that
is
there's
something
like
other
than
what
I
originally
thinking
of
so
I
saw
this
pole
and
I
just
marked
as
I'm
I
vote
for
I'm,
not
quite
sure.
If
I
ever
did
something.
F
Can
I
answer
that
Camilla
go
for
it
yeah
attorney,
you
were
absolutely
right.
A
cube,
Builder
is
a
generic
engine
and
it's
a
generic
project,
but
that
was
the
whole
reason
of
bringing
up
a
plug-in
plug-in
is
a
tool
in
which
users
can
customize
based
of
their
own
needs.
F
I
completely
understand
your
point
of
view,
but
the
idea
is
to
make
it
external
so
that
if
Cube
Builder
users
who
are
adopting
Prometheus
would
like
to
add
the
best
practices
they
can
use
this
plugin
someone
else
could
have
some
other
plugin,
which
is
based
on
Cube
State
metrics.
It
is
up
to
a
user
to
either
take
a
Prometheus,
take
Cube,
State,
metrics,
I,
think
Cube
State
metrics
also
uses
some
kind
of
services,
but
yeah.
F
It
is
just
an
available
ability
of
an
external
plugin
to
let
users
use
it
if
they
need
or
not.
But
yes,
as
you
said,
having
it
as
a
monitoring,
plugin
may
not
be
the
right
word.
Probably
the
right
word
would
be
to
use
as
a
Prometheus
Plugin
or
a
Prometheus
best
practices
plugin.
Something
of
that
sort
naming
convention
is
something
which
we
need
to
be
careful
of,
but
we
can
definitely
work
towards
it,
but
the
idea
is
to
get
a
feel
of.
F
A
E
I,
just
have
a
very
silly
question:
I'm,
trying
to
understand
the
relationship
between
the
operator
repository,
the
the
operator
SDK
repository
and
the
queue
Builder
like
why
some
stuff,
we
have
something
similar
with
cubillo
with
the
dashboard.
So
when,
like
some
stuff,
are
being
implemented
there
and
when
does
and
why
don't
we
merge
these
two
projects?
What
what's?
What's
going
on.
A
The
same
scaffolds,
however,
open
hmsgk
has
the
goal
to
Edge
helpers
on
top
meeting
with
the
proposed
to
integrate
the
project
with
their
Solutions
inside
of
your
Predator
framework,
like
competitor
lifecycle,
managed
for
your
distributes
for,
for
example,
openshift
openshift
shoes
uses
operator,
lifecycle
manage
also
operated,
sdgk
allows
people
create
projects
using
ansible
and
Helm,
and
that
is
not
a
desire
for
Kobe
Builder,
but
it
could
be.
Builder
provides
all
base
for
the
creation
top
and
javao
she's.
She's
achievements
is
that.
E
A
Think
I
think
LGB
very,
very
calm
like
a
plugin
for
operator
life
cycle
manager,
so
I
did
help
I,
don't
have
any
objection
to
that.
I
think
Jessie
raised
a
proposal
that
is.
E
A
A
A
We
have
a
if
it
actually
be
also
a
design
from
from
the
mountains
of
your
Predator
SDK
I.
Don't
see
why
not
I
think
we
need
to
do
a
proposal,
check
the
problems
and
check
the
cons
like
how
they
should
be
and
if
that
can
be
maintained
like
it
moving
the
forage
the
process
as
anything
here,
any
other
thing
else,
if
you
as
a
user,
would
you
like
to
have
kept
looking
inside
of
Kobe
Builder
that
she,
instead
of
he,
used
the
operator?
A
So
she
can
give
you
cheese
as
a
help
creating
an
issue
into
then
we
will
try
to
discuss
right,
I,
think
that
is
the
point
they're
going
to
discuss.
What
is
the
point?
She?
What
is
the
common?
So
what's?
What
is
the
pros
like
if
you
want
to
integrate
your
Predator
life
cycle
management
could
be
building
or
would
you
better
use
the
sdgk?
If
you
create
your
uploading
copy
Builder
could
just
important
as
well.
It
needs
to
be
responsible
to
maintain
copy
Builder.
So
all
these
discussions
need
to
go
foreign.
A
F
Yeah,
nothing
else
from
my
end,
Camila
covered
it
all.
There's
a
lot
more
work
going
on
the
olm
side
of
things,
but
yeah
having
it
as
a
plug-in
having
it
support
as
a
plug-in
makes
sense.
I.
A
I
think
it's
something
like
a
helper,
maybe
it's
very
nice
right
for
for
operator
framework,
maybe
I,
don't
know,
I
need
something
to
think
about,
maybe
like
if
I'm
Kobe,
Builder
user
and
I
won't
use
the
cheese.
Plugin
choose
operator,
lifecycle,
manager-
and
you
know
I,
don't
know.
Maybe
it's
a
good
idea
at
least
shows.
A
A
In
The
Proposal
meeting
the
best
I,
don't
know
if
that's
changed
has
few
things
that
she
in
my
humble
opinion,
would
you
feature
better
in
a
library,
prometrius
projects
I,
don't
know,
but
she
I
try
to
push
all
information
that
we
have
about
this
subject
so
far
and
please
feel
free
if
that
is
from
your
interest,
to
check
and
vote
and
give
your
thoughts
as
well.