►
From YouTube: Kubernetes KubeBuilder Meeting 20201020
Description
KubeBuilder Meeting for 2020/10/20. See https://sigs.k8s.io/kubebuilder for more details.
A
Usually,
usually,
we
say
something
like
welcome
to
the
and
then
the
meeting
so
like
the
q
builder
triage
meeting
for
what's
today,
the
20th
of
october
and
then
yeah.
This
is
going
to
be
recorded
and
posted
on
youtube.
B
All
right
I've
started
recording,
so
we
can
get
this
show
on
the
road.
B
Yeah
so
camila
and
I
have
like
the
introduction
that
sally
said
yeah,
so
welcome
to
the
very
first
q
builder
bug,
triage
meeting
or
issue
triage
meeting
on
october
20th.
This
will
be
recorded
and
this
meeting
will
be
recorded
and
potentially
put
in
venues
like
youtube.
So
just
make
sure
to
be
nice
and
I'm
looking
forward
to
seeing
how
we
started
formalizing
triaging
processes,
all
right.
C
We
are
open
to
suggestions
as
defined
a
new
flow,
the
milestones
that
we
have
now
artists
one.
So
we
have
the
v3
which,
in
my
mind,
should
be
just
you
test.
Is
there
to
be
a
block
before
we
do
that
release
the
v3
plugin
things
that
should
be
specifically
for
to
be
addressed
in
the
v3
plus
booking
into
the
next
one,
which
would
be
anything
next.
A
Just
a
note,
I
think
we
kubernetes
in
general
has
added
like
a
triage
bot
that
adds
the
untriaged
label,
and
then
we
can
add
the
triage
label.
I
think
milestone
probably
will
work
fine
for
us
as
well,
but
if
we
wanted
to
do
that,
that
might
make
it
also.
That
is
also
another
alternative.
C
C
So,
okay,
let's
start
this
one,
it's
one
mission
that
she
I
raised
yesterday
in
the
sgk
meeting
we
was
speaking
about.
The
tristana
shows
that
she
for
instagram
operators
for
some
cluster
is
not
possible.
Run
making
run
entities
was
the
solution,
so
the
idea
of
a
cheese
issue
is
just
she
tracked
that
we
might
need
to
shift
it.
So
that
is
just
check
it
if
they
make
it
run,
runs
in
this
kind
of
clusters
and
if
not,
we
could
try,
see
everybody
agree
using
the
rest
of
the
same
solution.
D
So
the
fix
for
this
was
adding
a
an
import
for
the
was
it
the
gcp
plugin.
That's
like
a
dummy
import
so
that
the
the
auth
plugin
works,
and
I
I
think
what
we
were
talking
about
yesterday
was
adding
this
by
default
to
the
scaffolded
main.go
in
coup
builder.
Is
that
correct.
C
C
A
Solution
what
it's
worth,
I
I
believe
we
do
import
at
least
the
gcp
plug-in
by
default.
I
don't
remember
about
the
other
ones
in
the
scaffolding.
I
think
off
the
top
of
my
head.
I
can
double
check,
but
if,
if
we
don't
it's
it,
it
seems
if
we,
if
we're
doing
that
and
we're
not
doing
it
for
the
others,
it
seems
reasonable
just
to
switch
to
the
the
general
off
in
the
scaffolding.
C
Would
you
like
you
to
put
in
some
milestone
or
we
don't
need
to
work
with
the
milestones?
That's
all.
A
Let's
put
it
put
it
in
in
the
next
milestone,
I
guess
I
think
we'd
have
a
yeah.
I
don't
think
there's
any
particular
pressure
on
this,
but
we
can
move
it
once
it
gets
accepted
or
once
it
gets
implemented.
D
Just
a
quick
question
about
labeling:
I'm
of
the
opinion
that
we
shouldn't
be
adding
good
first
issue
or
help
wanted
to
anything
unless
it's
been
triaged
or
if
we're
like,
really
sure,
because
issues
being
created
might
not
actually
be
things
that
we
want
worked
on
before
discussion
happens.
D
D
Yeah,
I'm
of
the
opinion
that
help
wanted
and
good
first
issue
should
be
added
until
it's
been
agreed
on
that
we
actually
want
to
work
on
an
issue
and
that
discussion
has
been
concluded.
A
Yeah,
I
can
say,
for
I
guess,
for
what
it's
worth,
I
think
it's
probably
usually
fine,
for
we
might
want
to
have
an
additional
label
for
like
this.
This
needs
investigation.
A
On
things
that
are
like
a
little
bit
more
obvious,
they
don't
need
as
much
discussion
initially,
sometimes
those
turn
into
things
that
you
need
occasionally
by
accident,
but
I
think
that's
fine.
I
know
I
think,
in
the
context
of
more
complicated
things,
getting
agreement
is
probably
good
who.
How
was
good
first
issue?
Were
you?
It
was
this
just
prompted
by
this
particular
issue
or.
D
I
just
noticed
it
on
a
few
things
that
I
would
like
to
discuss
before
we
actually
start
working
on
them.
I
I
just
don't
want
to
have
people
from
the
community
starting
to
work
on
stuff
and
then
we're
like
wait.
We
actually
don't
want
this,
because
it's
not
a
great
experience
for
them.
I
don't
think
that
happens
very
often,
so
maybe
my
fear
is
unfounded.
D
I
just
like
to
default
to
not
adding
those
unless
I
know
they're
very
it's
a
very
simple
issue
to
resolve
and
that
it's
pretty
well
agreed
on
the
the
solution.
A
That's
that's
fair.
A
I
I
guess
I
think
the
main
question
is:
is
the
assumption
that
we're
only
doing
triage
during
these
meetings
then
or
is
the
assumption
that,
like
the
the
kind
of
the
people
on
the
approvers
list
or
whatever
are
going
to
be
doing
intermediate
triage
during
the
week
if
they
see
stuff
come
up
and
notice,
it's
interesting
because,
like
that's,
I
think
that's
the
case
where,
like
that
question
comes
up
more
right
like
if,
like
an
issue,
comes
in
and
you
read
through
it
because,
like
it
happens
to
come
in
while
you're
looking
or
something
like
that
or
like
someone
pings
you
about
it
and
it
looks
interesting,
do
you
do
you
say:
okay,
I've
read
through
this,
I'm
gonna,
I'm
gonna
mark
it
as
triaged,
because
someone
pinged
me
and
I
read
through
it
and
in
that
case,
like
maybe
you
add,
help
wanted.
A
D
D
A
No,
so
triage
triage
accepted
means
someone
has
read
through
it.
So
like
the
default
state
of
an
issue
and
there's
a
there's,
a
pro
bot,
we
need
to
turn
on
for
this.
The
default
fit
of
the
issue
is
untriaged
or
needs
triage.
I
think
I
think
it
needs
triage
and
then
once
we
triage
accepted
means
someone
has
marked
this
as
confirmed
that
this
is
an
interesting.
A
Not
it
not
an
interesting
issue
but
like
this
is
an
issue
that
we
like
in
some
capacity,
have
decided
to
like
tackle
or
something
right,
and
then
there's
other
ones,
there's
like
triage
duplicate,
which
indicates
that
it's
cl
we're
going
to
close
it
and
it's
a
duplicate
of
something
else.
Triage
needs
more
information
which
means
kind
of,
as
it
says.
On
this
hand,
we
haven't
confirmed
that
it's
an
issue
we
want
to
deal
with
because
we're
waiting
for
more
information,
etc.
A
So,
in
this
meeting,
we'd
probably
end
up
just
looking
at
triage
needs
triage
and
anything
that
had
triage
accepted
or
triage
needs
more
info
information
or
whatever
we
we
wouldn't
look
at
we
might
follow
up
on,
needs
more
information
to
see.
If
they've
been
updated
but
like
triage
accepted
means
we're
done
with
triage,
basically,
okay
thank.
A
C
A
Yeah
I
I'm
kind
of
in
agreement
with
camilla
here,
I
think,
probably
our
false
po.
I
think
our
false
positive
rate
is
pretty
low.
I
guess
like
for
people
doing
intermediate
triage.
Just
you
know,
exercise
your
best
judgment,
probably
I'd.
Also.
A
I
don't
know
what
the
permissions
on
the
on
the
good
first
issue
and
help
wanted
are,
but
like
I
think
we
probably
would
should
roughly
restrict
those
to
reviewers
and
approvers
in
the
repo-
and
I
think
probably
with
that,
we'll
end
up
with
like
a
pretty
low,
false
positive
rate,
and-
and
you
know
if
we
do
end
up
starting
to
get
a
higher
false
positive
rate.
I
think
we
can
revisit
it
and
do
something
like
what
you
suggested
originally
eric
but
like.
A
A
It's
pretty
simple,
the
other
thing
that
I
usually
do
with
good
first
issue.
I
haven't
been
doing
it
as
much
on
this
repo,
but
I
usually
try
to
do
it
in
control
run.
Hyman
controller
tools
is
like
if
you
notice
one
of
the
things.
The
good
first
issue
bot
says
is
make
sure
you
have
a
clear
description
and
or
if
it's
testable,
have
a
clear
description
of
tests.
I
usually
just
try
to
add
a
comment
at
the
very
bottom
or
you
can
do
it
in
here
too.
A
That
just
says,
like
the
thing
you
probably
need
to
do
like
the
place
you
need
to
start.
Looking
is
x.
A
The
things
you'll
need
to
have
at
the
end
are
y,
so
we
you
can
go
back
and
do
that
afterwards
or
we
can
do
it
now,
just
as
an
example,
it's
actually
probably
good
to
have
an
example
on
the
on
the
meaning
recording.
So
why
don't
we
do
that
now,
but
just
usually
just
like
for
this
one.
A
I'd
probably
say
something
like
you
know,
edit
the
scaffold
for
v3
to
replace
the
slash
gcp
line
with
the
slash
offline
linked
above
and
then
regenerate
the
golden
files
and
make
sure
they
look
good.
A
And
that
way
someone
comes
in
and
is
not
just
like
either
like
this
is
my
goodness
first
issue,
but
I
don't
have
the
context
of
where
to
start
and
they're.
Also
not.
We
also
don't
like
end
up
being
like
well.
Why
didn't
you
like?
Do
this
thing?
That's,
like
I
guess,
maybe
obvious
to
people
who
have
been
working
on
the
project
for
a
while,
but
might
not
be
obvious
to
like
a
new
person
like
this
is
like
the
the
confirmation
step
or
whatever.
C
I
have
got
to
choose
one,
I
don't
make
you
know
what
you
hear
enjoy
you
try
to
provide.
You
more
details
feel,
like
you
said,
okay,.
C
Okay,
the
next
one
going
to
be
okay,
this
one,
I
put
it
the
first
issue.
I
remove
okay.
C
C
For
webbing
hooks
should
be
approaching
the
webbing
hooks
plugins,
because,
as
a
developer,
I
can
decide
to
consume
the
kobe
builder
package
and
they
have
chosen
the
image
plugging
into
the
creative
grouping,
creativity
so
or
as
a
developer
like
he
encouraged
a
project,
and
you
don't
use
webby
hooks
at
all,
and
I
think
that
he
better
also
saying
something
like
this
another
day.
That
is
very
confusing
for
the
users
having
so
many
configuration
files,
so
why
we
should
discover
the
files
that
they
are
not
using
actually.
A
A
Keeping
track
of
webhooks
yeah
we
we
because
we
need
to
like
only
do
it
on
the
first
webhook
right
or
probably
only
do
it
on
the
first
webhook
so
like
as
long
as
we
have
a
list
of
like.
Does
this
project
have
webhooks
or
so
as
long
as
we
have
some
way
to
tell
does
this
project
have
webhooks
enabled?
I
think
this
is
probably
fine
for
the
v3
plug-in,
but
do
other
people
have
any
other
thoughts.
C
D
Yeah,
I
think
it's
fine.
I
think
this
also
ties
into
the
larger
discussion
about
how
we're
going
to
pass
in
weapon
conservative
versions,
but
this
is
a
step
in
the
direction
of
using
or
passing
in
that
information
into
the
create
web
hook.
Plugin.
A
I
think
some
of
this
might
also
tie
into.
We
had
some
discussions.
I
think
at
some
point
around
maybe
thinking
about
changing
how
we
do
our
configuration
like
assembly.
A
I
think,
a
few
months
ago
we
had
a
meeting
with
phil,
where
phil
discussed
some
issues
and
potential
solutions.
I
think
we
can
maybe
con
like
that.
Also
kind
of
ties
in
I
don't
know
what
the
timeline
is
for
us.
Thinking
about
that,
but
like
at
some
point,
we
may
also
want
to
re.
Like
start
considering
that.
D
C
Okay,
so
about
to
the
labels,
she's
had
someone
that
she
starts
to
work
on
that
she
has
a
question,
but
she
shows
very
simple
as
well.
Can
we
put
it
good
first
issue
or
help
you
want
it.
D
A
Yeah,
I
usually
save
good
first
issue
for
things
that
are
either
fairly
small
or
fairly
self-contained.
C
C
C
C
D
Yeah
this
is
like
directly
related
to
932
and,
I
would
say,
is
not
well.
I,
like
I
wouldn't
do
anything
with
this
until
932
is
resolved
in
some
way,
because
then
we're
just
adding
like
more
like
optionality
into
the
like
baking,
that
into
the
binary
itself
and
should
be.
This
should
be
done
in
space
two
plugin.
Instead.
A
I
don't
think
it
needs
to
be
a
duplicate.
I
think
it
can
just
be
a
blocked
on
so
to
speak.
A
I
think
if
there's
not
already
a
blocked
on
triage
label,
we
can
request
that
they
get
added
one.
We
can
always
add
our
own
too.
You
should
have
permissions
to
do
that.
I
think
call
like
triage
blocked
or
something
like
that.
C
A
This
is
pretty
straightforward:
we
have
a
collection
of
like
weird
api
server
things
that
people
can
run
into
one
of
the
ones
I've
been
seeing.
A
lot
recently
is
like
upgrading
a
v1
beta,
1
crd
to
a
v1
crd
kind
of
by
accident,
because
that
you
forget
that
you
have
a
v1
beta
1crd
installed
and
then
the
api
server
returns.
Some
really
weird
errors,
but
there's
a
few
things
in
this
category,
and
I
think
we
probably
just
want
to
have
an
faq
section
in
the.
A
C
A
C
I
think
it
is
maybe
it
could
be
good,
good
first
issue
as
well,
because
it's
just
a
document,
but
yes,
maybe
always
beginner.
A
I
didn't
add
good
first
issue
initially
just
because
it
requires
a
little
bit
more
knowledge
of
like
what
the
particular
weird
issues
are
yeah,
and
so
I
figured
it
would
be
slightly
better
for
someone
with
a
little
bit
more
kubernetes
experience
like,
even
if
you
have
to
come
and
ping
me
and
be
like.
What
did
you
mean
here
just
having
a
little
bit
more
background
on
like
what's
going
on,
is
probably
helpful.
C
So
next
one
it
is
one
it's
a
question.
C
So,
let's
see,
I
think,
yeah.
I
need
a
comment.
Okay,
the
user
follow
the
documentation
to
create
the
the
watch.
I
understand
that
you,
the
user,
would
you
like
to
create
a
watch
it
for
a
resource
that
she
is
not
managed
by
the
controller?
C
Maybe
so
you
have
a
better
context
why
we
did?
It
is
implementation
or
engineering
use
the
attributes
because
we
have
a
an
in
the
water,
is
like
his
controller.
I
believe
that
you
should
not.
You
should
do
the
same.
If
you
replace
policy
here,
I
would
be
able
to
watch
resources
that
are
not
managed
by
the
controller.
A
I
mean
sorry,
I'm
gonna
have
to
look
at
this
issue.
I
think
in
more
detail
asynchronously
I
can.
I
can
add
it
to
my
list
of
things
to
look
at
if
you
want.
C
I
I
was
checking
here,
I
think
it's
like,
so
I
put
you
to
react.
You
need
the
information
just
to
hear
my
understanding
as
well
about
his.
B
This
issue
looked
a
little
more
like
a
support
question
than
anything
else,
just
related
to
what
the
person
expect
what
the
person
expected
in
terms
of
like
how
quickly
something
in
the
api
and
how
quickly
something
would
actually
be
deleted.
And
it
looks
like
we're
just
looking
for
like
a
path
forward
or
like.
What's
the
best
approach
to
handling
this
issue,
rather
than
this
being
something
that
is
on
cube,
filter
to
fix
or
introduces
a
feature.
A
C
When
we
teach
the
multi-group
support,
they
we
have
with
the
controllers.
When
we,
when
we
change
the
project
to
support
multi-groups,
we
change
the
layouts,
so
is
no
longer
controllers
and
the
switch
that
she
gonna
file.
So
the
package
in
the
default
is
calls
controllers,
but
when
we
change
the
lg
for
the
multi-group,
where
we
have
the
path
of
the
group,
the
the
file,
the
switch
test
file
has
the
package
contact,
others
as
well,
so
tc
user.
Would
you
like
it
to
just
you?
Have
the
package
name
the
group
name?
C
D
I
mean
I,
I
could
even
see
just
having
the
the
package
name,
be
the
group
package
name
at
all
at
all
times.
I
don't
know
why
it
needs
to
be
static,
but
this
change
works
too.
I'm
fine
with
it.
C
C
Okay,
this
one
is
a
supportive
question,
the
problem
of
the
user-
it
is
it
shows
more
controlling
in
time
than
kobe
builder,
but
they
have
a
lot
of
crgs
and
they
usually
like
to
filter
their
watch
by
a
label
which
is
related.
I
believe
we
have
work
that
she
has
been
doing
the
controller
on
time
as
well.
C
C
C
D
Right
so
this
is
a
controller
runtime
specific
thing,
and
there
is
already
an
issue
open
for
that.
I
don't
know
if
you
looked
like
that,
like
it's
like
244
or
something.
A
Yeah,
I
think
we
can
either
close
this
as
triage
accepted.
A
Oh,
this
is
a
real
old
one:
oh
no
yeah
yeah,
so
I
think
we
can
either
close
this
as
triage
accepted
or
not
traffic
except
trio's
duplicate,
or
I
think
we
have
a
misfiled
label
that
we
use
for
things
that
are
filed
in
the
wrong
repo.
A
But
we
can,
we
can
do
triage,
we
can
do
triage
duplicate
for
now
and-
and
you
know
that
yeah.
D
A
Information,
yeah,
you
you
wouldn't
really
already
they
were.
They
were
asking
about
predicates
I
it's
you
might
want
to
make
it
more
clear
that
we
also
have
where
244
is
working
on
non-predicate
like
not
on
watch
predicates.
A
Sure
yeah,
I
think
I
think
you
I
think
they
were
just
confused
because
you
mentioned
predicates,
which
are
client-side
filtering.
C
B
A
And
they're
currently
just
not
running
because
they
haven't
been
updated.
We
should
either
update
them
or
remove
them.
I
think
we
may
want
to
remove
them
just
because
or
replace
them
with,
like
actual
nightlys
that
run
every
night
or
something
which
I.
B
A
Is
still
useful
but
yeah
someone
just
needs
to
go
in
and
update
the
configuration
they're
currently
using
like
a
weird
hack
ego.
113
hacked
together
build
of
the
releaser
from
a
while
ago,
and
so
we
just
need
to
figure
out
our
new
release
process
and
update
that.
C
A
D
Out
of
curiosity,
why
are
we
moving
away
from
go
releaser
and
I
I
asked
because
I
recently
wrote
a
proposal
for
operator
sdk
that
adds
go
releaser.
So
I'm
just
curious
to
hear
your
opinion.
A
I
don't
think
we
necessarily
need
to
move
away
the
way,
but
at
the
very
least,
for
release
tooling.
We
we
have
like
a
separate
process
for
controller
runtime
and
controller
tools
that
we've
been
using,
and
so
we
wanted
to
unify
and
the
separate
process
produces.
Honestly.
I
think
nice
release
notes
and
I'm
not
just
saying
that,
because
I
wrote
the
other
thing
but
yeah.
So
the
the
separate
process
produces
slightly
nicer
release,
notes
and
so
just
like
having
the
same
release
notes
process
would
be
nice.
A
A
It
like
six
to
one
half
dozen,
the
other
if
we
actually
like,
if
the
releaser
is
still
useful.
For
that,
I
know
at
some
point.
There
were
some
serious
issues
with
it,
which
is
why
we're
using
a
custom
version
that
whose
source
code,
I
think
has
unfortunately
disappeared
somewhere.
But
if
like,
if
it's
working
now-
and
we
just-
and
we
can
tell
it
to
not
generate
release,
notes
for
us
and
use
our
normal
release,
notes
workflow,
then
that's
fine.
A
D
Gotcha
yeah,
thanks
for
clarifying
that
I
from
what
I
understand,
you
can
use
a
custom
change,
log,
that'll,
post
those
release,
notes
in
the
release
and
that's
a
major
reason
why
we're
okay
using
it
in
the
sdk,
but
don't
want
to
hold
up
triage
for
too
much
longer,
because
we
have
two
more
minutes.
Thanks
for
thanks
for
talking
about
that.
A
Let
me
know
eric
if
you
have
any
questions
on
the
release.
Tooling,
I've
been,
I
uploaded
a
whole
bunch
of
stuff
into
the
q
builder,
release
tools,
repo,
including
some
stuff-
that's
just
like
templates
for
use
in
the
individual
repos.
A
C
Okay,
so
next
one
okay,
so
she's
one,
it's
something
that
she
his
contributor
was
speaking
with
me
and
we
decided
to
raise
his
issue
together.
C
Now
we
have,
it
is
related
to
too
much
group
support,
because
when
we
change
the
default
also
has
api
in
the
single
and
when
we
move
to
much
group
support
which
is
getting
apis
in
the
pluto.
So
would
it
be
very
simple
if
you
have
the
same
in
domain
platform
for
both,
I
don't
see
any
reason
for
we
change
the
name
of
the
detector.
C
We
just
simplify
things
like
we
have
the
gt.
Looking
the
gtg
command
that
she
can.
We
could
do
something
smart,
which
is
something
that
it
is
contributor
was
working
on.
He
would
like
that
when
we
run
eg
to
allow
the
multi-group
support,
all
files
would
be
fixed
by
the
two
in
the
new
layout
instead
of
be
required,
doing
manual
steps,
so
things
like.
D
That
so
I
personally
like
well
from
from
what
I
understand
the
the
api
controller
versus
apis
controllers
directories
are
like
the
the
former,
are
a
lot
simpler
and
most,
I
think,
most
use
cases
use
non-multi-group,
so
single
group,
and
so
the
default
project
is
quite
simple
and
pretty
intuitive.
And
then,
when
you
want
to
move
to
multi-group,
you
have
the
wider,
which
is
a
little
bit
more
complex,
but
necessarily
so
because
you
need
more
sub
packages
for
your
different
groups
for
controllers
and
apis.
D
So
I
think
that
the
idea
is
that
you
don't
add
complexity
until
you
need
it
yeah.
That's
why
I
like
the
current
structure.
I
you
are
right,
though
it
does
add
complexity
when
transitioning
to
multi-group,
but
on
the
flip
side
like
maybe
there
should
be
like
known
complexity,
and
you
should
know
exactly
what's
being
changed.
A
Sorry
go
ahead.
Did
I
are
you
off
with
me
in
there?
No,
I'm
done
all
right
so
to
cut
you
to
sorry
to
add
on
to
that
a
little
bit
too,
like
api
versus
apis
is
a
visual
indicator
or
a
word
indicator
nomenclature
indicator
of
what
you
can
expect
the
hierarchy
to
be
right
like
if
you
see
apis,
you
can
expect
it
to
be
apis,
slash
group,
slash
version,
and
if
you
see
api,
you
can
expect
it
to
just
be
api
version.
A
This
this
pattern
is,
is
used
elsewhere,
a
bit
in
kubernetes
too.
Okay,
like
for
people
who
are
used
to
it,
the
differentiator
between
api
and
apis.
Is
it
is
a
nice
quick
like
what
can
I
expect
out
of
this
directory,
which
I
think
is
a
nice
list.
C
C
A
Yeah
so
yeah
maybe
just
leave
a
quick,
a
quick
comment
of
like
discussed
in
the
meeting
or
like.
If
you
want
to
go
back
later
and
leave
a
reason
but
yeah
we
can
just
close
the
yeah
close
it
afterwards.
That
way.
There's
history.
C
Okay,
so
has
a
reason,
so
it's
fine,
okay,
the
other
one
was
like
question:
okay,.
C
How
to
test
external
external
types
or
third
parts
apis?
So
I
think
he
I
answered,
and
you
know
I'm
waiting
for
the
feedback.
A
So
I
think
I
actually
think
this
should
be
marked
as
as
as
a
something
we
should
accept.
I
think
this
is
a.
I
get
asked
this
question
a
lot
too.
I
think
this
is
a
like
indicative
of
a
gap
in
our
book
documentation.
I
think
we
should
probably
just
make
a
like
under
like
the
cookbooks,
like
a
cookbook
section
of
like
short
recipes
or
or
like
under
the
tutorial
section
or
something
we
should
just
have
like
a
short
recipe.
I
can
comment
as
such
too
hold
on.
A
What's
the
issue
number,
we
should
have
a
short
recipe
for
this,
though,
because,
like
I,
I
see
this
all
the
time
and
I
think
like,
if
you
understand
actually
what's
going
on
under
the
hood,
then
this
may
be
kind
of
obvious,
but
if
you're
just
starting
out
or
you're
you're,
not
realizing
that
oh
they're,
all
just
types,
people
people
don't
understand
that
you
can
just
do
this,
so
I
think
we
should.
We
should.
C
A
C
C
We
have
just
in
three
minutes
as
well,
so
I
think
we
can
call
the
day
unless
someone
has
something
to
say.
Would
you
like
to
share
some
idea,
make
some
question.
B
How
should
we
handle
the
remaining
backlog
because
we
only
got
through,
like
a
small
subset
of
all
the
issues
open
right
now,
I
could
take
a
look
at
these
and
asynchronously
like
see
which
ones
I'd
like
to
assign
to
myself
and
then
maybe
in
future
meetings
we
can
bring
up
which
issues
we're
interested
in
and
then
from
there
sort
out
which
milestones
you'd
want
to
consider
putting
them
in
slash
which,
if
we
want
to
actually
address
them
at
all,.
C
D
I
would
bet
that
we,
given
that
we
will
be
doing
our
triaging
during
the
week
in
between
these
meetings.
I
I
would
bet
that
we
will
get
through
all
of
these
currently
open
issues
within
a
few
meetings.
Oh
yeah,
I
think
I
think
the
rate
of
us
triaging
in
these
meetings
is
faster
than
they're
actually
being
created,
including
the
the
rate
at
which
we
actually
deal
with
them
during
the
week.
So
I
say.