►
From YouTube: Carvel Community Meeting - October 14, 2021
Description
Carvel Community Meeting - October 14, 2021
We meet every Thursday at 10:30am PT. We'd love for you to join us live!
The team went over the latest release from the Carvel tool suite and had lengthy discussion on speeding up docs PRs, as well as other topics. Full notes here: https://hackmd.io/F7g3RT2hR3OcIh-Iznk2hw?both#October-14-2021-Agenda
A
A
If
you
are
watching
this
as
a
recording
on
youtube,
we
invite
you
to
come
and
join
us
live
we
meet
every
thursday
at
10,
30
a.m.
Pacific
time
it's
an
opportunity
for
you
to
come
and
listen
in
on
what
the
maintainers
are
working
on.
What
they're
thinking
about
and
present-
and
you
can
you
yourself-
can
present
any
ideas
that
you
want
the
maintainers
to
work
on
regarding
the
carvel
tool
suite.
A
If
you're,
not
a
member
of
that
slack
workspace,
you
can
request
an
invitation
to
gain
access
and
we
are
also
found
on
twitter
at
carnival
underscore
dev
when
you
are
interacting
with
us,
whether
it
be
in
the
meeting
in
slack
on
on
twitter
on
github,
we
require
that
you,
please
read
and
abide
by
our
code
of
conduct
all
right
next
up.
If
you
are
attending
the
meeting
live.
Please
put
your
name
in
any
organization
that
you
are
representing
down
here
in
the
agenda.
A
This
helps
us
keep
track
of
those
that
are
attending
that
are
outside
of
the
maintainer
circle,
and
we
can
keep
the
lines
of
communication
open
if
you're,
using
those
tools
for
the
carvel
tools.
We
also
invite
you
to
input
the
details
about
how
you
use
karvel
into
this
pinned
issue.
That's
on
the
carnival
repo
and
you
can
put
your
logo
for
your
organization
on
there
as
well,
and
it
helps
us
learn
more
about
the
way
that
others
are
using.
The
tools,
as
well
as
help
promote
your
organization
as
well.
A
So,
if
you
haven't
registered
to
attend
coupon
yet
and
want
to
attend
virtually,
there
is
still
time
if
you're
attending
in
person,
we
encourage
you
to
go
and
drop
by
the
the
last
day
of
keynotes
to
check
out
this
five
minute
keynote
covering
carville
shadaroopa
is
going
to
be
doing
a
short
demo
and
once
you
you
view
that
feel
free
to
find
us
back
on
slack
and
give
us
any
questions
that
you
might
have
after
watching.
It.
A
What
you'll
need
to
do
is
keep
an
eye
out
for
the
the
oktoberfest
topic
labeled
on
our
repos
right
now.
We
just
have
it
on
image
package,
repo
and
specifically
labeled
hacktoberfest
labeled
issues,
and,
if
so,
if
you're
participating
in
oktoberfest
and
want
to
help
out
the
karvel
team
go
to
image
package,
repo
check
out
these
issues
and
help
us
out,
and
you
can
also
earn
a
free
t-shirt
once
you
hit
four
pull
requests.
B
Right
one
note:
there:
cap
also
has
some
issues
so
I'll
drop,
a
link
there.
A
I've
got
some
releases
to
cover
here.
We
got
ytt
0.37.
C
This
one's
an
exciting
one,
we've
got
a
couple
of
improvements
here.
First
of
all,
thanks
to
cobra,
we've
got
command
completion
like
shell,
shell
command.
Completion
kind
of
just
comes
for
free,
almost
there's
some
docs
that
we
want
to
do
around
that
to
like
make
that
a
little
bit
easier,
that's
forthcoming,
but
the
command
is
there
is
present.
C
The
other
thing
is
we
got
a
contribution
for
a
new
module
in
our
standard
library
for
parsing
ip
addresses
and
cider
blocks.
So
that's
all
there
to
go.
It's
documented
go
check
that
out
super
cool,
appreciate
the
contribution
and
then,
as
alongside
that
that
same
group
implemented
very
small
but
rather
powerful
new
feature
in
the
assert
modules.
C
This
try
to
feature
which
allows
you
to
attempt
to
do
something
so
like
parson,
json
and
you're,
not
sure
whether
or
not
it's
going
to
succeed,
and
you
don't
want
it
to
be
a
failure.
If
it
fails,
you
don't
want
it
to
end
your
ytt
execution.
You
just
want
to
learn
about
that.
Oh,
there
is
an
error,
and
maybe
you
try
something
else
differently
now
you
can
do
that
with
try
to
it's
great
for
helping
out.
If
you
want
to
write
tests
where
you
have
functions
or
return
errors,
also,
there's
documentation
there
about
that.
C
Yeah,
this
one's
also
contributor
led
as
well
as
great
we've,
got
a
a
whole
new
builder
in
cable.
Now
right,
because
k
build,
helps
you
build
container
images
and
resolve
them
to
digests,
and
you
plug
in
many
of
your
favorite
tools
to
do
the
actual
container
image
building.
Now
you
can
use
bazel,
which
is
a
really
well
known,
commonly
used
general
purpose
build
tool.
The
integration
is
there
so
go
check
that
out.
C
We
do
have
a
breaking
change
in
the
the
body
of
an
annotation
with
this
release,
where
we
used
to
refer
to
something
called
metas
and
now
we're
using
a
term
which
we
feel
is
much
more
intention
revealing
origins
where
the
information
contained
within
the
origin
annotation
tells
you
from.
Where
did
this
container
image
come
from?
So
if
it's
built
from
source,
you
get
to
see
the
get
repo.
C
That's
the
the
data's
always
been
present,
git
repo,
whether
or
not
your
directory
was
dirty
when
you
did
the
build
or
from
other
sources
that
information's
always
been
there,
we're
just
now
referring
to
them
instead
of
metas,
as
origins
so
check
that
out
and
then
finally,
like
a
really
important
piece
is
actually
plumbing
that
very
data
all
the
way
through
into
your
image
package,
lock
files.
C
So
now,
when
you
are
using
that
full
carbon
flow
from
build
to
deploy,
including
cabled
in
your
in
that
flow,
those
annotations
will
get
attached
to
your
resources
as
they
go
into
the
cluster.
And
so
now
you
can
inspect
your
resources
and
see
your
origins,
information
there.
D
I
can
talk
about
this
one
yeah,
so
basically
there's
a
new
flag
added
debug,
it's
a
global
flag,
so
this
is
useful.
If
you
know
you
get
a
generic
error
back
from
image
package
and
you
want
to
know
like
a
bit
more
details
and
what
went
wrong
because
image
package
deals
with
multiple
registries.
D
It
basically
tells
you
the
flow
between
image
package
and
each
of
the
registries
you're
interacting
with
and
and
it
helps
you
narrow
down.
Potentially,
you
know
an
issue
in
that
flow
between
image
package
and
a
specific
registry.
So
it
shows
all
the
http
course
being
made
it
redacts
the
sensitive
information,
thanks
for
nicholas,
stimulate
for
adding
that
for
folk
using
you
know,
image
package
as
a
library,
we've
bumped
golang
to
1.17
and
also
there's
a
bug
fix.
D
So
if
you're
using
windows
and
using
image
package
on
windows,
it
now
loads,
the
roots
d8
store.
So
if
you're
using
tls,
you
know
https
on
a
registry
and
it
was
signed
by
a
trusted
authority,
then
it
now
is
able
to
verify
that
certificate
and
those
docs
in
terms
of
if
you
want
to
also
use
your
own
custom
ca
and
also
be
able
to
load
the
root.
Ca
store
refer
to
the
docs
and
how
to
do
that
for
windows,
specifically.
A
All
right,
it
doesn't
look
like
there's
anyone
here
to
go
over
cap's
stuff.
D
E
F
B
That's
the
one
sorry
I
was
gonna
add.
I
think
that
they've
been
working
on
the
cluster
fulfilled
resources,
epic.
B
And
yeah,
it's
I
think
it's
just
in
flight.
I
don't.
I
don't
have
tons
of
insight,
but
that's
something
that
they've
been
focusing
on
for
their
next
release.
A
Well,
thanks
aaron,
and
then
I
was
just
gonna
say
for
those
watching
at
home.
The
cap
team
is
based
in
india,
so
this
time
is
a
little
bit
difficult
for
them
to
attend,
but
sometimes
we
do
have
one
or
two
of
them
on
here
to
help
answer
any
questions
about
cap
that
you
may
have.
E
E
E
I
also
expect
that
we'll
have
a
few
more
stories
added
to
ytt
around
open
api
in
the
near
future.
So
if
you're
curious
about
features
on
the
upcoming
keep
an
eye
out
for
that.
A
Okay,
carnival
backlog.
I
think
this
is
left
over
from
previous
meeting
unless
I'm.
A
Is
there
anything
folks
want
to
go
over
particular
with
cat
control
or
a
vendor
that
they're
working
on.
H
Well,
I'm
you
want
to
how
much
you
want
to
bet.
We
were
going
to
say
different
things
or
that
anyways.
I
was
going
to
call
out
that
we
were
even
one
step
closer
to
moving
our
image
repository
to
the
the
github
container
registry
and
that
that
pull
request
is,
you
know,
poised
to
be
merged
in
the
near
future
and
then
we'll
we'll
start
cutting
releases
that
are
stored
on
github
container
registry.
C
H
No,
I
don't
that's
a
good
question.
We
probably
should
move
them
consistently.
I
know
like
after
we
move
cap
controller.
The
next
thing
is
to
move
secret
gen
controller.
I
would
imagine
we'll
just
move
everything
to
be
consistent.
I
think
that's
part
of
the
vision
is
to
get
rid
of
the
docker
hub
account
management
eventually,
but
that's
not
necessarily.
G
I'll
just
add
into
that
neil
has
or
worked
on
a
quick
story
for
basically
removing
proxy
credentials
that
were
showing
up
in
some
cap
controller
logs.
So
we
had
a
small
issue
with
just
basically
that
we
would
actually
like
log
the
proxy
in
and
of
itself,
and
that
basically
did
not
account
for
the
fact
that
you
know
there
can
be
credentials
actually
as
part
of
that
proxy
format
that
were
included
in
those
logs.
So
there's
been
a
quick
pr.
That's
been
put
in
to
just
basically
remove
that
information.
A
B
Yeah,
this
topic
came
up
from
some
conversation.
I
had
with
some
people
that
might
be
here,
but
we
talked
about
how
it
can
be
challenging
to
get
docs
prs
to
land,
probably
due
to
a
number
of
things.
So
maybe
maybe
it's
worth
elaborating
on
what?
What
are
the
problems
that
we
are
facing?
Maybe
before
we
try
to
solve
it?
B
C
H
I
think
stuff
can
linger
right,
so
it's
like
you
can
put
like,
because
we
don't
have
deadlines.
You
could
say
so
right
like
if
you
go.
Look
at
the
carvill
website
right
now,
there's
like
quite
a
number
of
open
pr's.
Some
of
them
are,
are
pretty
old.
Some
of
them
are
from
this
week
and
the
in
in
none
of
those
cases.
Is
it
really
clear
like
right?
H
You
can't
just
say:
oh
well,
the
unit
tests
are
passing
and
it's
been
up
for
48
hours,
so
I'm
going
to
merge
it
and
in
fact
who
is
responsible
for
for
merging
them.
H
G
I
think
that
sometimes
I
wonder
if
like
more
tightly
coupling
like
docks
to
like
an
actual
feature
or
some
type
of
fix,
like
the
actual
story
itself
and
like
the
issue
is
not
actually
closed
until
like,
like
the
docks,
should,
in
essence
like
close
the
pr
depending
on
like
in
what
order
they
happen
right
like
if
a
if
a
story
has
docks,
then
the
the
issue
itself
is
not
closed
until
you
submit
a
docs
pr
and
it
like
actually
closes
out
that
issue.
G
So
in
some
ways,
I
I
think
like
putting
like
an
actual
like
reason
to
like
write
the
docs
themselves
is
I
at
least
think
like
that's
like
a
sometimes
a
pain
point
from
that
cap
controller
standpoint
is
there,
there
hasn't
always
been
like,
I
think,
like
sometimes
like
docs
can
get
lost,
sometimes
because
they're,
not
always
like
included
as
part
of
stories.
D
Yeah,
that's
that's
been
my
experience
with
an
image
package
which
is
you
know
the
last
two
things
I
talked
about
the
debug
flag,
the
windows
bug
each
of
them
as
part
of
the
stories
working
on
them.
We
deemed
it
necessary
to
add
documentation,
so
the
actual
code
change
pr
was
approved
by
image
package
approver.
D
Similarly,
when
we
raised
a
pr
for
the
carville
documentation
and
kavel
repo,
the
person
who
approved
it
was
also
the
you
know,
image
package
approver.
So
that's
how
the
model
that
we,
that
you
know
we
used
and
it
seemed
to
work
well,
meaning
that
it
got
merged
before
the
release.
So
we
could
link
the
release,
notes
to
the
documentation.
C
So
I've
heard
kind
of
two
positions
on
that
one
is
put
it
in
draft
and
then,
when
you're
ready
to
go,
put
it
in
draft
so
that
it
doesn't
get
merged
accidentally
ahead
of
time
and
then,
when
you're
ready
to
go,
pull
it
out
of
draft
and
have
it
merged
like
we
have
that
be
part
of
the
release
process.
So
there's
some
kind
of
approval
before
that
moment
and
then
the
other
is
well.
If
you
put
a
version
caveat
in
the
docs,
then
you
can
merge
at
any
time.
C
And
then
to
complicate
things,
we
actually
had
a
proposal
to
talk
about
a
way
of
using
versioned
docs
went
through
a
whole
process,
approved
the
proposal,
but
there's
work
to
to
be
done
to
like
make
that
a
thing,
and
we
haven't
done
that
work.
So
there's
sort
of
some
ambiguity
around.
How
do
we
handle
versions.
F
F
B
B
Don't
yet
have
I
guess
any
sort
of
like
I
guess
does.
Does
anyone
have
a
mitigative
step
or
action
here
that
we
can
consider.
B
C
Agree,
I
mean
I
have
a
snarky
response,
which
is
what,
if
we
could
get
a
dock
rider
in
here.
Like
you
know,
a
professional.
A
I've
been
told
that
that
that's
kind
of
in
the
works-
and
I
think
the
dark
writer
would
be
under
my
team,
like
under
jonas's
team,
because
it
seems
like
a
lot
of
the
dark
writers
are
on
the
cet
team.
So
I
think
it's
dap
budget,
but
under
us
or
some
something
weird
like
that.
But
I
know
that
he
has
talked
to
pulkit
and
others,
and
I
think
there
is
approval.
B
Cool
check
and
done
cool
well
for
now.
I
think,
there's
there's
some
level
of
acceptance.
Writing
docs
are
challenging,
but
there
is
a
next
step
here
is.
Could
we
could
we
find
someone
who's
more
dedicated,
more
experienced
here,
so
the
next
one
I
had?
I
think
it's
along
the
lines
of
kind
of
one.
So
there's
like
two
ones
here.
One
is
a
pr
I
guess
done
and
then
the
other
one
is
when
it's
done.
When
do
we
merge
them?
So
maybe
we'll
start
with.
When
is
a
pr
done.
H
You
know
all
voice,
I
agree
with
it
enough
dimitri
who's,
not
here,
and
I
I
agree
with
it
enough
to
repeat
it
said.
Basically,
even
bad
docs
are
better
than
missing
docs
and
so
the
the
argument
is
like
you
know,
we
should
merge
quickly
and
then
go
ahead
and
revise,
but
that
that
we
might
be
lingering
too
long
in
the
like.
H
Oh
you
know,
let
me
make
sure
like
neil
did
you
want
me
to
change
that
from
to
a
two
you
know
or
I'm
just
I'm
picking
on
neil,
but.
B
B
Is
there
anything
further
that
we
want
to
do
like?
Actually,
I
don't
know
have
like
a
policy
that
we
want
to
enforce
with
ourselves.
Like
you
know,
if
I
don't
hear
feedback
within
x
amount
of
hours
days,
I
can
just
move
forward
or.
C
C
So
I'd
like
to
hold
back,
because
I
want
other
certain
people
to
like
put
put
their
fingerprint
on
it
in
some
way
or
their
perspective
on
it.
Then
then,
like
cool,
more
power
to
you,
but
maybe
it's
just
like
why.
Why
not
just
leave
it
sort
of
wide
open?
It's
a
website
that
we
all
edit.
H
I
think
you
know,
I
think
the
one
like
wrinkle
there
could
be
there's
a
big
difference
between
I'm
adding
a
dock,
I'm
correcting
a
bug
in
a
dock
versus
I've
decided
to
re-architect
all
of
our
sidebars
right.
H
So
it's
maybe
there's
an
argument
to
be
made
that
refactors,
especially
like
sweeping
ones,
probably
warrant
more
discussion
than
I'm
at
I'm
documenting.
My
recent
work.
B
Okay,
maybe
that
helps
for
the
the
when
it's
done
another
one
then
or
the
next
one
here
is:
who
is
responsible
for
merging
them?.
B
Cool
and
then,
as
for
the
the
other
one,
when
do
we
merge
them?
I
think
the
couple
things
were
mentioned
using
the
draft
to
signal
a
pr
is
not
yet
ready
to
merge.
I
think
that's
a
great
strategy.
The
version
doc
proposal
is
probably
like
the
real
long-term
solution.
It's
not
yet
prioritized,
but
is.
Is
there
anything
else
here
that
folks
want
to
share.
C
Yeah
I
I
kind
of
dug
what
neil
was
sharing
that
like
you
could.
Even
just
I
mean
especially
if
you're
putting
inversion
caveats
around
stuff,
then
it's
that
helps
it
from
being
in
some
way
ambiguous
and
so
then
there's
no
timing.
And
so
then
you
just
do
it
it's
easier
to
not
miss,
because
if
you
write
it
and
you're
like
hey,
that's
good
enough,
it's
got
the
virtual
cowbetch
merchant
you're
not
trying
to
remember
to
do
that
later.
F
A
Awesome
and
I
got
a
quick
update
on
the
doxwriter
wreck-
it
will
most
likely
be
happening
in
the
new
year.
So
don't
hold
your
breath.
C
A
A
H
I
can
talk
through
that
real
quick.
Actually,
do
you
mind
if
I,
if
I
share
a
screen,
so
that
I
think
that's
the
pull
request
link
is
in
the
minutes,
and
I
wanted
to
not
in
great
detail
but
briefly
like
at
a
table
of
contents
level,
the
the
contents
of
this
document.
It
is
it's
under
this
shared
docs.
I
honestly,
I
don't
know
how
to
navigate
to
this
page
from
our
actual
home
page,
but
I
do
know
how
to
type
shared
docs
latest.
H
Maybe
we
need
to
link
anyways
the
this.
I
everyone
on
the
team
has
been
through
at
least
one
discussion
of
this
previously
and
but
you
know,
if
someone
is
actually
watching
this
recording
first
of
all,
hi,
I'm
surprised,
but,
second
of
all,
maybe
you
care
what's
what's
on
here,
that
you
know,
there's
information
about
how
to
name
things,
how
to
structure
the
code
and
like
we
have
some
some
opinions
about
when
to
be
dry
and
when
not
to
be
a
little
bit
of
planning.
There's
in
some
ways.
H
Maybe
the
the
juiciest
section
is
like
we
have
a
few
specific.
You
know
opinions
about
how
to
write,
go
and
then
there's
there's
some,
hopefully
helpful
instructions
on
you
know,
testing,
branching,
releasing
etc,
and
so
I
I
am
gonna
leave
this
up.
H
Even
despite
our
discussion,
I'm
gonna
leave
this
as
a
pull
request
for
another
day
or
two
and
if
I
think,
there's
a
few
helpful
comments
on
it
earlier
in
the
week
and
maybe
I'll
I'll
plan
on
merging
it
on
monday
or
tuesday
and
next
week,
unless
there's
you
know,
someone
finds
some
issues
but
anyways
open
for
comment
for
a
little
bit
longer
and
then
subject
to
anyone
changing
it.
H
So
that's
thanks.
I
don't
know
I'm
happy
to
take
comments
now,
but.
A
I
didn't
even
know
that
existed
so
definitely
need
to
find
a
way
to
make
it
easier
to
navigate
to
on
the
website.
B
B
Cool
well,
if
anyone
has
ideas
for
how
to
how
to
make
this
more
easy
to
discover
a
pr
is
a
good
place
to
start
that
discussion.
B
I
have
I,
I
think,
the
last
question
which
today
we've
had
some
really
great
discussions.
At
times
I
felt
that
we've
hit
the
discussion
part
of
this
meeting
and
haven't
had
time
or
haven't,
had
things
that
we
want
to
chat
about.
A
quick
search
of
our
backlog
tells
me
of
our
cumulative
backlog.
Tells
me
we
have
more
than
120
issues
with
the
discussion
label.
What,
if
we
used
this
time
to
start
slowly
going
through
those
and
like
having
the
discussion
to
clarify
what
we
want?
Some
of
these
issues
to
be.
H
I
guess
like
right,
it's
always
possible
that
we
discussed
something
about
well
cap
controller
and
like
neil
and
daniel,
and
I
have
one
level
of
access
of
knowledge
right
and
then
I
think
that's
that's
going
to
be
true
for
any
tool
is
there's
going
to
be
one
to
three
people
at
the
meeting
who
have
kind
of
specialist
knowledge
of
that
tool
because
they're
in
that
code
base
regularly.
H
So
I
guess
I'm
curious
what
the
expectation
would
be.
Would
it
be
like,
and
today
we're
we're
just
talking
about
ytt
issues
you
can
drop.
If
that's
not
going
to
be
helpful
for
you
or
would
it
be
like
you
know
today
we're
going
to
do
one
from
ytt
one
from
cap
controller
and
the
point
is,
you
know,
mesh
a
little.
B
Yeah,
that's
a
good
point.
I
I
guess
I
just
added
a
note
here
that
we'd
want
to
ensure
the
folks
in
the
meeting.
Are
the
quote
unquote
right
folks
for
the
discussion-
and
I
do
agree,
probably
like
blocking
them
together
by
some
sort
of
domain
or
some
interest
would
probably
be
more
effective
than
just
like
a
grab
bag
of
things.
So
so
yeah,
maybe
maybe
there's
a
little
more
preparation
here.
That's
needed
to
make
sure
that
we
invite
you
know
any
any
of
the
right
people,
but
something
something
to
consider.
D
Yeah
I'd
be
willing
to
try
it
out,
but
I'd
be
skeptical
like
mainly
because
I'm
wondering
if
we
have
all
the
right
stakeholders
as
well,
so
a
lot
of
these
issues,
at
least
when
which
package
discussions
came
from
outside
folk.
You
know
who
are
using
the
tool
and
it
might
be
two
or
three
outside
folk.
B
Yeah
that
that's
a
great
call
out,
maybe
that's
the
carrot.
They
need
to
attend
these
meetings.
B
Cool
cool,
wait,
I
don't
think
there's
anything
to
do
now,
it's
more
of
something
something
to
keep
in
mind
as
as
we
have
these
ongoing
discussions
and
if
we
want
to
try
and
nudge
them
forward.
This
is
a
great
time
to
do
so.
Synchronously.
A
Okay,
anything
else.
The
team
wishes
to
discuss.
A
Nope,
okay!
Well,
thank
you
everyone
for
joining
today
and
if
you're
watching
this
from
home,
I
would
love
to
have
you
join
us
live
and
with
that
enjoy
the
rest
of
your
week.
Thank
you.