►
From YouTube: Carvel Community Meeting - July 27, 2022
Description
Carvel Community Meeting - July 27, 2022
We now meet every 2nd and 4th Wednesday at 9am PT. Come join us live!
This week we went over several new releases to different Carvel tools, recent blog posts, and announced that we have officially submitted Carvel to be considered as a donated Sandbox project to CNCF! Full agenda here: https://hackmd.io/F7g3RT2hR3OcIh-Iznk2hw#July-27-2022-Agenda
A
Hi
everyone
welcome
to
this
week's
edition
of
the
carpool
community
meeting,
where
we
begin
our
new
day
and
time
of
meeting
every
second
and
fourth
Wednesday
at
9
00
a.m.
Pacific
Time
today's
date
is
July
27th.
If
you're
watching
this
from
home,
we'd
encourage
you
to
come,
join
us
live
it's
a
great
opportunity
for
you
to
come
in
and
meet
the
maintainers
meet
other
members
of
the
community.
You
can
come
listen
in
on
what
we're
working
on.
A
You
can
ask
questions
or
if
you
have
a
featured
request
or
or
anything
like
that,
we
would
love
to
have
you
come
join
us
live
and
get
to
know
you
for
those
that
are
on
the
meeting
today
and
you
have
something
you
want
to
discuss
with
the
team
just
be
sure
to
add
that
down
below
in
the
discussion
topic
section
and
we
will
get
to
it
towards
the
end
of
this
meeting.
A
If
you
are
unable
to
join
us
live,
and
you
still
have
questions
or
want
to
hang
out
and
get
to
know
the
community,
you
can
still
find
us
in
the
Carvel
channel
in
the
kubernetes
slack
workspace.
You
can
email
us
in
the
Google
Groups.
You
can
obviously
find
us
in
any
of
our
GitHub
repos
and
on
Twitter
at
Carvel
underscore
Dev.
A
If
you
are
using
the
cardboard
tools,
we
want
to
know
more
about
how
you
are
using
those
tools,
so
we've
created
a
pinned
issue
that
just
requests
that
you
give
us
some
more
details
about
yourself
and
how
you're
using
the
cardboard
tools
in
a
comment
on
the
issue
itself,
optionally.
We
also
ask
that
you
put
your
organization's
logo
and
we
would
get
that
added
to
the
adopters
file,
but
the
details
that
we
request
from
you.
A
It
just
helps
the
team
understand
more
about
how
others
are
using
the
Carvel
tools
and
helps
with
our
like
planning
for
development
and
our
roadmap
and
whatnot.
It
also
helps
the
community
understand
how
folks
are
using
the
tools
and
in
turn
like
gives
them
ideas
of
how
they
could
use
it
as
well.
So,
please
be
sure
to
add
a
comment
there.
We
would
we'd
love
to
know
more
and
those
that
are
on
the
call
today.
A
Just
make
sure
you
add
your
name
to
the
attendees
list,
it
is
optional,
but
we
would
love
to
to
keep
track
of
who's
attending
the
meeting.
Thank
you
on
to
announcements.
So,
since
our
last
meeting
we've
had
a
couple
of
blog
posts,
we
have
Captain
Dagger
by
renu
yarde
and
we
have
cap
Deploy
on
gke
using
keyless
Authentication
oidc
bayash
sethia.
A
So
these
are
a
couple
of
really
good
blog
posts.
I
highly
encourage
you
to
go
check
them
out
re,
give
them
a
read.
If
you
have
any
questions
about
them,
please
be
sure
to
find
us
in
slack
and
and
or
any
feedback
on
them.
We
would.
We
would
love
to
know
more
about
how
you're
how
you
enjoy
reading
these
blogs
that
we're
putting
out
upcoming.
A
We
have
something
by
Joe
Kimmel,
another
maintainer
of
Carvel,
giving
us
a
review
of
Carvel
in
July
and
then,
as
far
as
these
blog
posts
go
and
sharing
content
about
Carvel,
it's
not
limited
to
the
maintainers
to
to
put
out.
We
would
actually
love
to
know
and
hear
more
about
how
others
are
using
Carvel
and
sharing
the
knowledge
of
how
they're
using
carbo,
with
maybe
other
tools
or
or
anything
like
that.
A
A
If
they're
going
to
be
speaking
on
a
workshop
or
anything
like
that,
we
want
to
make
sure
that
we
we
know
about
it
and
we
can
promote
it
in
all
the
different
channels
that
we
have
and
in
turn
we
also
want
to
say
thank
you
by
sending
you
some
Carvel
swag,
we
have
T-shirts
and
and
hoodies
and
beanies
and
all
sorts
of
stuff.
So
you
know
just
we
would
love
to
to
know
more
and
have
others
outside
of
the
maintainers
contribute.
A
So
if
you
have
any
questions
about
that,
just
let
us
know
moving
on
to
some
releases,
we've
had
several
releases
since
we
last
met,
so
we
had
cap
with
o.50
Praveen.
You
want
to
give
us
some
more
context
on
this.
B
B
So
just
release
had
like
a
couple
of
bug:
fixes
one
of
them
was
related
to
the
TTY
flag.
So
by
default,
like
the
default
value
of
dt5
is
actually,
it
has
been
set
to
true
for
deploy
and
delete
commands,
but
and
because
of
that,
it
used
to
cause
some
issues
with
other
commands
as
well.
The
headers
and
photos
would
come
in
the
way
when
people
would
want
to
kind
of
type
the
content
to
a
file,
so
we
have
actually
set
the
default
value
to
false
and
only
for
deploy
and
delete.
B
That's
set
to
true.
Then
we
we
had
an
issue
with
exist
and
Nova
annotation,
so
these
annotations
are
basically
used
to
wait
for
resources
that
are
created
by
some
other
controllers,
and
these
would
not
be
part
of
the
actual
Gap
app
that
we
already,
including
in
the
Manifest.
So
there
would
be
conflicts
when
the
controller
was
cap
controller
in
this
case,
because
cap
controller
would
use
cap
to
deploy
those
resources
and
they
would
already
have
the
cap
labels.
B
B
So
let's
say
there
are
two
changes,
change
again
change
B
and
you
would
want
change
P
to
be
unblocked
as
soon
as
a
certain
condition
is
restraint,
change
a,
but
you
would
still
want
Capital
reconcile
change,
a
complete
detail,
another
condition,
so
you
could
use,
unblock
changes
for
a
particular
condition
and
then
all
the
changes
depending
on
that
resource
would
be
unblocked.
Next
stop
is
diff
change
of
sample.
B
This
is
flag
that
we
recently
added
and
with
this
we
users
could
basically
get
the
final
vaml
that
Gap
is
going
to
be
applying,
including
all
the
rebase
rules
are
exist,
operation
or
delete
operation
that
is
going
to
be
performed.
We
have
also
added
default
change
groups
and
change
rules
for
cap
controller
resources.
B
That
is
app
cos
package,
installs
and
service
accounts,
so
service
accounts
would
now
be
created
before
apps,
yours
and
packet
installs,
and
would
be
deleted
after
cap
controller
and
pack
after
app
CR
and
package
installs,
and
we
we
got
our
first
contribution
from
I.
Could
I
can
only
read
the
handle
right
now?
Victor
yeah
thanks
Victor.
A
Thank
you
so
much
Praveen
on
on
giving
us
some
more
contacts
around
that
and
thanks
Victor
for
for
your
contribution.
Any
questions
on
this
release
before
move
on
to
the
next
one.
C
Real
quick
one
Praveen
the
diff
changes
yaml
does
that
that
shows
the
total
yaml
that's
going
to
be
applied
as
opposed
to
other
output.
That
shows
the
Delta.
Is
this
showing
the
whole
thing?
Yeah
yep,
okay,
yep.
B
A
E
Right
so
I'll
do
it
just
for
you
cool
okay,
so
the
moment
came
for
the
Glorious
number
30
of
image
release
of
image
package.
So
here
we
go.
Let's
all
Rejoice
with
these
awesome
changes
that
happened
so
in
this
in
this
particular
release.
What
happened
so
we
had
two
ux
improvements
where
we
we
basically
give
more
information
to
our
users
of
what's
happening
with
image
package.
So
right
after,
like
a
lot
of
you
might
not
know,
but
right
after
we
push
all
the
images
go
to
the
registry.
E
What
we
do
is
we
tag
them,
but
in
the
end,
what
what
the
user
saw?
It
was
like
just
okay,
all
the
images
are
pushed
and
then
nothing
and
then
oh,
we
finished.
Now
we
have
more
information,
we're
telling
people
what
we're
doing
so
that
that's
cool
and
we
now
provide
more
information
about.
So
if
you
are
copying
non-distributable
layers
around
so
usually
with
those
layers
you
and
there's
like
there
was
a
warning
saying
like
there
are
non-distributable
layers
that
we
didn't
copy.
E
Now
it
tells
you
which
layers
they
are
and
from
what
from
which
image
so
that
that's
pretty
cool
and
two
bug
fixes
for
this
release,
one
of
them
thanks
for
being
for
it.
It's
now
our
bash
completion
now
works.
So
thank
you
for
that
and
the
other
one
is
something
that
we've
been
asked
for,
and
that
is
how
can
I
create
a
bundle
in
Windows
that
works
in
a
Linux
machine
right,
and
that
was
failing.
That
was
not
working,
and
now
it
is
working.
E
So
that's,
awesome
and
I
think
that's
it
I
don't
know.
Does
anyone
have
any
question
about
the
release.
A
C
Yeah
I
can
I
can
swing
this
a
little
bit.
I
know
that
this
was
a
request.
This
has
to
do
with
when
you
have
a
git
source,
and
you
really
just
want
the
latest
that
it
could
actually
be
expensive
to
either
pull
or
update
that
local
copy
of
the
repo.
But
now
you
can
manage
that
with
the
death
depth
configuration.
A
A
If
there
was
an
annotation,
no
okay,
all
right
secret,
gen
controller.
F
Can
try,
although
I
didn't
I
I,
think
it's
it's.
This
is
really
incremental
release,
so
it's
it's
mostly
the
release
process
itself
that
Neil
the
the
Neil
improved.
So
the
does
not
very
much
that's
exciting
from
a
user
perspective,
unless
you're,
an
arm64
user,
in
which
case
now
suddenly
you're
very
excited
because
before
there
was
nothing
for
you
and
now
there's
a
release
for
you.
F
Right
on
right
on-
and
you
know,
I
guess
the
other-
maybe
the
other
thing
where-
and
this
is
Neil's
work,
but
he
in
some
sense
secret,
gen
controller.
It
now
kind
of
shows
what
does
the
Carvel
packaging
team
think
a
generic
package
might
look
like.
So
it's
it's
a
little
bit.
It's
it's
an
attempt
at
making.
You
know
an
archetypal
package.
You
know
we're
really
thought
leaders
in
this
space
Neil
is
the
voice
of
a
generation
so
that,
but
that's
anyways
I
you,
you
can
be
inspired
by
his
work.
A
I
love
it
and
it
looks
like
we
had
some
new
contributors,
Glenn
Rogers.
F
C
Kind
of
rounding
out
the
celebratory
releases
here
you
might
think
that
this
version
gives
you
the
answer
to
life
universe
and
everything.
It's
just
shy
of
that.
We're
really
excited
about
two
huge
developments
here.
One
is
an
entire
new
module
if
you're
doing
any
floating,
Point
work,
which
is
not
very
common,
but
it
comes
up
from
time
to
time.
You
now
have
lots
of
the
basic
math
functions
that
are
available
to
you.
Do
note.
If
you
use
any
of
that
stuff,
floating
Point
math
can
vary
across
architectures.
C
So
buyer,
beware
with
that
one
you'll
get
a
a
proper
warning
if
you
do
do
that,
but
but
it's
still
available
to
you,
the
other
big
big
thing
that
is
coming
out
with
this
release
is
on
top
of
the
basic
Plumbing
that
we
have
in
place
for
ytt
data
values,
validation.
We
now
have
what
we
believe
is
that
core
set
of
out
of
the
box
validation
rules.
C
So
if
you
haven't
had
a
chance
to
kick
the
tires
on
this
still
experimental
feature
that
first
item
there
is
a
blog
post
that
we
put
out
that
walks
you
through.
Why
are
we
doing
this?
How
to
get
started
with
it
and
then
the
other
is
a
direct
link
into
the
docs
to
see
what
are
those
rules
that
are
available
now
note.
The
window
is
closing
in
terms
of
being
able
to
put
your
fingerprint
on
the
overall
experience
and
and
even
some
functionality
on
this
feature,
we're
interested
in
your
feedback.
C
That's
why
we're
putting
this
out
in
an
experimental
mode.
We
want
to
not
commit
to
the
API
at
first.
Give
you
a
chance
to
have
an
influence
on
that,
so
taking
advantage
of
that,
we're
going
to
want
want
to
do
some
validation
on
that
and
basically
firm
up
that
API
very
soon.
So,
if
you
have
any
opinions
about
that,
we'd
love
to
hear
them
big
or
small-
that's
not
it!
There
were
some
bug
fixes
as
well.
C
In
this
release,
the
first
one
was
where
there
were
a
number
of
cases
where
a
when
a
function
was
being
pulled
out
of
another
library
or
used
in
a
different
context.
It
would
actually
Panic
when
it
was
trying
to
report
some
kind
of
syntax
error.
That's
no
longer
the
case,
we're
able
to
basically
say
hey.
C
We
may
not
be
able
to
provide
the
exact
line
that
caused
this
error,
but
we
can
at
least
report
a
same
error
message
and
give
you
much
more
useful,
much
more
actionable
output
in
those
cases
and
then
Victor
swung
by
ytt
as
well
and
dropped
in
his
contribution
on
better
air
handling.
Whenever
we
have
hdprs
we'll
do
that,
when
we
you
can
source
files
into
ytt
through
the
dash
dash
files
flag,
you
can
supply
a
URL
there
so
yeah.
C
So
this
was
actually
like
a
number
of
contributions
from
folks
outside
the
team,
thanks
to
folks
who
actually
put
not
only
ported
over
the
starlark
math
module,
but
actually
adapted
it
to
a
number
of
ways
in
which
we're
implementing
those
modules
appreciate
that
Lars
dropped
in
a
really
cool
example
on
how
to
use
ytt
to
enhance
his
usage
of
a
home
chart,
it's
kind
of
cool,
it's
both
the
front
end
where
there's
validations
on
their
data
values
and
then
using
the
post
render
hook
to
do
some
overlays.
C
So
that's
a
nice
example
of
how
to
augment
your
using
Carvel
tools
to
to
improve
your
existing
tool
set.
Yeah
and
like
I
mentioned
Victor
thanks
again.
C
Cool
I
wanted
to
get
with
you
Scott
in
in
and
push
push
harder
on
some
of
the
edge
cases
and
see
how
that
works
for
you
so
I'm
coming
for
you
awesome,
I'm,
glad
I'm
glad
you
had
a
chance
to
kick
the
tires.
A
Sweet
okay,
moving
on
to
huge
announcement,
big
announcement,
reason
why
we
are
here:
we've
submitted
to
the
cloud
native
Computing,
Foundation
cncf
consideration
to
be
a
donated
sandbox
project,
so
we
submitted
two
weeks
ago
and
the
TOC
has
to
work
through
several
other
ones
that
have
have
submitted
ahead
of
us.
So
we
hope
to
hear
their
decision
with
the
next
few
months
on
whether
or
not
we
are
officially
a
Sandbox
project
with
cncf.
A
So
super
huge
big
news
where
we're
really
excited
for
this
opportunity-
and
you
know
it
just-
helps
us-
become
more
of
a
vendor.
Neutral
project,
hopefully
encourages
more
of
y'all
to
come
and
help
contribute
and
participate
in
in
the
development
process
of
Carvel
and
all
that
fun
stuff
so
really
excited.
We
hope
you
all
are
excited
too,
as
we
as
we
continue
to
evolve,
Carvel
and
and
get
more
involved
in
in
this,
the
cloud
native
community,
so
yeah.
So
big
news.
C
Yeah
this
is
this
is
really
great
Nancy,
like
one
of
the
things
that
comes
up
a
lot
when
we're
talking
about
like
project
Direction
and
things
like
that,
and
what
we're
focusing
on
it's
it's
round
the
table
that
we
want
to
make
sure
that
we're
participating
in
the
community
that
we're
contributing
back
and
that
we're
pushing
the
whole
ecosystem
forward.
So
this
is
exciting,
move
in
that
direction,
make
that
a
structural
commitment
and
not
just
aspirational
so
yeah.
This
is
great.
A
Yeah
yeah
so
stay
tuned
for
for
when
we
officially
become
sandbox
project.
Okay.
Moving
on
to
Aaron,
you
have
a
announcement
here
as
well:
contributor
model
working
group,
yeah.
H
This
is
just
a
bit
of
a
open
call,
but
our
Our
intention
here
is
to
make
the
path
to
maintainership
more
visible,
more
discoverable,
and
although
this
effort,
this
working
group
is
being
driven
by
maintainers,
it
would
be
great
to
get
input
from
the
community
as
well.
To
make
sure
you
know
it's.
It's
clearly
understood
how
you
become
a
maintainer
with
the
different
levels
of
being
a
maintainer
would
be
the
roles
responsibilities
all
that
good
stuff.
H
D
H
Sure
we
made
a
light
refreshing
of
the
road
map
a
lot
of
the
themes
there
are.
The
couple
themes
here
continue
to
be
the
same
package,
author
improvements
and
providing
stability,
as
John
mentioned,
the
the
first
one,
the
schema
validations
experimental,
that's
right
on
track,
Late
July!
So
please
play
around
with
that
and
give
us
feedback.
We
have
a
ga,
that's
planned
for
the
end
of
next
month,
based
upon
the
feedback
we
receive.
H
Let's
see,
then
we
have
a
couple
of
kept
controller
items
in
here
again
on
the
providing
stability
theme,
but
some
work
that
the
dentist
has
been
working
on
is
surfacing
name
spaces
for
deployed
resources,
so
this
was
a
a
request
that
we
have
heard
from
from
multiple
folks
just
trying
to
get
more
information
regarding
well
kind
of
like
what
it
says
regarding
the
namespaces
of
of
what
they
have
deployed
but
feel
free
to
click
into
there
and
learn
more
there's
another
track
of
work.
H
That's
that's
going
on
around
version
safeguarding
it's.
It's
essentially
a
collection
of
different
stories
that
all
have
the
word
version
in
it
to
some
degree
whether
it
is
providing
safeguards
for
you
know,
versions
that
are
being
used,
whether
that
is
the
cluster
kubernetes
version
or
yeah
I
guess,
like
I
know,
either
either
kubernetes
versions,
cluster
versions,
ways
to
either
protect
yourself
or
find
more
information.
Regarding
what
is
deployed.
I,
don't
know
Joe.
If
you
want
to
add
things
to
that.
H
Perfect,
thank
you
and
then
rainy.
You
want
to
jump
in
on
the
K
control.
I
Yep,
so
for
key
control,
we've
been
working
on
it
since
past
two
months
and
trying
to
now
we
are
at
the
stage
where
we
are
ready
to
come
out
with
Alpha
release.
I
think
mid-august
is
when
we
are
planning
it.
We
will
try
out
a
few
things
with
some
of
the
projects
that
we
have
thought.
That
would
be
good
examples
for
us
to
try
it
out
and
then
so.
I
Those
examples
would
also
make
it
into
the
take
control
repo
and
that
should
give
us
a
it's
not
so
request
you
all
to
give
us
feedback
when
we
use.
If
you
want
to
partner
and
try
out
the
tool,
let
me
know
we
would
definitely
love
to
partner
and
see
if
we
could
package
something
together.
I
Cap
I
have
not
put
in
here,
but
we
are
now
working
on
the
next
Milestone
I.
Think
Praveen
will
talk
about
what
we
are
working
on.
However,
one
thing
that
I
would
like
to
see
is,
and
try
out
is
if
there
are
something
that
the
community
would
want
to
bump
up
and
some
of
the
issues
that
they
would
want
us
to
see
working
more
on.
So
just
calling
out
to
the
computer
to
see
if
you
have
something
that
you
want
us
to
work
on,
yeah.
H
Cool
and
then
the
last
item
on
here
for
image
package
well
image
package
is
probably
the
incorrect
label
here,
but
the
the
general
idea
is
minimizing
registry
Communications
to
provide
a
better
user
experience,
especially
when
using
a
large
package
repositories.
E
Just
that
this
is
not
it
image
package.
This
is
Cap
controller
right,
but
yeah.
So
we've
been
we've
been
hearing.
People
complaining
about
like
cap
controller
is
starving
some
of
the
Registries,
with
a
lot
of
connections
always
trying
to
connect
always
trying
to
download
the
images
from
them.
So
this
is
an
effort
for
us
to
try
to
minimize
this
as
much
as
much
as
possible
so
that
we
can.
E
H
You
cool
yeah,
that's
everything
as
always,
if
you
have
feedback
for
us
regarding
this
roadmap
or
would
like
to
discuss
what's
coming
next,
this
is
an
ongoing
planning
activity
and
we
are
we're
working
on.
You
know
what
is
coming
next,
so
so,
please
do
not
hesitate
to
reach
out.
E
Oh
look
at
me
again,
so
basically,
we've
been
keep
kept
on
working
on
our
resume
flag,
so
we
are
adding
a
flag
to
to
our
copy
command.
When
you're
copying
it
to
a
tar,
you
can
provide
now
flag
called
resume
that
will
resume
the
download
for
your
for
your
bundles
when
you're
copying
them
to
a
tar.
So
like
there
are
some
complaints.
People
saying
like
oh
I,
have
a
very
big
bundle
when
I'm
copying
it.
For
some
reason
it
fails.
E
I
have
to
restart
everything,
so
this
little
flag
will
allow
you
to
start
from
where
you
were
and
do
not
have
to
re-download
all
the
images
so
we're
trying
to
make
it
better
for
our
users.
So
that's
that
what
we
are
it's
in
PR
now
so
I
am
assuming
that
somewhere
in
the
next
couple
of
days,
we'll
be
able
to
merge
that
just
to
let
y'all
know
like.
E
The
second
point
is
that
we
kind
of
stopped
a
little
bit
on
the
work
that
we
were
doing
for
the
is
authentication
that
we
were
doing
that
I
talked
about
a
couple
of
weeks
ago.
Eventually,
we
might
start
this,
but
there's
like
other
competing
priorities
as
well
related
to
the
the
roadmap
that
we
just
saw
so
we'll
see
where
things
fall
in
there
and
I
think
for
image
packages.
That's
it
cap
now.
G
E
B
B
Exciting
we
have
been
looking
for
some
time
now
and
your
kind
of
debating
about
how
the
output
and
the
flag
name
should
be
so
right
now,
whenever
we
are
applying
or
waiting
for
resources,
if
any
of
the
resource
errors
out,
we
kind
of
quickly
error
out
and
don't
apply
the
rest
of
the
resources,
but
with
this
change
all
the
resources
like
as
much
as
possible
in
the
sense
that
if
there
is
a
failing
resource,
then
the
resource
that
are
dependent
on
it
should
not
be
applied.
B
But
otherwise
cap
will
try
to
apply
all
the
resources,
wait
for
them
and
then
show
the
errors
at
the
end.
So
yeah,
as
I
mentioned,
we
are
mostly
debating
about
how
the
error
summary
should
look
at
the
end
and
what
should
be
the
flag
name
to
kind
of
disable.
This
feature
next
up
is
so
right.
B
Now,
if
we
don't,
if
cap
does
not
have
the
permission
to
list
namespaces,
then
it
won't
be
able
to
like
figure
out
what
figure
out
the
existing
resources
present
on
the
cluster
right
now
we
just
use
the
flat
the
namespace
provided
by
hyphen
and
flag
and
look
for
the
resources
in
that
namespace,
but
we
are
planning
to
use
a
namespaces
that
we
store
inside
the
cap
config
map,
while
deploying
the
resource
so
in
subsequent
deploys.
C
B
F
A
A
So
yeah,
if,
if
you
have
any
questions
on
it
or
comments,
be
sure
to
take
a
look
at
those
hyperlinks
to
the
pr
and
issues
or
if
you're
you
get
a
little
itch
to
help
contribute,
we
would
love
for
you
to
participate
in
that
way
as
well
all
right.
Moving
on
to
ytt.
C
Yeah,
so
we're
going
to
be
spending
the
bulk
of
our
time
right
now,
we're
preparing
to
be
able
to
do
usability
testing,
so
this
is
about
setting
up
situations
where
we
can
get
feedback
on
usability
takes
a
little
bit
of
work
to
do
that.
Look
out
for
us
soliciting
for
your
participation
in
something
like
this.
C
It
should
be
we're
going
to
make
it
so
it's
really
cheap
and
easy
to
do
if
you
want
to
just
pop
on
a
video
call
with
us,
15
minutes
or
so
something
like
this
just
to
be
able
to
try
something
out
to
get
some
feedback
from
you
on
how
it's
working
it
doesn't
even
have
to
be
a
formal
session,
though
we're
in
this
mode
right
now
of
trying
to
gather
feedback
about
pretty
significant
feature
and
we'd
love
to
hear
from
you,
because
what
we
want
to
do
is
provide
something
that
you
find
ergonomic,
that
it's
intuitive
it's
what
you
need.
C
It
works
the
way
you
want
it
to
work,
so
we're
we're
interested
in
any
kind
of
feedback
on
this
big
or
small,
on
the
ytt
validations.
So
that's
our
Focus
this
week
is
getting
all
that
prepared
and
starting
to
do
some
of
those
interviews.
A
A
There
hasn't
been
anything
put
into
this
discussion
topic
section.
Is
anything
popped
up
that
folks
want
to
discuss.
We
have
just
a
little
under
20
minutes
left.
F
D
Putting
on
a
spot,
but
do
you
have
any
kind
of
any
recent
conversations
with
your
customers
or
anyone
or
your
users,
which
might
be
interesting
for
as
it
relates
to
Carvel
it's
okay,
if
you
can
say
no,
but
that's
what
I'm
gonna
give
you
the
stage.
G
I
not
much
I
mean
I'm
using
it
more
and
more
my
customers,
a
lot
of
them-
are
definitely
interested
in
it
and
see
the
value
of
certain
parts
of
it,
the
main
one
that
I
was
never
using,
but
it
started
to
use
as
Vendor
by
a
lot
of
customers
and
they're.
It's
resonating
and
it's
definitely
helpful
very.
F
G
Helping
is
them
being
able
to
pull
in
from
multiple
different
sources
into
a
very
structured
repository
where
a
lot
of
them
are
bringing
in
things
from
upstream
and
may
have
their
changes
that
they're
saving
in
a
git
repo
and
an
easy
way
for
them
to
start
working
with
things
is
to
pull
down
from
us,
mostly
git,
but
they're,
also
using
sometimes
pulling
down
image
package.
G
So
a
lot
of
times
we're
using
that
as
just
an
easy
way
to
automatically
just
bring
everything
down,
so
they
can
start
working.
I've
also
used
it
in
certain
environments
to
help
them
set
up
bootstrap
machines
so
pulling
down
binaries
from
30
different
places,
pulling
down
scripts
pulling
down
whatever
it
is.
It
becomes
very
easy
to
maintain
that
way.
They
can
run
a
single
command
basically
and
get
all
of
the
dependencies
they
need
on
their
machine
to
start
working
in
the
environment,
something
that
before
was
a
very
tedious
task.
F
It
sounds
almost
like.
We
should
not
tell
your
customers
about
Docker
files,
the
hot
new
way
from
2014
to
download
all
the
things
that
you
need
for
a
development
environment
and-
and
you
know,
get
get
up
and
running
with
your
with
your
local
Dev.
F
G
So
you
may
be
shocked,
but
there
are
many
environments
out
there
where
Docker
on
your
desktop
is
not
allowed
due
to
company
restrictions
and
the
fun
of
working
on,
for
example,
vdi
machines
that
you
have
no
admin
access
to.
So
can't
install
Docker,
desktop
or
they're
non-persistent
things,
so
you
install
Docker
desktop
and
you
need
to
reboot
the
VM
for
it
to
work
and
yeah.
It's
gone,
there's
a
lot
of
fun
around
that
by
customers.
G
So
in
those
cases
things
like
vendor
are
solutions
that
make
it
easier
in
these,
let's
say:
restricted
environments
where
you
can't
just
spin
up
a
Docker
container
exec
into
it,
have
whatever
you
need
just
create
these
kind
of
immutable
use
of
get
pod
code
spaces
whatever
it
is
as
a
developer
experience.
G
G
For
sure,
and
no
and
we're
doing
a
lot
of
packaging
these
days
and
that's
always
fun.
G
G
I
may
happen
to
have
984
packages
in
a
package
repository.
Currently,
it
may
not
behave
the
best
possible.
D
D
To
sink
yeah.
F
G
F
G
F
Yeah
I
I,
we
would
we
would
love
to
hear
about
it
because
I
know
like
we.
We
did
some
benchmarking.
We
did
some
tuning,
but
you
know
benchmark
especially
this
Benchmark
was
intentionally
synthetic
like
it
was
like
every
package
was
just
empty
brackets,
so
you
know
I
think
we
we'd
love
to
hear
about
a
real
world
friction
and
where,
where
it,
how
it's
Landing
for
you?
Oh.
G
Mine
is
10,
ver
mine
is
15
versions
of
every
chart
in
the
bitnami
album
chart
Repository
or
the
commercial
version
of
the
bitnami
Helen
chart
repository
and
convert
it
into
using
image
package
and
K
build
and
whatnot
to
give
you
everything
and
schemas
filled
into
the
everything.
So
it
takes
some
time
that
are
quite
large.
G
G
A
Sweet
Scott
and
any
anything
else.
C
Yeah
Nancy,
really
quick,
Praveen
and
I
had
a
little
bit
of
an
open
thread
there
in
his
update
and
the
internet.
Wasn't
cooperating.
C
I'd
asked
about
that
point
in
cap
where
we're
working
on
this
feature
to
be
able
to
apply
as
much
as
we
possibly
can
and
collect
errors
and
Report.
Those
and
Praveen
was
asking
they're
planning
on
making
it
the
right
default
behavior
and
provide
an
option
to
opt
out
of
it
so
that
it
would
be
the
default
to
do
that.
I
was
wondering
what
should
that
behavior
be
so
I
have
I
have
an
opinion,
but
it's
it's
not
exactly
the
most
informed
in
it.
C
C
Maybe
that's
useful
to
be
able
to
run
it
in
that
state,
but
if
more
often
than
not
it's
better
just
to
try
and
deploy
as
much
as
possible
and
if
there
are
a
few
things,
a
few
components
that
aren't
going
to
make
it
that's,
that's
fine
actually
would
make
it
easier
to
learn
all
the
things
that
aren't
quite
working.
Maybe
that's
a
better
experience
but
I
feel
like
I'm,
not
terribly
qualified,
to
give
a
good,
authoritative
answer
to
that
I
wonder
if
others
on
the
call
might
have
an
opinion.
G
G
For
that
or
having
even
the
ability
within
the
cap
config
to
be
able
to
set,
not
only
is
this
enabled
or
not,
but
the
idea
would
be
to
be
able
to
say
certain
types
of
objects
that
fail.
You
should
continue
anyways.
There
may
be
certain
types
of
failures
that
may
happen
in
certain
cases
that
we
want
to
say:
okay,
continue
if
this
happens,
but
if
I
enter
a
crash
loop
back
off
of
my
container,
there
is
no
reason
for
me
to
continue
to
playing
a
service
in
an
Ingress
and
whatever
other
objects.
G
There
are
because
it's
not
going
to
expose
anything,
but
if
it's
some
auxiliary
object,
let's
say
maybe
I
do
want
to
continue
in
just
that
one
object
won't
be
created,
so
I
think
having
the
ability
to
set
which
objects
or
what
types
of
failures
you
would
accept
within
a
cap,
config
to
just
say
kind
of
hey
proceed.
Anyways
would
be
kind
of
the
middle
ground
there.
That
gives
on
one
hand
the
better
quick
feedback
to
someone
and,
on
the
other
hand,
allowed
for
a
bit
more
customization.
B
D
B
B
G
Yeah
I
think
that's
less
I
think
that's
a
nice
to
have
right.
That's
a
nice
enhancement
after
this
I
think
having
the
flag
on
its
own
is
a
huge
value
and
I
think
would
bring
huge
usability
Improvement.
So
definitely
a
huge
fan
of
that
I
think
that
the
adding
to
cap
config
would
just
be
a
nice
additional
feature.
I
guess
you
could
say
in
the
same
vein
down
the
road
as
a
extend
extension
of
this.
A
Okay
thanks
everyone
for
joining
us
for
our
new
day
and
time,
and
if
you're
watching
this
from
home,
we
encourage
you
to
come,
join
us
for
the
next
one.
We
meet
every
second
and
fourth
Wednesday
of
the
month
at
9am,
Pacific
time
so
with
that
enjoy
the
rest
of
your
week
and
we
hope
to
see
you
soon.