►
From YouTube: Kubernetes Community Meeting 20170510
Description
We have PUBLIC and RECORDED weekly video meetings every Thursday at 10am US Pacific Time.
https://docs.google.com/document/d/1VQDIAB0OqiSjIHI8AWMvSdceWhnz56jNpZrLs6o7NJY
OpenCompose Demo; Releases; SIG Apps, SIG BigData
A
B
B
A
B
A
E
E
So
since
Chow
our
branch
manager
to
castles
spend
their
time
to
fix
the
build
and
have
and
also
cuddly
needs-
and
please
help
us
to
pass
of
our
families
and-
and
we
plan
to
cut
to
alpha
for
for
the
1.7
our
next
Wednesday.
So
this
is
our
schedule
what
we
are
so
so,
if
you
have
some
feature
you
want
to
be
includes
editing
past
as
early
as
possible.
So
please
make
sure
your
merge
before
that
time
and
that's
the
kind
of
schedule
time
is
mix
with
it.
So
this
is
related
to
the
current
patent.
E
You
need
another
one.
It
is
one
of
the
pick
away
from
1.6
retrospective.
It
is
so
we
should
just.
We
should
have
site
site
the
appetizer
suites
as
early
as
possible,
so
we
release
in
the
request.
So
we
already
have
the
order
upgrade,
has
the
site
help
or
that,
for
example,
up
which
has
the
for
one
point
file
to
master
one
point,
six
to
master
and
all
those
kind,
and
also
done
through
the
front
of
master
to
the
previous
treaties.
So
the
past
is
available.
All
that
has
suite
is
available.
E
You
can
take
look
in
the
advertiser
grade,
so
there's
the
upgrade
to
master
school.
Has
this
week.
People
can
take
a
look
at
those
things
and
also
our
upgrade
the
tested
our
lead.
Our
Jeff
is
the
stat
monitor
those
tests
swing.
So
kindly
we
are
talking
about
things
we
already
have
those
available
and
we
want
to
capture
those
problems
as
early
as
possible
and
also
resolved.
So
so
we
thought
we
are
started
talking
about.
E
Maybe
we
put
those
up
with,
has
some
of
chooses
them
up
put
in
curl
up
with
pasture
to
our
P,
a
broker
PR
build
broker.
So,
but
this
is
animals?
Yes,
so
if
we've
done
as
we
with
an
email
to
the
community
so
a
matter,
it
ends
so
evil
or
remains
the
feature
manager.
Send
the
odd
announcement
to
the
spreadsheet.
For
the
one
point
famous
feature,
tracking
board
for
the
component
is
the
one
handsome
increase
we
expect
here.
E
So
we
want
to
start
a
dog
hiking
and
also
a
dark
process
as
early
as
possible.
So
this
is
another
thing
we
want.
Your
the
community
help
us
move
forward
felony.
It
is
because
the
I
personal
also
been
doing
bad
with
so
I
was
want
to
mention
that,
because
we
used
to
have
the
experience
engineer,
send
the
implementation
like
be
for
the
code
of
phrase,
feature
phrase
and
also
the
newer
canonical
everyone's
busy
and
a
reviewers
with
the
velocity
review
or
the
PS.
So
I
just
want
to
also
personal.
E
So
we
want
to
the
community,
have
asked
you
to
to
help
to
make
this
process
so
that
they
got
a
feature
development
and
a
review
process
more
healthy.
So
we
can
start
the
development
as
early
as
possible
and
also
attract
as
early
as
possible
and
a
review
as
early
as
possible.
So
we
can
also
even
passed
earlier.
So
that's
all
I
need
correct.
Do
you
want
to
fill
something
more.
E
A
E
A
G
All
right
so
say
God
says
honest.
One
year
old,
we
had
our
first
meeting
last
year
on
May
18th.
Since
then,
we've
had
42
meetings,
two
of
them
have
been
been
in
person
and
we
have
a
42
technical
gonna
happen
for
now.
This
big
has
been
a
place
to
talk
about
developing
and
managing
kubernetes
applications
for
applications
for
kubernetes.
We
also
make
this
a
place
to
provide
feedback
on
the
apps
experience.
This
is
also
a
friendly
place.
G
To
start
is
to
bring
the
day's
journey
and
a
lot
of
what
we
talk
about
is
a
higher
level
tools
on
top
of
coupons
that
help
people
with
the
ax
experience,
and
we
also
keep
up
with
and
champion
to
brands,
were
closed
features
and
improvements.
So
this
is
a
very
small
subset
of
those
features
and
improvements
that
the
whole
list
is
very
lengthy
and
details
and
I'm
happy
to
provide
a
link
to
that.
If
anybody
is
interested.
G
Currently,
there
is
an
open
TR
for
switching
deployments
to
a
new
hashing
algorithm
to
avoid
collisions,
there's
also
an
open
PR
for
moving
cron
jobs
to
beta
and
there's
also
an
emphasis
on
fixing
flaky
tests.
So
yeah
also
houses
several
on
brawler
projects,
so
there
are
seven
or
eight
of
them,
but
I'm
just
going
to
run
through,
like
five
of
them
today
pretty
quickly.
G
G
So
we've
started
pairing
with
new
contributors
who
have
you
know
somewhat
bigger,
pull
requests
to
ensure
they
have
everything
they
needed
or
been.
Excuse
me
to
make
sure
their
pull
request
is
successful.
G
We're
also
pushing
our
CI
to
the
limit,
so
we're
wondering
if
we
should
like
move
off
circle,
CI
and
kind
of
figuring
out
what
we
should
do
there.
That's
an
ongoing
project
and
we're
also
figuring
out
how
our
protein
should
process
like
how
you
add,
a
core
team
member
with,
like
the
official
process
for
removing
someone
who's
been
inactive,
we're
also
working
to
like
better
at
documentation
as
well,
because
all
the
documentation
live
in
the
same
repository
as
code
and
so
we're
working
on.
G
Creating
a
website
for
that
and
making
that
more
user
friendly
and
nice
to
interact
with
oh
I
didn't
have
time
to
mention
all
the
technical
details
for
the
last
few
releases
of
home.
But
I
did
want
to
call
out
home
plugins
because
we'll
see
a
little
pattern
in
the
next
slide,
so
home
plugins
allow
us
to
extend
the
functionality
of
home
for
personal
use
cases
or
like
for
the
use
cases
of
a
particular
organization,
and
recently
Adam
riess
of
the
home
team,
led
an
effort
to
create
a
home
plug-in
management
system.
G
So
now
you
can
like
manage
and
share
common
plugins
more
easily.
So
that's
been
really
cool
to
compose
is
a
tool
for
moving
from
docker
compose
to
kubernetes.
It's
also
a
project
that
we
house
it's
formally
in
the
incubator
process,
making
monthly
releases
and
it's
close
to
graduation
apparatus
tree,
which
initially
started
out
as
a
spec
for
pulling
pushing
and
discovering
bearnaise
apps
turn
into
what
it
was
previously
called
CNR
and
turn
into
add
registry.
You
might
have
heard
Brandon
Brandon
talk
about
it
at
qualitative
kind.
G
So
now
you
can
officially
push
poll
and
discover
kubernetes
apps
to
play
using
a
home
plugin,
but
you
can
use
to
login
play
push
charts,
display
and
sell
charts
and
play,
and
this
effort
was
largely
led
by
Jimmy
szalinski
and
his
team
from
core
lessons.
So
that's
been
really
cool
to
see
they're,
also
working
on
a
Quay
UI
integration
and
a
composed
integration.
G
The
official
kou
branch
arts,
repo,
is
also
growing
really
fast.
So
there
are
currently
70
stable,
charts,
9
incubator,
charts
in
over
70
open
pull
requests
for
new
or
improved
shirts.
So
clearly,
this
group
is
also
facing
scaling
challenges.
They've
added
several
new
chart
maintainer
to
their
team
to
help
with
their
pull
requests.
They
also
have
grand
plans
for
CI
and
testing
and,
if
you
missed
the
last
big
apps,
meaning
Vickie
Iglesias
gave
a
great
presentation
on,
like
common
chart,
patterns
and
dust
chart,
breath
packs
using
best
practices
for
charts.
G
Lastly,
on
the
charts
depository,
if
you
haven't
seen
to
box,
calm
and
you're
interested
in
like
search
and
discovery
of
the
official,
stable
charts,
then
check
that
out.
It
allows
you
to
easily
discover
and
download
charts
and
it
was
built
using
an
open
source
project
called
monocular.
So
if
you
have
your
own
private
repository
and
have
been
thinking
about
creating
like
a
like
a
UI
for
it
check
out
monocular,
you
don't
have
to
do
that
by
yourself
and
you
definitely
don't
have
to
do
it
from
scratch.
G
App
controller
allows
you
to
is
another
project
and
it
allows
you
to
like
install
resources
in
a
particular
order
or
with
some
sort
of
like
carrot,
child
hierarchy.
So,
basically,
when
you
define
your
kubernetes
resources,
you
can
you
know,
put
predefined
this
like
order,
and
hierarchy
and
it'll
create
a
resource
deployment
graph,
which
is
then
you
can
then
it'll
like
execute
on
that,
so
they're
also
progressing
you're,
creating
a
helm,
integration,
they're,
adding
more
examples
and
improving
documentation.
G
They've.
Also
that
team
watch
a
leads
that
effort.
He
also
created
the
ability
for
home
to
have
like
server-side
backends,
pluggable
backends,
so
that's
how
they're
actually
implementing
their
appcontroller
innovation.
Therefore,
our
questions
on
that
there
is
a
demo
in
QA
section
at
session
at
sea.
I'll
find
a
20
second
open
compose
we'll
share
about
today.
So
I
won't
go
into
too
much
detail
on
that,
so
we're
going
to
continue
doing
all
the
things
that
we've
done.
G
There's
also
an
emphasis
on
writing
west
camels,
so
different
people
in
the
group
with
different
organizations
if
there
are
different
efforts
being
made
to
kind
of
like
solve
the
same
problem.
So
I've
listed
a
few
of
them
here
today
and
there's
a
design
like
talking
about
and
comparing
these
different
efforts
in
the
sig
and
seeing
if
there's
any
way,
we
can
consolidate
efforts.
G
Logistically
we're
still
doing
weekly
meetings.
We
alternate
between
two
different
meaning
formats,
so
one
week
we'll
do
a
general
in
stand-up
and
then
the
next
week
we'll
do
a
gentle
in
a
discussion
topic.
Our
meetings
are
slotted
for
an
hour,
but
usually
they
end
a
little
bit
early,
so
they're
about
45
to
50
minutes
each.
G
As
you
can
see,
the
scope
of
the
sig
is
very
large
and
we've
talked
several
times
about
splitting
this
thing,
but
it
never
made
sense
because
of
all
the
cross
projects
and
cross
team
collaboration.
That's
going
on
so
instead
of
splitting
the
sig
I
want
to
mention
that
we
added
a
new
sig
asshoie.
His
name
is
Adnan
abul
hassein.
He
is
an
anchor
engineers
evening
at
fit.
Nami
he's
a
chart.
G
Repository
maintainer
in
need,
he
has
lots
of
experience,
designing
and
deploying
different
types
of
applications
on
kubernetes,
so
Matt,
Serena
and
I
are
very
excited
to
welcome
him
and
his
expertise
to
our
team
and
going
forward
we'll
just
iterate
on
what
we
have
more
discussions
around
best
practices
and
patterns
with
different
types
of
people
who
come
to
this
save.
So
it
could
be.
You
know,
in
depth.
Like
you
know,
in
the
weeds
two
grants
developers,
it
could
be
people
who
are
building
the
higher
level
tools.
G
It
can
be
someone
who's
deployed,
their
first
app
on
kubernetes
and
so
setting
really
good
context.
Setting
very
clear
context
on
around
the
demos
and
discussions
have
become
increases
increasingly
important,
so
working
on
making
sure
we
have
good
solid
feedback
loops
and
more
detailed
notes
and
making
sure
that
our
meeting
formats
are
satisfactory
for
the
overwhelming
majority
of
the
participants
of
our
sake
and
indefinitely
have
today,
I
think
that
might
be
on
my
team
but
I'm
happy
to
take
questions
offline.
If
anybody
needs
oh.
C
Samsung,
ok,
hi
there,
so
I
think
two
questions,
one
was
I'll,
keep
them
both
up
and
you
can
see
how
much
of
them
we
want
to
cover.
One
was
a
question
about
circle:
CI,
lows,
work,
but
but
I
think
there's
been
some
rumbling
about
a
CI
Singh
starting
out,
but
that
seems
to
be
pretty
directly
strongly
connected
to
what
you're
working
so
so
that'd
be
one
thing
to
comment
on
and
the
other
is
I'd
say
the
word
operator
missing
from
your
deck
is
pretty
notable,
and
you
know
I
think
this
is
miss.
C
I
think
that
operators
as
far
as
I'm,
seeing
it
kind
of
come
in
two
flavors
one,
is
like
operator
to
use
something:
that's
kind
of
fundamental
to
kubernetes
itself
and
then
the
other
is
there.
Are
operators
popping
up
for
doing
more
like
application,
kinds
of
things
from
databases
and
other
things
and
I
was
wondering?
I
was
wondering
if
your
group
is
working
specifically
an
operators
or
not
yeah.
G
C
Operators
right
did
you
look
at
there's
many
applicants.
This
is,
and
it
was
just
curious
that
the
presentation
talked
about
helm
and
helm,
charts
and
didn't
have
anything
about
the
I
guess.
Can
we
call
it
a
kubernetes
operator,
but
anyway,
I
know
it's
a
new
evolving
space,
but
I
think
it's
one
that
has
some
interest
and
I
was
curious.
If
your
stay
is
what
yes,
let.
G
Me
let
me
address
that,
so
that's
a
really
good
question.
First
of
all,
so
I
know
there
has
not
been
a
lot
of
emphasis
on
the
operator
story
in
this
particular
sake.
This
far
only
because
nobody
has
started
up
yet
we
have
a
list
of
like
demos
that
we
want
to
request.
So
it's
just
been
on
ur
to-do
list
to
request
an
operator
demo
from
you
know
someone
from
core
OS
or
someone
who's.
You
know
running
the
STD
operator.
G
There
is
a
home
chart
for
the
FDD
operator,
I
know
for
sure,
but
I
don't
know
if
there's
more
concerts
for
other
types
of
operators,
we
should
have
that's
a
very
good
point
and
we
should
be
talking
a
little
bit
more
about
operators
in
our
in
our
sig.
So
thank
you
for
that.
On
the
CI
note,
the
dive
wasn't
aware
of
this
API,
but
I
can
be
definitely
tell
tell
the
person
who's
working
on
circle
CI
from
the
home
team
about
it,
so
they
can
go
and
check
that
out.
G
C
That
just
to
be
clear
about
my
question,
I
think
I've
heard
some
discussions
about
whether
there
should
be
a
CI,
sinned
or
not,
and
since
you've
been
since
you're
talking
about
maybe
like
aware
and
how
to
split
the
opposite,
because
I
agree
there's
an
awful
lot
there.
That
would
be
worth
it
to
throw
that
into
your
list
of
considerations.
C
G
Yes
gotcha,
so
then,
then
my
question
would
be
like
what
is
this
say,
focusing
on
I,
don't
like
CI
in
kubernetes
or
the
CI
for
commands
into
Bernie's
related
projects.
I
think.
C
C
F
Cool
so
cig
big
data
we're
fairly
new
and
there
used
to
be
a
different
iteration
of
this.
That
I
think
gave
it
off
and
hammer
started,
but
we
revived
this
on
the
1st
of
February,
which
is
like
three
or
four
months
ago.
You
had
floating
14
meetings
so
far
and
lots
of
interest
in
it
started
off
kind
of
backwards.
F
So
we
do
have
some
overlap
with
cig,
apps
and
scheduling
because
we
do
batch
computing,
which
kind
of
falls
in
the
scheduling
space.
So
our
goal
broadly
is
to
focus
on
deploy
now
operating
big
data
applications
and
mostly
integrating
them
into
kubernetes.
Well
enough
that
people
feel
like
ESB,
you
know,
kubernetes
is
a
great
place
to
run,
not
just
your
stateless
workloads,
but
everything
so
far.
We've
done
some
work
on
SPARC
and
HDFS.
Most
of
this
stuff
happens
outside
and
external
repository
store
issues
which
map
to
it
in
the
main
repo,
but
mostly
its
outside.
F
F
F
So
far,
yeah
like
I,
said
externally.
Both
we
have
about
ten
contributors,
lots
of
different
companies.
We
have
interest
from
cloudera
collaborators,
building
a
third
body
resource
based
operator
for
some
of
their
stuffs
for
experimenting
with
it.
Data
breeches
involved
because
they're
the
upstream
part
community,
which
is
helping
us
get
this
upstream.
Our
efforts
on
spark
are
probably
going
to
go
into
Apache
spark
itself.
F
We
have
Google
that
is
us
and
I've
an
high
profile
with
lots
of
their
companies
and
we're
still
growing
this
list
yeah,
and
we
have
two
upcoming
talks
at
sparks
on
it,
which
is
going
to
be
fun.
We
have
one
introducing
spark
on
coronaries
natively
and
one
more,
which
is
going
to
be
HDFS,
how
to
run
it
different
benchmarks
and
just
progress
on
that
front.
The
issues
I'd
say
that
we're
facing
at
least
ones
that
I
see.
F
F
E
F
A
B
Thank
you
for
a
patient
with
after
let's
2017,
so
we
have
to
fight
Weiland
and
XR
so
anyway,
oh
so,
we
wanted
to
quickly
show
you
what
we
have
achieved.
In
sum
up,
whenever
we
are
going
on
it's
like
a
sort
of
a
side
project
where
we
are
working
on
it,
we
call
it
open
compose.
We
have
discussed
about
this
at
the
last
two
cube
cons
and
shift
conversion
and
cube
con
Seattle.
B
Also.
We
have
present
at
once
and
to
see
gaps
meeting.
So
let
me
quickly
tell
you
what
it
is.
So
we
all
know
the
cube
analysis
is
a
very
powerful
tool
and
it
is,
it
is
powerful
and
it
is
complex
and
complexity
is
because
of
its
power,
but
we
have
seen
internally
and
externally
lots
of
people
lots
of
developers
just
want
to
deploy
their
applications.
We
can
think
in
terms
of
machine
learning,
developer.
Who's.
Writing
is
my.
He
is
getting
happy
riding
his
bike
on
our
and
most
to
deploy
common
kubernetes.
B
He
really
does
not
need
to
learn
everything
about
given
route.
Api
object,
something
with
that.
So
many
of
these
of
our
developers
are
not
system
engineers
which
few
hundreds
develop
of
themselves,
are
and
very
good
at
work.
So
should
they
have
to
learn
the
few
related
complexities?
That's
the
question
that
we
were
trying
to
answer
when
we,
how
can
we
make
their
life
easier,
so
this
here
at
cube
conference
at
the
see
gaps
meeting
and
then
there
were.
B
He
said
in
his
talk
and
it's
a
brilliant
talk,
if
you
haven't
seen
like
really
urge
everyone
to
see
it
is
that
human
is
the
system
stuff
that
they
were
built
by
system
engineers
or
system
in
developed,
and
this
is
the
problem
that
we
have
seen
it
outers
when
we
have
talked
to
quite
a
few
developers
internally
and
externally.
So
what
we
did
is
when
we
saw
this.
We
tried
to
solve
this
problem
and
over
last
few
months
and
our
first
pry
was
well
first,
we
investigated
bunch
of
tools
that
existed.
B
This
is
around
eight
eight
nine
months
back.
One
of
them
was
composed
and
rejoined
countries
and
decided
to
the
great
tool
because
it
gave
a
great
user
experience
that
developers
who
are
using
docker
compose
who
had
a
mindset
of
traffic
compose.
It
really
gave
got
them
on
to
cable
edges.
So
we
really
we
when
I
say
we
I
mean
me
and
my
team
started
working
on
even
later
on
compose
and
we
have
been
adding
features
on
us.
We
have
done
bunch
of
releases,
but
we
soon
reached
a
limit
offer
and
spun
to
composes
limits.
B
It
is
the
limit
of
the
docker
compose
language
per
se,
and
it's
fair
enough
because
it's
made
for
that
thing.
It
does
not
have
one-to-one
mapping
to
to
burn
it.
So
we
were
hacking,
features
around
using
labels
and
whatnot
and
that's
the
different
story
altogether,
but
we
we
realize
that
it's
switching
the
limits
that
we
would
like
to
go
beyond.
That
is
the
point
when
we
started
thinking.
What
can
we
do
about
it,
and
so
we
started
thinking
and
we
start
used
our
learnings
and
we
thought
it
was.
B
It
is
a
good
idea
to
have
a
different
language
altogether
with
the
great
user
experience.
Hopefully,
who
should
be
extensible
when
I
say
great
user
experience
at
it.
Speed
simple,
is
it
explained
in
terms
of
the
contribution
language
that
we
will
use
and
the
tooling
and
the
management
that
means
they
are
having
to
generate
a
star
out
of
this
higher
level
abstraction
aware,
so
it
should
been
higher
level
abstraction,
hopefully,
which
will
generate
the
actual
Cuban
etic
objects
so
that
the
application
can
be
deployed.
B
The
developer
does
not
really
need
to
know
everything,
probably
almost
nothing
about
given
orders.
He
or
she
should
be
concentrating
only
on
their
ass
and
basic
configuration
about
their
application,
and
we
should
definitely
not
become
another
resource
and
hopefully
have
some
more
things
that
would
make
easier
bail
on
this
topic
a
little
bit
later.
So,
as
I
said,
we
have
been
working
on
this
as
a
prototype.
We
have
been
talking
to
other
people
as
well
getting
some
feedback
whatever
we
have.
H
So
this
is.
This
is
how
an
open
compose
spec
looks
like
okay.
A
lot
of
these
things
are
our
MVP
OC
stage
but
yeah.
This
is
a
descriptor
good
idea
of
where
we
want
to
reach
right.
So
this
this
open
opposed
file
is,
is
for
a
simple
container
for
this
application.
So
the
first
one
is
the
solar
for
services,
data
rate,
and
you
can
see
that
there
are
environment
variables
that
are
defined
and
there
are
environment
variables
which
we
are
mounting
using
secrets
which
will
be
translated
to
to
covalent
a
secret.
H
This
is
the
database
file,
then
down
there.
That
is
asking
Emma
messing
up
some
of
the
output
but
yeah.
So
this
is
the
second
service
which
is
the
web
service.
Again,
you
can
see,
there
are
environment
variables
and
there
are
ports.
There
are
volume
up
with
turbo
volume
outs
in
the
database
and
all
of
that,
so
all
of
the
application
definition
goes
into
this
and
there
in
the
end
of
the
file,
we
have
the
root
level
volumes
in
secret.
The
definition
most
of
this
should
be
self
explicitly.
H
That's
the
volume
way
of
creating
of
the
of
the
given
size.
Those
are
the
secret.
This
will
translate
to
the
secret
name
and
coburn.
It
is
these.
Are
the
data
keys
inside
of
kubernetes
things
evolve
evolve?
All
sorts
of
inputting
are
supported,
so
this
is
the
open
compost.
Pipe
I
will
quickly
get
out
of
the
thing
and
and
using
the
open
compost
tool
if
I
can
die
privately
I
will
convert
this
file
to
kubernetes
objects.
H
That's
some
digestive
measure
output
facility,
the
more
command
to
open
composer
s,
the
file
name
convert
and
oh
because
they're
directly,
to
which
I
want
to
output
the
to
which
I
I
want
to
output
equivalent,
as
are
tech
artifacts.
So
once
I
run
the
command
all
of
these
kubernetes
artifacts
are
generated
and
you
using
respect
that
you
just
say
just
saw
and
I'm
just
quickly.
Gonna
run
a
few
say
today,
so
all
of
these
kubernetes
resources
get
created
and
when
I
was
recording
randomly
my
database
decided
to
get
stuck
in
container
gated
so
I
quickly.
H
E
H
A
successful
connection
to
database
as
we
need
I'll
get
the
service
IP
and
then
quickly
try
to
connect
it
using
e
links
of
it's
a
command-line
browser,
thingy,
so
yeah.
So
so
the
application
is
our
wordpress
is
running.
I
press
enter
okay,
so
so
this
will
give
you
other
the
definetely.
Your
wordpress
reggie,
like
when
you
open
unity
in
the
in
the
browser.
So
that's
the
demo
that
identical.
B
Functional
so
of
as
you
see,
we
have
been
building
the
abstraction,
as
we
call
it
of
the
language
whatnot,
but
obviously
there
are
challenges
and
we
have
been
working
on
some
of
them.
We
know
that
abstraction
can
velocity
or
leaky
they
completely
aware
of
that,
and
we
are
thinking
what
we
could
do
about
this.
We
also
know
that
kubernetes
press
and
various
attributes
in
APL
of
different
bill
for
reasons
so
who
and
they
have
been
booked,
they
have
been
discussed.
We
have
been
proposals
and
that's
why
they
are
there.
B
So
that's
the
thing
that
we
are
working
on
and
then
also
we
need
to
always
remember
that
we
should
not
make
it
another
difficult
resource
to
remember
by
the
developer,
so
we
shouldn't
be
shifting
the
problem
domain.
So
obviously
we
should
always
keep
it
simple.
I
know
it's
a
little
bit
of
ribbon
at
this
point,
but
that's
very,
very
important
and
paramount
for
us.
So
what's
the
current
status
and
next
step,
we
have
seen
what
we
have
this
demos.
B
If
you
go
to
the
github
repository,
the
link
is
at
the
end
of
the
slides
I'll
show
to
you.
You
will
see
that
there
is
some
documentation,
the
language.
What
we
have
there
is
more
or
less
defined
in
the
documentation
that
we've
member
well
in
the
separate
documentation
folder.
There
are
some
examples
we
are
working
on
for
examples
upwards,
but
there
are
other
things.
B
So
I
am
talking
this
from
giove
discuss
again,
but
this
is
all
this
is
what
I
really
like
that
about
the
talk,
and
one
of
the
things
is
that
with
line
walls
change
when
the
easy
things
should
be
easy
and
hard,
things
should
be
made
possible,
and
this
is
the
goal
that
we
want.
We
really
want
to
keep
things
simple
as
well,
and
we
don't
want
just
because
you
are
trying
to
abstract
out
things.
I
don't
want
to.
B
We
don't
want
to
may
miss
out
on
the
points
or
miss
out
on
the
features
or
the
use
cases,
whereas
people
with
advanced
care
use
cases
are
intermediate,
use
cases
are
ignored,
so
we
we
don't
want
to
do
that.
Keeping
that
in
mind,
as
Michelle
said,
that
there
are
various
initiatives
around
it
like.
If
you
plot
with
dissonance
and
whatnot,
they
all
initiates
each
and
every
initiative.
This
kind
so
have
there
take
on
the
problem.
We
are
definitely
talking
to
them.
B
There
is
a
health
anticipations
and,
on
the
figures,
I
have
some
calls
around
days
between
a
bunch
of
us.
A
this
team
score
with
bitNami
I
feel
quite
a
few
people
love
us
for
boxing
but
anyway.
So,
let's
see
where
that
goes.
But
our
goal
is
to
make
sure
that
the
thing
is
really
help
developer.
So
we
would-
and
we
will
reach
from
our
roadmap
that
we
will
add
integrations
with
IDs
using
language
server
protocol
or
whatever
it
is
the
various
tooling.
B
We
have
some
ideas
about
integrating
it
meet
this
maze
and
maybe
because
the
punch
up
places
well,
java-based
micro
services
would
be
can
be
deployed
on
various
privileges
or
shift
Buster's
and,
if
you'd
like
to
think
in
terms
of
workflow
management
as
well.
That's
what
I
mean
extension
comes
in,
so
that
developer
can
push
his
hell.
Her
open,
compose
file
or
application
file
into
the
CIO
for
the
free,
and
somebody
else
can
take
it
and
also
add
in
little
bit
of
an
effect.
B
You
are
Yama's
that
would
take
it
to
our
production
and
things
like
that.
But
and
to
be
honest,
we
don't
have
all
the
answers.
Yet
we
don't
know
if
our
solution
is
the
perfect
solution.
I,
don't
think
we
have
the
monopoly
over
it,
but
we
are
working
towards
it
working
towards
to
finding
a
good
solution
that
works
for
as
many
cases
as
possible,
and
we
are
definitely
hoping
to
get
some
feedback.
This
is
where
we
generally
hang
out
on
the
gap
channel
and
various
of
the
cumulative
slack.
We
have
a
github
repository.
B
A
A
Alright
doesn't
doesn't
look
like
any
questions.
Thank
you
guys
very
much,
all
right
so
now
on
to
the
tail
end
of
our
community
meeting
notices
all
right,
so
it
looks
like
we
have
a
notice
about
the
time
of
cig
testing,
changing.
A
I
Used
to
be
at
9:30
a.m.
Pacific,
we've
now
moved
it
to
1
p.m.
Pacific
so
as
to
avoid
conflicts
with
all
the
other
states
having
a
meeting
on
Tuesdays.
So
if
you
wanted
to
attend
sick
note,
sick
windows
or
was
a
clustered
lifecycle,
I
think
or
sick
cluster
offs,
one
of
the
I
sorry
I
always
get
those
two
confused.
A
few
words
heading
those
and
we're
sad
that
she
couldn't
attend
sick
testing.
Now
you
can
so
come
join
us
at
1:00
p.m.
looking
forward
to
seeing
you
all
thanks
right.
F
Just
wanted
to
echo
on
that
I
think
like
one
of
the
hardest
things
to
do
in
this
egg
is
accommodate
all
of
the
time
zones
and
all
of
the
people
in
like
all
of
the
community,
and
it
is
really
hard-
and
you
know,
as
we
were,
trying
to
be
a
global
team.
Like
you
know
the
earlier
comments
about
like
the
fact
that
some
of
the
people
and
release
managers
are
saw,
you
know
they're
doing
everything
that
they
can.
F
It's
just
scheduling
it's
tough
and
I'm,
beginning
to
find
at
least
on
my
side
that,
including
everybody
in
all
of
the
meeting
is,
you
know
almost
more
untenable
than
just
getting
all
the
work
done.
All
the
information
passed.
So,
what's
really
great
is
you
know
it
seems
like
everyone's
trying
to
do
things
and
trying
to
be
flexible,
and
you
know
keep
it
on
that.
Ball
is
important
for
the
things
that
makes
kubernetes.
F
I
Totally
agree
with
that,
this
was
sort
of
some
one
question
feedback
so
just
to
like,
if
you
feel
like
you're,
not
able
to
kind
of
say-
and
you
could
really
like
to-
please
speak
up
about
that-
let
us
know,
because
if
people
don't
actually
say
something
about
it,
we
feel
like.
Maybe
there's
not
a
need
to
try
that
out.
I
know
the
community
meeting
had
at
one
point.
I
We
tried,
like
shuffling
I'm
like
once
a
month
to
a
more
Asian
friendly,
Khan's
own
and
I,
know
that
sig
note
alternates
between
a
one
time
and
another
to
accommodate
the
Asia
timezone
I
weekly
basis,
so
we
definitely
have
options
to
be
flexible.
If
there's
demand
for
it,
you
know
specific
asks
where
what
precipitated
this
change
for
sig
testing.
If
you
want
to
see
what
the
full
schedule
is
for
all
the
SIG's,
you
can
go
check
out
the
community
github
repo
there's
a
syphilis
signatures,
also
kubernetes
io
/
community.
I
Has
this
really
awesome
calendar
that
has
all
the
different
signe
things
that
you
can
add
to
your
town?
Keep
it
like.
In
course,
you
can
do
color,
which
was
what
I
used
to
figure
out
a
time
that
didn't
can
click
on
Tuesdays.
Oh
that's
great!
It
is
that
in
a
minute
at
all
or
I,
didn't
know
about
that.
Now,
if
you
go
to
kubernetes
dot,
io
and
then
click
on
the
community
tab
up
top
it's
there
yeah.
F
Cool
thanks,
yeah,
I,
guess
I
just
wanted
to
chime
in
again
and
just
wanted
to
say,
especially
when
it
with
the
release.
Manager
updates,
which
are
I
alluded
earlier,
are
really
important
again,
not
that
these
people
aren't
doing
everything
they
can
should
and
more
because
being
a
release.
Manager
is
such
a
hard
thing
to
do,
especially
for
the
patch
releases,
I
think
when
the
attention
kind
of
focused
on
the
new
release.
But
you
know
it's
scheduling,
hard
and
I.
Think
the
thing
that
we
need
to
focus
on
is
you
know.
E
Have
the
ones
asking
a
question
so
we
used
to
have
to
build
clock,
but
all
is
the
Google
internal
rotation,
but
we
share
with
the
community.
So
there's
a
lot
of
those
you
clients
that
I
saw
those
email
come
back
and
talk
about
to
have
the
community
people
include
as
the
build
cop
and
so
right
now
longtime.
E
We
we
used
to
have
to
build
hard
money
for
all
those
past
the
straight
and
they
need
to
the
remains
or
not
really
into
ginnis,
and
also
for
the
PR
builder
and
submit
view
and
now
listen,
close
the
scope
and
Joe
the
role
response
pin.
It
is
not
when
and
clear
defined
lately
so
and
so
I
just
show
this
cleansing
and
the
should
we
just
kick
out
a
new
process.
Now
you
have
the
notation
among
the
community
members
and
as
the
beautiful
and
and
we
define
the
role
and
responsibilities.
Builder
cups,
I.
I
Idea
so
I
have
two
thoughts,
maybe
that
something
that
falls
under
sig
testing
to
maybe
that's
something
that
falls
under
sig
release
since
there's
sort
of
a
person
who's
in
charge
of
triaging
testing
bugs
during
the
release.
Three,
the
kubernetes
leadership
summit,
if
you've
been
invited
to
that
I,
think
that's
a
place
for
a
lot
of
governance,
related
issues
and
how
to
spread
the
load
away
from
Grimble
to
the
rest
of
the
community
is
a
really
great
forum
for
that
discussion
and
then
for
I.
Think
the
build
top
ultimately
has
like
great
power.
I
The
ability
to
like
push
things
into
the
repo
to
unblock
the
queue
or
you
know
they
can
sort
of
go
through
all
the
safeguards.
So
you
would
really
want
to
make
sure
the
pool
of
people
who
are
rotated
through
for
that
are
highly
trusted.
Members
of
the
community,
so
I
agree
with
you.
We
should
define
what
that
role
is,
so
we
can
define
who's
responsible
enough
to
have
it.
E
That's
good
and
appoint
yeah,
so
I
think
I'm.
Actually
I
do
so.
We
so
it
means
I
know
we
just
resume
our
kind
to
build
cop
as
yes
and
notation
is,
and
also
connect
the
Idaho
for
the
responsibilities.
I
notice
that
last
couple
weeks
will
come
in
to
act
like
the
old
to
build
cops.
The
house,
because
we
have
a
lot
of
confusing
I,
can
still
connected
effort
and
definitely
the
role
so,
but
there
are
certain
things
is
not
to
have
the
coverage.