►
From YouTube: Kubernetes SIG API Machinery 20170621
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
B
A
C
A
D
The
last
community
meeting
when
a
partner
was
going
to
the
list
of
things
she
brought
up
CID
going
directly
to
beta
and
a
bunch
of
people
thought
that
she
was
talking
about
most
the
discussions
about
CRT
going
directly
to
beta,
but
people
thought
about
a
derogation
and
that
spawned
a
discussion
of
some
people
were
surprised
that
API
aggregation
was
on
by
fault.
There's
a
discussion
in
the
community
features
repo
about
whether
it
is
appropriate.
It
was
out
of
tree
in
1/6
an
alpha
state.
D
A
It's
kind
of
weird
the
way
we
talk
about
API
sync
data
without,
like
like
the
status
of
the
API,
has
really
literally
nothing
to
do
with
the
status
of
the
features
yeah,
the
the
API
being
beta
just
means
that
we're
willing
to
support
it
and
be
backwards
compatible
under
the
bit
future.
The
question
here
is:
if
the
features
data,
quality
or
not.
D
Well,
and
not
only
that
but
like
I,
think
I
mean
I
would
say
that
people
often
complete
the
API
the
feature.
It
is
an
important
thing
for
the
project
to
use
this
as
a
precedent
to
make
the
discussion
about
graduation
a
little
bit
clearer
so
that
you
know
if
the
next
time
this
comes
up,
there's
precedent
to
fall
back
on
and
it
recorded
whatever
happens
here.
I
have.
E
C
Well,
I
also
think
we
should
just
talk
about
what
it
means
when
we
say
it's
owned
by
default
and
what
we're
I
think
what
that
sentence
means
is
that
you
will
pass
through
aggregation
code
on
a
request.
If
you
actually
want
to
aggregate,
you
have
to
turn
on
the
API.
You
want
to
aggregate
right,
so
I
think
we
need
it
just
in
case
people
are
unclear
right
there.
There's
too
many
things
you
could
be
make
up,
and
you
say
the
aggregation
is
on
by
default.
C
D
Time
we
read
well
I,
think
it's
fun
I!
Think
it's
really
important,
though,
to
distinguish
with
maybe
have
an
internal
refactor
that
impacts
everybody.
When
we
talk
about
the
impacts
to
everybody
else,
how
do
we
message
that?
And
how
does
everybody
kind
of
get
by
and
because
that's
that
part
of
the
concern
right
is
if
cube,
admin
went
and
changed
how
the
API
server
was
set
up,
and
this
say
didn't
like
have
a
discussion
with
it.
There's
a
there's,
a
question
of.
E
E
D
F
D
It
new
to
this
discussion,
excitable
or
total
context
and
background
is
like
we're
trying
to
change
from
a
model
where
a
bunch
of
independent
teams
do
whatever
they
have
to
do
to
get
to
the
next
gate,
and
as
long
as
it
works,
it
doesn't
matter
to
one
we're
at
a
minimum.
We
are
explicitly
communicating
and
making
sure
we
have
a
clear
you
know
set
up
residence
across
the
different
things
specifically
so
that
we
have
a
looser
model.
A
D
A
Yeah
I
I,
don't
know
I,
don't
know
how
much
sympathy
I
have
for
that,
like
no
I
even
made
it
I.
Don't
even
think
this
deserves
a
features.
Free
poetry,
but
I
made
one
and
stated
at
the
like
months
ago.
That
said
it
was
going
to
be
on
by
later
and
on
my
default
is
our
target.
I
was
looking.
We
have
done
to
spread
this
even
more
widely
to
children.
The
community
leading
up
and
bid
already.
G
A
A
A
A
Okay,
so
I
don't
know,
I
don't
feel
like
we
got.
D
An
action
item
out
of
that
really
so
I
think
the
action
item
would
be
to
make
sure
that
all
the
I's
and
teams
are
crossed
now
and
the
justifications
are
recorded
in
the
issue
so
that
we
have
something
to
fall
back
on
and
then
be
open
to
continue.
The
discussion
I
mean
really
just
was
I.
Don't
think
it's
unreasonable
for
somebody
to
come
in
and
ask
us
a.
Are
you
jumping
directly
to
data,
and
this
is
close
like
I
mean
I?
D
A
D
It's
still,
gated
to
a
reasonable
degree.
Captain
is
often
the
requirement
for
beta
I've,
never
seen
the
recommended
as
a
requirement
for
beta
I.
Think
my
point
is:
is
that
we're
crossing
really
really
like
like
I,
understand?
You're,
absolutely
understand
this.
This
is
like
we
feel
like
we're
hot
dogging
it,
and
the
ask
is
to
be
to
be
better
at
communicating
to
be
not
so
hot
dog
looking,
and
so,
if
that
involves
us
say
I
like
this.
D
The
impact
of
this
is
very
minimal
because
you
can
simply
remove
all
a
great
extensions
when
you
fall
back
and
we've
had.
You
know,
do
much
so
time
on
the
core
code.
It's
just
a
refactor,
therefore
we're
okay.
That
might
be
enough,
but
that's
not
what's
being
communicated,
we're
just
kind
of
saying
well,
it
was
already
an
alpha
right,
so
yeah.
E
I
think
that
could
be
a
follow-up
item
here
as
well
answer
the
question
with
us
earlier.
You
know,
which
is
what
could
we
have
done?
Like
literally?
Is
it
a
set
weaknesses?
Music,
where
you
were?
You
have
a
like
an
obvious
place
to
run
this
app
people
right,
I,
don't
know
where
that
is
exactly
or
like
what
we
would
have
done
it,
but
maybe
Barnett's
ourselves
right
waiting
for
some.
E
D
F
A
A
D
Doesn't
mean
that
people
aren't
impact,
that's
what
I'm
saying
it's!
Okay
for
us
to
do
this!
We
just
need
to
get
better
at
communicating
and
messaging
right.
It's
like,
if
we
had
said
hey,
we're
going
to
refactor
the
whole
codebase,
and
you
know
publicly.
When
we
do
these
things,
we
just
need
to
get
better
at
pre-loading
it
that
isn't
post
loading
it
so.
A
You
know
I
I,
don't
necessarily
want
to
be
a
full-time
politician.
Right,
like
like
I,
mean
welcome
to
open
source
right.
Yeah
I
mean
a
little
bit
is
fine
but
yeah.
It
was
very
hard
to
get
things
done
in
this
project
and
I
think
a
lot
of
people
like
their
stakeholders
and
that's
good
and
you
know
I-
want
people
to
feel
invested
in
and
all
that.
But
you
know
if
you
come
to
a
point
where
you
have
to
get
buy-in
from
50
people
to
do
any
things
and
nothing.
D
Is
going
to
happen
right,
so
the
new
model
I
mean
the
model
that's
been
discussed
is
that
we
are
moving
more
aggressively
to
like
lazy
consensus
and
I.
Think
what
to
the
point
about
reflectors
that
don't
touch
anybody
else.
I
think
those
are
like.
Those
are
pretty
clear
and
we
just
should
communicate
that
aspect
of
it
better
and
then
we
do
cross
boundaries.
D
We
need
to
be
more
aggressively
communicating
it
so
that
lazy,
consent
system
do
its
work
right,
like
I,
think
you're
right
like
the
it's
not
ask
for
forgiveness,
but
it's
we're
making
everything
better.
Here's
how
we're
going
to
do
it
we've
taken
into
account
concerns,
and
we
do
at
least
make
sure
you're
aware
of
it
like
that
kind
of
stuff.
I
think,
is
some
of
the
iceberg
to
the
more
recent
discussions.
H
A
H
A
I
also
think,
just
as
someone
who,
like
kind
of
vaguely,
pays
attention
to
new
features
in
this
area,
but
it's
not
keen
on
the
call
all
the
time.
Sometimes
it
feels
like,
like
things
are
advertised
a
little
bit
later,
and
people
would
be
a
little
bit
less
surprised
if
they
were
like
kind
of
aggressively
advertised
at
the
beginning.
Hey
we
like,
for
instance,
API
user.
Hey.
We
have
this
alpha,
API
aggregator,
please,
please,
please
go
check
it
out
because
it's
probably
going
to
become
more
central
in
a
release
or
two
versus
oh
hey.
A
G
Have
a
concrete
notion
of
that
right
because
the
proposal
was
updated.
Last
November
we
updated
the
proposal
again
early
in
the
year.
In
March
we
announced
at
the
community
meeting
that
this
would
be
intent.
We
had
a
feature
issue
that
got
created
in
1:7
and
in
the
first
weeks
of
1-7
it
was
available
so
that
was
communicated
seven
months
ago.
It
was
completed
five
months
ago.
It
was
going
to
get
four.
A
A
D
Often
like
oh
this,
this
is
not
just
like
that.
This
is
coming
up
in
a
lot
of
context.
I
think
it's
for
better
or
worse
API
machinery,
because
it
goes
and
gets
so
much
done,
and
because
it
has
so
much
impact
and
because
it
is
so
important
like
it
gets
to
the
example
in
these.
But,
like
you
know,
cluster
lifecycle
changing
how
Cuban
men
did
a
che,
or
some
of
those
things
are
like
great
examples
of
like
alright
I.
A
E
Sorry
I
think
that
it's
possible
that
we
simply
didn't
say
the
right
stuff
at
the
earliest
opportunity,
like
we
just
recounted
been
like
10
months
ago
and
n
minus
4
months
ago,
and
that's
fine
as
possibly
do
need
to
say
it
that
many
times,
but
we
didn't
say
the
right
stuff
early
enough.
So
we
didn't
say
the
right
place.
I
think.
E
D
Up
the
emails
and
discussions
I
will
note
that
the
proposal
still
doesn't
talk
about
aggregation.
So
does
a
like
we
like
to
the
point
about
maybe
the
right
place
for
those
we
talked
about
it
in
many
forums,
but
maybe
we
need
to
do
a
slightly
better
job
of
focusing
some
of
the
places
where
we
record
these,
so
that
we
do
have
that
record.
I'm
done
might
be
one
small
bit
of
feedback
Oh,
so.
A
Starting
to
keep
feeding
the
third-worst
but
I,
imagine
the
hypothetical,
where
we
communicated
too
much.
How
would
we
know
I
think
people
would
actually
still
complain
that
we
haven't
communicated
enough
like
how
do
we
know
that
that
that
content
pop,
how
do
you?
How
do
we
really
overcorrecting,
basically
like
thankless
I,
think
we
probably
could
have
communicated
better
but
I
also
feel
like
it's
really.
You
need
to
get
into
a
pattern
where
no
matter
how
much
you
communicated
everybody
says
you
should
communicate
more.
Like
oh
I
didn't
hear
about
this.
A
G
D
Hospitals,
burger
and
that's
a
great
point,
I
think
what
I
would
say
is
there's
smoke
I'm,
not
saying
that
we've
got
a
giant
fire,
but
there
is
smoke,
and
so,
like
maybe
I,
would
say,
I
think
that
the
hypothetical
is
is
absolutely
valid.
If
we're
not.
If
we're
going
to
target
communication
more
effectively
or
we're
going
to
do
more
at
central
communication
of
like
here's,
the
big
changes
coming
to
cube
from
from
say
gay
bi
machinery-
maybe
that's
one
you
know
concrete
is
it
is
it
something
is
more
accessible
than
oh
yeah?
D
E
Going
to
propose
one
more
thing:
that's
not
really
an
action
at
the
game!
Is
that
I'm
not
sure
who
should
call
upon
it?
It's
an
idea,
as
I
need
easier
to
keep
track
of
this
sort
of
interesting
communication
sort
of
like
you
have
done
them
up
stuff.
It's
coming
world
yep,
instead
of
coming
in
the
form
of
both
file
at
random
intervals
or
features
or
community
meetings.
E
D
C
C
We
say
it
once
in
a
newsletter
or
whatever
it
is,
and
we
all
agree
that
if
it's
been
said
there
you
can
collect,
you
can
ask
questions
whatever,
but
at
that
point
you
know,
Daniel
is
essentially
off
the
hook
and
I
think
that's
important,
because
if
we
the
more
places
we
keep
repeating
this
stuff
and
the
more
things
you
have
to
say,
I
think
impacts
to
Daniel's
earlier
point.
The
messages
are
going
to
the
messages
will
get
lost
like
you're,
not
going
to
know
about
something
because
it
becomes
too
long.
Didn't
we
yeah.
F
So
I
I
think
just
writing
down
how
these
things
are
communicated
would
help
a
lot
I
think
there's
one
kind
of
communication
for
breaking
some
changes
to
an
API
application.
For
example.
That
requires
a
lot
more
legwork,
because
there
is
absolutely
going
to
be
impact
and
there's
another
bucket,
which
is
there's
no
intended
impact.
There
may
be
additional
functionality
and
is
how
those
will
be
communicated
and
we
can
skip
alpha
now,
I
think
in
Daniel,
beautiful
green
future.
F
A
I
think
for
me,
in
this
project,
probably
a
year
ago,
at
this
point
at
some
point,
I
had
to
come
to
the
company's
area.
Come
to
be
okay
with
the
fact
that
I
can't
keep
track
of
everything
and
people
are
going
to
go
off
and
do
things
but
I,
don't
agree
list
and
I
just
have
to
make
make
my
peace
with
that
and
I.
A
E
It
won't
be
inviting
you
won't
be
able
to
turn
in
whether
or
not
that's
really
true
Yeah
right
like
if
it
if
it
becomes
2006
or
whatever
right
like.
We
won't
really
be
able
to
know,
and
so
I
think
it's
really
important
to
figure
out
some
way
to
let
people
know.
We
don't
think
this
to
break
you
but
telling
you
that,
and
you
also
like
it,
avoid
people
complaining
yeah.
G
I
was
actually
open
to
a
demo.
I
have
the
aggregators
set
up
I've
just
launched
a
local
up
cluster
and
I'd
like
to
demonstrate
a
particular
API
server
and
it
won't.
Let
me
share
my
screen.
Yeah
I
probably
have
to
I,
probably
have
to
do
something
says
you
need
to
stop
sharing
Susan
resting,
oh
well!
Thank
you
all
right.
Okay,
all
right
I
am
now
looking
for
me.
I
think
you
should
all
be
looking
at
a
terminal
right
now.
I
G
Right
so
that
aggregator
that
people
don't
necessarily
want
on
by
default
I,
am
going
to
go
ahead
and
show
how
it
works
right.
One
of
the
problems
that
we
have
today
is
that
when
you
use
our
back,
you
try
to
so
divide
your
cluster.
You
end
up
subdividing
permissions
on
link
spaces,
but
you
have
a
choice.
You
can
either
have
everyone
see
all
available
namespaces
like
the
cluster
admin.
You
can
do
right
here
that
I
demonstrated
or
you
make
it
so
they
cannot
see
new
spaces.
G
So
if
I
try
to
do
the
same
request
is
David
David
can't
see
the
namespaces,
but
what
you
probably
want
is
for
someone
to
be
able
to
see
the
namespaces
they're
allowed
to
see
I'm
going
to
do
a
list,
and
so
I'm
going
to
a
great
in
an
API
server.
That
actually
does
this
in
OpenShift
can
call
these
projects
so
I'm,
creating
a
namespace
and
I'm.
G
G
Sata3,
that's
where
that
is
the
bottom
of
the
screen.
Okay,
there
we
go
I'll
keep
everything
about
there,
so
the
API
service
is
ready
to
use.
You
can
see
that
I
have
pod
technique.
You
guys
can't
see
what
I'm
typing
all
that.
So
now
we
have
any
spaces
we
have
and
so
I'm
going
to
a
get
projects
and
as
the
cluster
admin
you
can
see,
I
have
a
few
projects.
If
I
do
the
same.
G
Call
as
David
you'll
see
that
I
have
I,
have
no
projects,
I
mean
because
I
have
no
Commission's
to
see
any
of
these
namespaces.
So
now,
I
need
to
get
a
new
space
as
David,
and
so
there
is
a
way
to
do
it.
I'm
not
allowed
to
create
any
in
space
that
I
want,
but
I
can
create
a
resource
and
I
have
an
example
of
the
resources.
Link.
I
can
create
a
project
request
resource
and
this
will
actually
provision
a
namespace
one.
G
So
I
am
going
to
go
ahead
and
you
want
to
up
and
it's
a
bit
of
create
a
project
request
and
then
I
get
the
project's
at
David.
You
can
now
see
there's
a
project
in
the
list.
If
I
get
namespaces,
you
can
see
that
the
project
is
actually
a
namespace
and
I'm
getting
a
chases
as
the
cluster
admin
here
than
any
place
exists,
and
it's
perfect
now
I
have
another
user
say
Daniel
Daniel
will
not
be
able
to
see
my
project.
G
Clearly,
you
look
at
god-knows-what,
make
everything
beta
a
student
as
an
editor
to
me
to
one
of
my
existing
projects,
and
this
is
just
using
a
normal,
are
back
role
binding
and
now,
when
I
get
projects,
as
in
I,
am
able
to
see
the
overshift
project
that
I
created.
So
this
is
just
an
example
of
an
API
server
that
can
be
built
with
aggregation.
G
E
G
A
One
of
the
so
one
of
the
I
think
that's
actually
a
pretty
big
demonstration
of
what
the
concept
is
for
I
mean
it
could
be.
It
could
be
any
any
pads
like
it
created
as
an
SP
OPM,
but
it
is
open
ship
for
build
modular
lis
such
that
logical
resource
groups
were
serves
out.
A
single
API
is
like
you.
Could
you
could
mix
and
match
yeah.
F
G
A
I
think
explicitly.
That
is
why
we're
making
the
the
aggregation
layer
like
like
we,
we
think
we
should
do
that,
not
just
with
open
ship
resources,
but
also
tremendous
resources
right,
like
the
the
ones
that
are
that
are
optional,
you
should
be
able
to
choose
which
are
which
are
installed.
This
is
a
good
way
to
do
it,
yep
all
right
cool.
Thank
you.
J
Typescript
yeah
so
I
actually
designed
right
here,
so
he
actually
asked
you
can
can
act.
I
stood
by
you
know
fun.
You
guys
know
that
that's
would
be.
A
blank
table
has
hit
their
code
but
auto
ship
emotion,
so
pay
attention
to
joystick
and
been
an
accomplishment
type
type
script
to
be
clear,
easily
like
a
rapidly
record
of
scripts.
That
period
I,
take
checks
or
typing
classes,
IMF,
grunt
and
and
come
time
to
jealous,
but
also
in
jealous
excited.
Why
I
just
be
done
already,
but
do
you
need
a
leader.
J
A
J
A
J
A
C
A
K
Do
think
that
that
is
true,
that
there
is
I
think
there
is
technically
one
I,
don't
think
it's
a
special
group.
Great
I
do
want
to
say
that
my
own
personal
opinion
is
that
the
aggregator
stuff
is
fantastic
and
I
will
just
like
as
a
quick
aside
for
those
that
might
be
watching
on
YouTube
or
in
the
college.
Already
know
me,
like
I've,
been
leaving
the
Service
Catalog
sig
over
the
last
I
guess,
like
ten
months
now
and
I
did
a
very
hard
sell
of
the
aggregator
to
a
very
skeptical
crowd.
K
When
we
started
down
the
road
of
yes,
we
do
need
to
build
a
user
API
server
instead
of
relying
on
like
GPRS,
because
we
need
some
features
that
we
can
only
get
that
way
and
for
whatever
to
earth.
To
the
folks.
On
this
call
like
like
I
said,
there
was
a
lot
of
skepticism
and
now
that
people
are
seeing
the
aggregated
working,
we
had
a
demo,
the
aggregator
aggregating
together
the
core
API
server
and
service
catalog,
a
piece
of
API
server
of
I.
Think
people
are
in
service.
K
K
You
go
into
the
notes.
I
would
have
said
effusive,
but
you
know
if
you
want
to
use
gushing
totally
fine,
so
yeah
good
job
on
the
api
aggregator
and
like
it
really
is
pretty
awesome
to
see
like
the
vision
for
this
stuff
that,
like
a
lot
of
us,
have
had
in
our
heads
for
a
long
time
coming
together
and
being
something
that
we
can
show
people
math,
so
big
good
job
guys,
given
what
you
just
said,
I
feel
a
huge
butt
is
coming
in
here:
no
there's
no
buts,
so
I
here's!
K
My
story
is
that
we
had
started
building
service,
catalog
API
server
on
the
I
think
this
must
have
been
like
one
five
candidate
machine
Ernie.
He
had
a
rebase
four
one,
six
candidate
machinery
and
then
first
thing.
Monday
morning,
when
I
checked,
my
email
I
saw
that
somebody
maybe
last
week,
don't
remember.
I
saw
very
early
in
the
morning
some
day
that
someone's
first
contribution
to
the
Service
Catalog
incubator
repo
was
rebasing
on
to
the
one
seven
candidate
machinery,
which
I
certainly
applaud
the
ambition.
K
K
In
thinking
about
that,
I
realized
that
and
I
I
will
say
in
advance.
I
do
not
claim
to
have
read
every
piece
of
written
communication
since
that
has
been
a
theme
of
this
meeting.
I
come
to
you
as
an
ignoramus
clothed
in
burlap
dirty
part
of
the
unwashed
masses.
My
question
to
this
group
would
be
how
do
I
know
what
specific
versions
I
should
I
should
rebase
on
to
and
yeah
so
I
one
sorry,
one
corollary,
forgive
me
Daniel
my
brain
as
much
slower
than
yours,
but
it
gets
the
job
done.
K
If
you
give
it
enough
on,
the
corollary
would
be
like
what
would
be
like
the
communication
channel
we'd
use
to
communicate,
communicate
things
like
there's
a
bug
that
I
know
that
David
had
fixed
where
we
were
putting
things
into
the
wrong
place
in
TV.
So
I
would
do
that
as
like
a
like.
You
definitely
want
this
fixed
type
of
situation.
What's
what's
clear
communication
channel
for
that
POS,
okay,.
K
A
A
We
formalize
that
a
little
bit
more
and
and
distributed
amongst
our
other
reposts,
because
I,
don't
think
reversion
them
super
well,
I.
Think,
probably
what
it's
going
to
look
like
is
we're
going
to
maintain
release
branches
and
development
and
development
master
and
we'll
probably
be
making
incompatible
changes
in
master
and
people
need
to
vendor
in
a
branch
and
as
we
find
bugs
we'll
fix
them
and
after
and
then,
but
you
know
just
just
like
it
just
like
any
other
library
we're
going
to
have
to
have
the
whole
nine
yard.
A
C
K
That
is
excellent
and
I.
Remember
you
having
been
in
the
same
room,
listen
ill
dire
in
Prior
meetings.
So
that's
good
I
think
the
the
only
facet
of
this
that
hasn't
already
gotten
into
the
notes
and
into
the
conversation
so
far
in
the
meeting
is
it
would
be
as
a
user
of
the
machinery.
It
would
be
awesome
to
be
able
to
understand
the
changes
that
I
would
have
to
make
when
I
do
the
rebase.
K
So
as
an
example,
like
I
said,
we're
actually,
we've
been
on
this
pre
old,
a
candidate
1
6
candidate
machinery,
and
when
this
person
came
along
and
did
the
rebase
they're
like
oh
well,
this
takes
away
insecure.
I
did
not
expect
that
and
I
don't
think.
That's
a
failure
of
anybody.
I'm
only
illustrating,
like
those
kinds
of
things,
would
be
awesome
to
know
like
like
hey.
There
is
a
new
version
that
all
works
together.
A
Got
15
minutes
left
until
three
items,
so
I
want
to
kind
of
run
this
down.
Oh
I
think
the
things
you
are
asking
for
are
things
that
people
have
asked
for
in
the
clients
code
repository,
like
literally
everything
your
X's
so
like
like
yeah.
That
doesn't
surprise
us
right
now
is
basically
I
have
a
functioning
change
log
and
actually
populated
correctly.
A
You
have
a
change
log
in
the
crank.
Oh,
we
are
still
working
on
figuring
out
how
to
populate
correctly
I.
Think
our
plan
for
the
client
bill
repository
is
they
have
like
is
to
ask
for
people
to
be
a
client
really
so
in
npr's,
but
I
think
once
we
actually
get
it
element
working
here
in
these
repositories,
we
can
have
the
release
notes
mechanism
legally
required.
Just
one
release,
no
mechanism
like
you're
really
so
goes
into
their
into
the
change
log.
For
that
repository,
yeah,
so
I
think
I.
K
K
E
E
It's
pretty
clear
zone
than
were
willing,
Alya
I'm,
going
to
use
a
community
normally
they
would
be
based
in
my
mind
and
announcement
and
in
Google
Mail
with
I
know
that
Daniel
would
love
there
be
more
man,
honest
expresses
their
communication
to
each
other
on,
but
one
of
the
things
is
when
you're
going
to
lose
intersection.
Someone
can
unsubscribe
from
the
global
committee
amendments
and
focus
on
how
our
users
straight
down.
What
does
that
sound
right
to
followers?
Not
that's.
K
J
A
F
I
added
that
this
is
Jago
I
joined
Google
a
couple
months
ago,
as
an
engineering
manager
on
this
team
and
as
we
get
close
to
1/8
I,
just
wanted
to
refresh
everyone's
memory.
It
looks
like
you
guys,
put
together
one
seven
dock
with
a
bunch
of
great
ideas
that
was
a
copy
of
a
one
six
stock,
so
I
would
expect
them
the
next
week.
So
maybe
I
can
just
copy
over
the
17.1
there.
Yeah
I
will
fill.
A
F
A
A
Yes,
okay,
we're
using
API
server
counts.
Okay,
yes,.
A
B
So
this
is
econ
from
course,
and
so
empanada
is
o
interns
is
similar
and
so
walk
to
give
him
this
issue
to
welcome,
but
not
sure.
What's
the
extra
status
of
this
one
saw
a
bunch
of
a
PRS
link
to
the
one
and
I
think
there's
one
PR
that
using
the
config
map
to
income
the
ratio
to
fixation,
but
it
is
being
closed
and
I'm,
not
sure.
What's
the
final
position
is
what
I
think
we
try
to
come
up
with
the
proposal
and
I'll
try
to
discuss
what's
a
direction.
We
should
go
here
so.
B
H
So
if
you,
especially
if
you
have
one
that
goes
down
for
whatever
reason
and
you
try
and
hit
the
service
for
kubernetes,
if
one
of
those
backing
API
servers
is
down,
it's
still
going
to
try
and
direct
traffic
there
and
it
will
fail,
and
so
this
particular
github
issue
is
about
making
the
list
of
API
servers
and
the
endpoints
that
back
the
service.
That's
a
kubernetes
service
dynamic,
as
opposed
to
having
a
static,
API
server
count
and
not
being
able
to
route
around
failures.
When
API
servers
go
down.
Okay,.
A
E
Maybe
I
got
to
do
it,
which
is
quick.
What
is
it
that
you
would
like
to
get
from
bringing
this
up
in
this
meeting
rate
Afghan
requests
as
a
psychologist.
G
F
G
I
G
A
I
probably
do
have
a
lot
of
nice
office,
but
I
don't
know
that
I
am
going
to
have
like
just
realistically
speaking,
I.
Don't
know
that
this
is
going
to
follow
the
top
of
my
list,
so
I'm
not
sure
what
to
say
off
the
top
of
my
head.
I
G
F
Suggestion
would
be
to
get
the
review
in
soon,
since
this
is
a
going
to
be
freed
up
time
and
the
release
cycle.
If
it
comes
in
after
the
end
of
July
it
it
will
get
stuck
in
the
queue
I
saw
that
it
was
an
internship
project.
So
I
would
say
here:
are
the
design
for
photo
comes
through?
The
more
likely
will
be
15
X,
1m
yeah.
A
F
Maybe
something
less
contentious,
or
maybe
an
op
like
three
options
for
a
internship
project,
and
you
might
be
more
successful
in
that.
Like
what
else
I
know,
I
was
suggesting
that
you
come
up
with
three
options
and
post
the
new
users
group
that
will
be
created
sometime
in
the
next
week
yourself,
but
this
one
sounds
like
it's
somewhat
contentious
and
just
seeing
how
those
contentious
features
go
that
are
not
high
priority.
The
silence
you're
hearing
is
probably
not
a
good
sign.
Yeah.