►
From YouTube: Kubernetes 1.12 Release Team Meeting 20180827
B
A
A
That
I
was
hoping
that
steven
augustus
would
be
here,
but
it
looks
like
he's
not
so
there
there's
been
some
mention
of
a
few
things
coming
all
through
the
cycle
that
were
potentially
going
to
be
exceptions,
and
those
would
never
come
in
so
I
wanted
to
get
a
confirmation
from
Steven
that
that
was
basically
not
happening
at
this
point
right
because
tomorrow
is
code,
slush
and
code.
Freeze
is
the
week
after
so
looking
like
these
couple,
exceptions
aren't
going
to
be
accepted.
Just
wanted
confirmation
on
that.
A
A
C
A
And
the
there'd
been
a
couple
things:
I've
kind
of
wondered
about
their
status,
so
I
have
attended
sick
meetings
and
been
keeping
an
eye
on
them,
and
people
still
seem
confident
that
they're
moving
forward.
So
we
will
see
the
next
the
next
two
weeks,
especially
we'll
be
telling,
as
we
we've
got
four
weeks
remaining
and
you
get
halfway
through
that
you're
you're
really
down
into
the
point
where
it's
either
done
or
it's
not
happening.
So
we
will
see
that
I
guess
there.
C
Was
one
that
sort
of
made
it
to
light?
I
saw
it
on
a
different
slack,
but
apparently
the
windows
GA
might
be
pushed
to
1:13
I
saw
something
that
might
have
happened
in
signo,
so
I
think
that's
one
of
the
things
that
we
were
targeting
as
one
of
the
the
big
milestones
to
talk
about
and
I
think
it's
gonna
slip
yeah.
They.
A
It's
a
little
unclear
to
me.
They
had
a
set
of
things
so
that
the
one
bucket
said
stable
GA
and
that
obviously
applies
a
lot,
but
within
that
the
set
of
things
that
they
had
identified.
I
think
we're
all
done,
but
as
they've
talked
to
other
SIG's,
there's
quite
a
few
other
things
that
are
also
coming
so
yeah
I.
Think
that
one
there,
if
you
look
at
the
specifics
of
what
they
had
in
plan,
that's
what's
happening,
but
the
then
the
actual
notation
is
probably
not
even
gonna
say.
Probably
it's
not
happening
in
this
release.
A
Okay,
I
move
on
to
test
and
front,
so
our
current
set
of
test
failures
to
kind
of
briefly
talk
about
that
before
we
really
get
into
a
CI
signal.
We
have
a
lot
of
test
failures
right
now.
A
lot
of
them
are
happening
on
the
Google
platform
specifically,
but
because
so
much
of
our
test
happens
there
anyway,
some
of
them
it's
hard
to
split
and
as
I
look
at
the
symptoms,
I
feel
like
some
of
them
are
things
that
are
maybe
not
operator.
A
Specific
I
mean
they're
there
operator
problems
as
we
move
the
project
forward,
so
I
feel
like
gke,
maybe
a
slightly
a
bit
of
a
canary
there
and
I'm
just
a
little
bit
worried
to
say
like
because
in
the
past
release
we
said
well,
let's
remove
some
of
these
always
failing
or
always
flaky
gke
things
from
release
blocking.
But
if
they're,
our
only
source
of
signal
on
things
that
potentially
are
very
impactful
for
an
operator,
we
actually
want
visibility
on
that.
So
it's
kind
of
a
balance,
I'm
curious.
If
anybody
else
has
thoughts
on
that.
A
I
think
having
it
in
the
test
grid
for
blocking
like
we
should
be
getting
more
stuff
in
there
as
opposed
to
taking
it
out.
So
then
I
think
we
as
a
team,
have
to
look
at
it
and
decide
on
a
case-by-case
basis.
Should
we
consider
this
one
actually
blocking,
even
though
it's
in
test
great
blocking
yeah.
D
D
One
thing
that
up
generally,
if
it
has
like
come
picked
up
great
ones,
they'll
have
a
GPS
and
then
they'll
have
a
GC
computer
part
of
it.
If
that
one
passes
on
just
gke
once
turned
red,
then
yeah
at
least
from
in
111.
We
never
have
blocked
on
those
considering
that
we'll
be
taking
care
during
the
GK
3d
stream.
So
but
yeah
I
agree
that
that's
a
that's
our
one
of
our
provider
signals
you.
A
A
A
All
right,
I
will
move
along
then
so
last
week,
I
went
through
and
looked
at
the
the
failures
and
tried
to
kind
of
bucket
eyes
them
a
bit
more
and
and
there's
a
link
to
a
Google
Doc
there
and
we're
at
the
point
where
each
of
those
failure
symptoms
needs
to
be
its
own
distinct
issue
so
that
we
can
go
out
to
a
specific
state,
get
a
specific
owner
and
say
hey
this.
This
problem
needs
an
owner
needs
a
resolution.
A
If
we
have
too
broad
of
an
umbrella,
it's
hard
for
anybody
to
focus
in
on
anything
in
particular,
and
it's
easy
then
also
like
with
a
lot
of
these,
maybe
being
GAE
specific
somebody
to
kind
of
glance.
I
didn't
be
like
GK
issued,
not
gonna,
look
further
into
it,
but
as
I
looked
through
a
lot
of
them,
I'm
not
convinced
and
I've
only
got
a
few
that
I
marked
as
definitely
clearly
gke
or
google
specific
in
my
initial
triage,
so
I
think
Mohammed
you're
working
on
getting
issues
open
for
each
of
these
today.
A
F
E
F
Yeah,
so
that's
something
that
it
is
actually
quite
a
quite
a
problem
that
we
cannot
tell
exactly
what
the
failure
is
so
for
the
four
that
I
have
been
actually
trying
to
take
out
and
eliminate
them.
So
I
also
can
look
at
the
document,
and
yes
for
some
of
them,
we
still
don't
have
the
issues
for
tracking
issues
which
I'm
going
to
go
for
in
them.
There
is
a
terror
plot
of
one
for
the
multiple
upgrade
failures
that
we
had
I'm
going
to
break
them
up
to
them.
A
That's
where
it's
easier.
If
there
are
specific
issues,
so
we
can
we'd,
say
issue
six,
seven,
eight
nine
needs
looking
at
and
any
of
us
can
go,
look
at
the
status
there
and
github
and
understand
who's
been
being
to
hasn't
and
we
can
reach
out
to
additional
people
or
the
same
people
if
they
haven't
responded.
So
that's
where.
A
Helps
right
now,
like
I,
mean
no
workload,
balancing
issue
like
who's
who's
been
talked
to
so
far
on
it.
We
don't
have
any
artifacts
of
the
history
of
the
discussion
and
that's
what
the
the
github
issue
for
that
specific
thing
gives
us.
We've
talked
to
these
people.
They've
assessed
it
to
be
this
person.
We've
talked
to
them,
they
haven't
responded.
It
gives
us
the
history,
so
we
can
follow
up
efficiently.
A
F
So
yeah
for
the
issues
that
are
you
open
them
like,
for
example,
I
know,
th
was
open
for
them
for
that
go
and
make
sure
that
that
issue
is
used
for
following
it
up
like
they
take
in
the
case
of
I,
believe
it
was
just
that
so
for
one
of
the
testers
already
an
issue
open
one
by
a
person
so
I'm
just
using
that
for
tracking
I
guess
I
guess
this
was
for
scalability
storage,
so
felt
the
one
where
the
27,
the
GDC
PVD
thing.
So
there's
Audion
issue
open,
so
I
thought.
F
A
Already
one,
then
we
just
we
need
that
that
linked
and
like
there
that's
where
an
umbrella
issue
helps
like
we
can
say,
here's
our
current
test
failures,
C
sub
issues,
one
two,
three,
four,
five
six,
it
does
point
there's
13
of
them,
so
we
have
or
the
spreadsheet
that's
been
used
in
past
releases
so
that
we
have
an
oversight
of
the
short
list
of
things.
But
then,
from
there
we
can
tunnel
into
detail.
Information
on
on
who's
been
paying
to
hasn't
yeah.
F
D
E
All
right
so
most
of
the
same
bugs
from
last
week,
with
the
exception
of
the
one
I
mentioned
that
was
related
to
test
failures,
that.
E
A
A
A
E
Being
worked
on
and
while
it's
still
open,
I
saw
some
progress.
Yeah
this
one's
2
weeks
old
and
only
recently
got
added
to
the
milestone.
I
haven't
seen
it
yet
I.
Should
it's
it's
a
little
bit
yeah.
There
is
not
really
a
good
search
tool
to
find
issue
and
feature
related
bugs.
As
of
this
point,
I
have
to
sort
of
rely
on
people
labeling
it
correctly
and
that's
why
there's
only
just
now
popped
up
yeah
if.
A
E
E
E
D
F
D
F
D
E
A
That
one
I
find
particularly
interesting
because
one
another
looking
back
to
where
we've
started
the
meeting
like
how
much
do
we
worry
about
gke,
specific
things.
Some
of
that
depends
on
how
much
peer
coverage
we
have
like.
If
we
have
some
a
be
testability
like,
we
can
upgrade
using
cube
ATM
on
the
Google
platform,
but
we
can't
with
our
the
cluster
directories,
scripts.
Okay.
Well
then,
probably
it's
the
cluster
directory
scripts
right
and
rubidium
is
proof
that
it
works
Google
versus
Amazon.
A
We
have
the
test
point
that
there
is
the
set
of
cops
tests,
but
then,
if
those
are
flaky
and
have
historically
been
in
flaky,
we
we
don't
know,
we
don't
have
that
a
be
testability
and
even
then
it's
not
consistent
right.
You'd
want
cube
ATM
on
each
of
them
and
or
cops
and/or,
the
cluster
of
a
cube,
a
diems
of
future
direction.
So
that's
what
we're
focused
on
and
there
we
don't
have
coverage.
So
this
is.
A
This
is
very
much
a
almost
a
kind
of
mind,
reading
or
tarot
cards
or
whatever
the
the
analogy
would
be
to
just
try
to
suss
out
whether
something
matters
or
not.
So
anything
anybody
sees
like
that
like
when
this
is
a
great
example
of
like
hey
how
we
got
to
have
the
conversation
and
have
a
bunch
of
eyes.
Look
at
it
figure
out
how
worried
we
should
be
yeah.
E
Nothing
that
I
can
tell
there's
a
couple
of
them
that
are
getting
closer
and
I
prodded
a
whole
bunch
of
them
me
because
I
didn't
see
any
walk
on
them
since
last
week.
So
come
code
freeze
either
some
people
are
going
to
get
very,
very
hasty
or
very,
very
sad.
A
As
much
as
I've
expressed
some
concern
about
where
we
stand
on
the
features,
I
have
actually
been
pleasantly
surprised
at
seeing
quite
a
bit
of
feature,
work
really
coming
in
and
not
seeing
additional
destabilization
and
tests,
so
I
mean
we're
the
right
things
are
happening
it.
It
can
be
easy
to
get
overly
worried
on
the
release
team
because
you're
focusing
on
all
the
problems,
but
there
is
good
signs
of
positive
work
happening.
Yeah.
E
It's
I
was
just
a
little
bit
concerned
where
my
quick
notes
go
here.
There's
a
couple
of
critical
urgent
ones,
there's
three
of
them
and
I
haven't
really
seen
well,
there's
a
new
one,
but
there's
two
of
them
that
I
haven't
really
seen
any
activity
on
and
it
seems
to
be
pretty
like
on
one
of
them.
People
are
still
just
pushing
the
ball
around
so
I.
Don't
have
high
hopes
for
that.
One
on
another
I
think
there's
a
there's.
Actually
a
couple
open
PRS.
A
The
two
storage
ones,
I
feel,
like
storage,
has
a
very
detailed
list
of
features
that
they're
working
to
land
in
112
and
and
113,
because
the
CSI
container
storage
interface
work
is
aiming
to
hit
their
stable
GA
milestone
in
113.
So
they
they
have
I
feel
like
they've,
got
pretty
good
artifacts
of
working
or
kind
of
steaming
forward
on
things,
but
that
may
also
mean
that
they're
they're
very
busy
and
scattered
and
attention.
If
you
have
something
that's
priority
critical
urgent.
A
But
the
issue
is
open
on
June
5th
that
doesn't
come
across
as
critical
important
and
obviously
that
one
is
an
example
that
didn't
start
out
as
critically
important
back.
Then
it
was
important
soon
and
then
got
elevated,
so
they
are
they're
coming
along
on
their
work
and
of
the
one
of
the
things
that
we've
done
in
the
past
and
I
I
haven't
always
been
fond
of
it
because
I
feel
like
the
data
set,
is
small,
but
we've
tried
to
give
sort
of
a
histogram
with
what
Josh
called
with
the
the
leader
board.
A
The
leader
on
problems
and
which
wit
SIG's,
are
carrying.
The
most
risk
in
terms
of
issue
counts,
and
it
is
informative,
though,
to
see
imbalance
that
verges
versus
feature
work
and
merges,
and
things
happening
so
that
you
get
a
sense
of
how
activist
it
is
or
whether
they're
they're
not
and
six
storage,
at
least
the
activity
levels.
Give
me
a
sense
of
comfort.
Actually
I
would
say
that
they're
they're
on
top
of
stuff
right
now.
G
Well,
I
mean
aside
from
building
software
that
will
poke
them
for
me
repeatedly,
which
probably
would
be
helpful
I'm,
not
actually
all
that
disheartened
at
the
level
of
lack
of
Docs,
open
PRS
and
it
I'm
sure
what
I
was
thinking
it
was
like
I,
probably
could
have
explained
the
process
better
so
that
there
could
be
some
confusion.
Coming
from
that,
but
I
started
a
drive-by
shooting
of
the
Cooper.
Nice
features
repo
and
just
kind
of
basically
spanned
everybody
and
said
all
of
you
were
late
and
so
10
people
kind
of
responded
and
said
yeah.
G
Let's,
let's
work
on
that.
There's
a
really
the
point
of
having
the
docs
open
deadline
is
just
honestly
to
gauge
activity
on
the
feature.
If
somebody
is
prepared
to
open
a
PR
literally
with
nothing
in
it,
just
open
a
PR
on
the
docs
website,
they're
fairly,
confident
that
this
release
it's
going
to
be
coming
in.
So
that's
the
litmus
test
that
this
is
really
used
for
I'm
person,
I
mean
I'll,
start
some
PR
reviews,
but
I'm
not
going
to
look
at
most
of
this
until
two
weeks
from
today
or
tomorrow.
G
When
is
the
is
a
week
from
tomorrow?
Yes,
KO,
trees,
yeah
so
a
week
from
tomorrow's
code
freeze
and
that's
when
Doc's
deadline
is
I,
think
so
so
that's
when
we
really
start
potentially
dropping
features
or
saying
okay.
This
is
super
bad,
so
there
are
just
four
those
were
so
you
know
out
of
64
features
that
we
have
there's
36
that
do
not
have
open
PRS
for
documentation,
which
is
pretty
serious.
Although
36
features
in
a
single
release
is
still
quite
high
or
wait.
No
64
close
enough.
G
A
number
that
is
greater
than
0
features
is
a
decent
release,
so
I
think
we'll
have.
If
I'm
just
kind
of
like
opening
up
a
crystal
ball.
You
mentioned
the
tarot
cards
we
might
as
well
get
into
all
of
the
black
magic
I.
Think
we'll
probably
have
18
that
probably
don't
make
it
based
on
this
list
because
of
Docs
or
other
related
code.
Freeze
like
hey.
Actually
you
know
this
isn't
gonna
work
out.
So
that's
what
I
thought
like
I
said
having
started
reviews
just
so
buggin
people
and
I'm
I'm
decently.
G
A
Her
that
and
I
think
it's
really
useful
to
call
out
like
you.
We
can
make
automation
that
pokes
people
and
we've
had
that
fun
issues
and
PRS,
and
all
this
stuff
and
we've
kind
of
started
walking
away
from
that,
because
that
stuff
is
all
filtered
and
that
that
one-on-one
ping
or
reaching
out
to
people
is
where
we're
getting
better
responses.
My.
G
A
A
A
H
A
Cool,
so
the
next
couple
things
I
wanted
to
talk
about
briefly
some
process
stuff.
So
we
talked
about
removing
or
dropping
the
the
requirement
on
status
approved
for
milestone.
The
lazy
consensus
was
still
later
today
on
that,
but
I
haven't
been
I,
haven't
seen
any
objections,
and
quite
a
few
plus
ones
on
it,
so
that
should
be
going
forward.
I
need
to
scrub
the
github
team.
Still
the
Mount
zone,
maintainer
x'
list
I'd
created
a
spreadsheet,
actually
should
drop
a
link
in
the
document.
I
went
through
and
looked
at
the
team
and
compared
it.
A
Maintainer
x'
alias
in
it
that's
completely
unused,
but
these
two
sets
were
like
hardly
even
overlapping
really
so
people
have
been
confused
on
what
it
takes
to
set
milestone
and
somehow
somehow
things
have
worked,
despite
that
and
I
I
think
that's
actually
proof
that
a
large
part
of
the
community
gets
their
stuff
on
time,
so
they're
not
gated
by
code
freeze
and
then
that
the
sig
leads
where
they
have
stuff
late,
the
sig
leads
or
the
sig
leadership,
whoever
it
is,
whether
it's
a
sig
chair
or
some
higher-level
approver
from
the
group
or
whatever
active
people
are,
are
sufficiently
getting
milestone,
sets
and
not
out
of
trouble.
A
I
feel
like
in
the
past
releases.
We've
only
had
a
few
things
where
somebody
really
really
struggled
figuring
out
how
to
something
set
and
we
needed
to
step
in
tests.
Infra
has
a
little
bit
of
a
fix
up,
so
the
way
that
we
had
done
code
freeze
in
the
past
was
partly
through
the
the
github
milestone
or
not.
A
Yeah
I
get
milestone
under
I,
keep
forgetting
which
one
to
his
name
is
which
so
we
stopped
using
that
Munder
and
tied
is
what's
going
to
be
used
for
code
freeze,
but
we
realized,
as
we
were
talking
about
changing
these
labels,
that
there
was
a
slight
gap
in
what
the
automation
could
actually
do
to
enforce
and
call
his
got
a
PR
going
there
Owen
I
had
I
had
meant
to
leak
that
as
well.
I
will
drop
a
link
in
there
after
the
meeting
yeah.
E
Regarding
the
loss
of
status
labels,
which
will
be
real
in
what
a
few
hours
we're
still
waiting
for
3
p.m.
Pacific,
and
we
you
need
to
update
all
of
the
job
descriptions
for
team
like
release
team
roles,
or
at
least
the
ones
where
those
are
relevant,
so
I
think
the
easiest
way
to
do.
This
is
if
every
team
lead
could
just
issue
a
PR
against
that
and
correcting
yeah.
E
A
E
A
E
I'll
get
that
pin
later
today
after
3
p.m.
Pacific
well,
that'll
be
midnight.
Maybe
tomorrow
exid's.
A
It's
it's
not
a
big
rush,
and
actually
this
is
something
I
had
been
thinking
about
this
weekend
too.
That
I
should
do
today.
It's
kind
of
a
this
is
a
good
time,
so
we're
we're
just
finishing
the
second
month
of
the
three
month
cycle.
Everybody
should
glance
through
their
their
role
handbook.
Give
it
a
read
through.
Is
there
anything
in
it
that
you've
noticed
over
the
last
two
months
is
out
of
date
and
go
ahead
and
drop
a
PR
in
there
cuz
the
next
month?
A
A
So
the
next
one
around
process,
stuff
cherry-picks
traffic's-
have
been
a
mess
for
a
while.
Gwyn
has
been
working
on
cleaning
up
the
developer
guide
with
respect
to
that
and
then
Nikita
also
noticed
some
things
and
put
a
PR
against
test
infra
and
that
was
merged.
But
then
it's
caused
a
follow-on
issue,
so
I've
put
links
there
I
think
everything
unless
I'm
cold.
Do
you
have
any
comments
on
the
needing
to
bump
the
prowl
image
that
had
come
up
and
slack?
But
I?
Don't
know
that?
There's
an
issue
specifically
I.
E
A
I
guess
it's
a
relatively
minor
issue,
just
it's
blocking
cherry-picks
on
past
branches,
but
there
were
I
think
the
the
next
branch
update
that's
scheduled
is
Monday
of
next
week.
So
the
the
branch
monitor
that
or
the
patch
release
manager
probably
probably
wants
to
be
able
to
be
merging
things.
But
it's
not
like
we're
blocking
master
or
112
release
of
the
it's
slightly
lower
pressure.
I
would
say
next
on
the
list.
E
A
E
E
E
F
A
A
E
E
E
A
E
A
I
D
E
A
It's
linked
in
the
the
the
agenda
also
the
devil,
guide,
PR
link
there
here.
B
B
Mean
just
to
give
some
great
context
right,
like
before
I
quit,
devious
employer
to
work
for
a
new
employer.
I
did
try
to
collect
everybody's
gripes
with
cherry-picks
and
put
them
under
one
super
old,
cherry-pick
issue
that
was
called
improvements
to
the
cherry-pick
process,
to
make
it
easier
to
them
for
them
to
track,
or
something
like
that
and
within
that
I
linked
to
a
bunch
of
different
github
issues,
as
well
as
a
mailing
list
that
was
sorted
by
Jordan
Liggett,
to
try
suggesting
how
we
could
try
using
better
labels.
B
How
we
definitely
wanted
to
support
the
use
case
of
like
cherry
picking,
a
single
PR
for
a
master
into
multiple
release
branches.
At
once
things
of
that
nature
there
is
prowl
I
forget
if
it's
an
external
plugin
or
just
a
different
crowd
command,
but
openshift
has
been
using
it
on
their
project
where
they
can
issue
a
slash,
very
technical,
where
their
PRS
and
that
will
automatically
create
a
cherry-pick
pull
request.
B
And
we
could
augment
that,
to
like
add
the
appropriate
shiny
labels
and
have
the
labels
automatically
do
the
things
according
to
our
rules,
processes
and
whatever
that
are
still
in
the
midst
of
being
documented
and
revised,
to
simplify
everything
so
that
all
sounds
cool
I.
Don't
have
time
for
that.
I.
B
Don't
think
she's
on
the
call
here,
but
I
want
to
give
nakita
like
all
the
thumbs
off
I,
possibly
can
for
working
to
move
some
of
the
cherry-picking
logic
over
to
a
month
out
of
munchkin
hub
and
over
to
prowl
and
like,
if
she's
the
person,
to
continue
doing
that
stuff
awesome.
But
we
could
use
somebody
who
has
like
that
level
of
technical
depth
and
you
can
kind
of
stay
committed
to
this
for
a
couple
of
weeks
to
really
solve
all
of
those
problems.
B
A
That's
nicely
and
I
it
reminded
me
of
the
that
there
is
do
you
recall,
is
that
issue
that
you
had
opened
the
umbrella
one?
Is
it
a
testing
for
our
community,
its
intestine?
For
us
this
doesn't
break
it
up
and
drop
it
in,
though
I
can
find
it
too.
It's
it.
I
know,
I've
commented
I,
think
on
it
as
well,
so
I'll
drop
it
in
the
a
link
in
the
agenda.
A
B
One
other
thing
related
to
that.
Like
one
of
the
there's,
there
are
a
few
remaining
munters
in
lunch,
github
that
are
involved
with
running
this
thing
called
the
cherry-pick
you
and
I
want
to
go
talk
to
all
of
our
patch
release,
managers
and
branch
managers
and
ask
if
any
of
them
actually
use
that
thing,
because
I
don't
think
they
do.
If
they
don't,
you
can
go
ahead
and
turn
that
down
and
that's
less
functionality.
B
B
Yeah,
if
you
get
her
like
you,
okay,
if
you
go
to
cherry-pick,
let
me
see
if
it's
one
word
or
two
words:
yeah
like
this
is
probably
another
one
of
those
things
where,
like
Dee
Jerry
do
Pat
release
managers,
even
though
this
thing
exists,
you
get
a
cherry,
pickpockets
thought
I.
Oh,
it's
something!
That's
supposed
to
show
like
pull,
requests
against
release
branches
that
have
a
different
set
of
labels
that
blah
blah
blah
and
if
nobody's
we
shouldn't,
have
it
well.
B
A
That
goes
the
bigger
picture.
We
we've
got
a
smattering
of
tooling
that's
kind
of
incoherent,
not
consistently
used.
If
we're
gonna
clean
up
the
process,
we
should
define
a
good
process
that
works
for
the
people
who
exercise
it
and
then
have
Automation
that
follows
it
like
people
process
automation
in
that
order.
B
100%
I'm
trying
to
take
because
because
of
my
bandwidth
issues,
I'm
trying
to
sort
of
take
the
lower
effort
approach
of
instead
of
reinventing
the
cherry-pick
process
and
putting
the
new
shiny
thing
in
place
right
away.
If
people
aren't
using
the
old,
busted,
undocumented,
broken
thing,
I
think
it's
easy
to
do
that,
but
there's
an
issue
for
that
shiny
new
thing,
I
just
personally
can't
own
it.
I
A
As
much
as
I
say,
people
process
tools,
that's
like
aspirational
right,
I
mean
in
the
meantime,
all
of
these
cleanups
are
awesome
and,
like
you
said,
Nikita,
and
also
go
in
the
stuff
that
you're
doing
this
is
awesome.
Progress,
SoCo,
slash
tomorrow
again
as
a
reminder,
and
then
one
cool
thing
last
week
and
this
week,
Doug
who's,
our
branch
manager
is
on
vacation
and
his
shadows
are
actually
doing
some
of
the
work
for
him
on
doing
branch
fast-forwards,
and
we
found
a
number
of
documentation
issues
and
one
cool
thing
is
well
I'm,
not
sure.
A
If
Jung
Lee
is
on,
we
actually
did
today.
The
branch
fast
forward
from
China
and
one
of
the
things
that
we're
trying
to
figure
out
here
is
how
much
of
the
process
is.
We
get
it
into
the
community
beyond
Google.
Can
we
ensure
that
we're
using
tooling
that
is
exercisable
by
the
broadest
community
that
we're
not
excluding
certain
folks?
So
that's
actually
a
really
cool,
finding
and
and
potentially
one
of
the
betas
in
September.
We
might
try
and
have
young
bill
just
to
see
what
happens.
E
A
A
E
A
That,
because
I
I
was
a
little
feeling
bad
today
that
I
started
the
agenda
like
five
minutes
before
the
call
or
something
I'm
moving
super
slow.
This
Monday,
my
my
niece,
is
hiking.
The
Pacific
Crest,
Trail
and
I
went
and
took
her
food
this
weekend
and
I
hiked
12
miles
or
something
with
a
50-pound
backpack
and
I
barely
made
it
to
my
desk.
E
A
I
was
gonna
wait
till
next
week
to
ask
for
that.
But
actually
this
is
oh.
This
is
one
other
thing
then
so
next
week
Monday
is
a
holiday.
I
will
send
out
a
cancellation,
but
then
starting
this
week,
we're
going
to
the
three
times
a
week
meetings.
So
the
invites
for
that
have
gone
out
and
I
tried
to
pick
a
time
that
splits
the
difference.
I
will
do
my
best
to
have
the
agenda
doc
updated
for
each
day.
J
So
what
I
was
saying
real,
quick
is
I've
dropped
the
link
into
the
chat
the
first
time
we
did
Google
Doc
and
we
had
a
lot
of
trouble
with
last
minute
conversion
back
to
markdown
and
then
last
time
we
did
just
straight
markdown
and
github,
and
that
was
unworkable
because
of
time
issues
and
it
took
too
long,
etc,
etc.
So,
this
time
we're
going
to
split
the
difference
and
do
markdown
in
a
Google
Doc
and
see
if
that
does
any
better,
so.
J
B
I
We
have
not
discussed
it
played
recently
now,
but
I'd
still
like
to
push
for
tied,
yeah
I.
It
should
be
ready
for
us
to
essentially
start
trying
it
out.
The
only
last
change
that
we're
waiting
for
is
the
front
end
change
to
just
display
the
branch
status
and
milestone
status.
The
PR
is
already
out
for
that,
though,
and
we
can
start
testing
out
that
I
can
handle
the
load
and
everything
in
dry
run
mode.
B
I
understand
I,
don't
know
if
today
is
the
right
time
or
forum
to
have
that
discussion,
but
I
want
to
make
sure
if
we're
planning
on
doing
that,
we
have
the
right
stakeholders
involved
to
a
go/no-go
decision
on
that.
The
conservative
part
of
me
wants
to
say
no
part
of
me
that
wants
munch
get
up
to.
You
know
warm
us
by
the
glow
of
the
bonfire.
Like
you
think
that
says
we
should
we
should
move
forward
and
make
sure
that
we
over
communicate.
A
Based
on
my
chats
with
Cole
and
you
I
have
taken
the
impression
that
we
are
going
to
use
tide,
but
we
haven't
given
a
specific
like
to-do
list
there
so
I'm
cold.
Would
you
be
able,
like
by
the
Friday
meeting,
to
give
sort
of
an
assessment
on
like
a
breakout
of
what
you
think
needs
done
at
this
point
and
we
could
review
and
discuss
from
there
yeah.
I
E
A
Gonna
send
out
a
cancellation
for
it
since
we're
meeting
three
times
a
week
right
now,
plus
of
the
alternate
Tuesday
meeting
so
next
week
we
would
still
have
three
meetings
it
would
effectively
be.
Thursday
is
the
only
day
we
don't
have
one
besides
Monday
being
the
holiday,
so,
like
I,
think
I
think
we've
got
pretty
solid
coverage
at
this
point.
Okay,.
E
J
A
So
I
will
send
out
a
cancellation
for
next
Monday's
meetings.
You
should
all
have
seen
invites
for
the
meetings
being
added
on
Wednesday
and
Friday
all
right,
I
think
we
are
definitely
at
the
end
of
the
documented
agenda.
We
still
have
a
couple
minutes
but
ready
to
call
it
a
meeting
going
once
twice.