►
From YouTube: Kubernetes Community Meeting 20170126
Description
We have PUBLIC and RECORDED weekly video meetings every Thursday at 10am US Pacific Time.
https://docs.google.com/document/d/1VQDIAB0OqiSjIHI8AWMvSdceWhnz56jNpZrLs6o7NJY
1.6 features; 1.5 release estimated 2/10; Demo Kubernetes Visual Studio Code integrations; Contribex WG update; SIG Auth update; Sig Apps update; governanace discussion
C
E
A
A
B
F
I
had
a
brief
update
forum
june
from
my
android
device
went
on
the
lips.
I
can
share
the
screen,
so
we
had
a
brief
recap
a
few
minutes
ago.
We
finished
it
with
one
of
these
features.
I
will
start
reviewing
the
features
and
also
tasted
the
lead
to
the
spreadsheet
one
that
succeeds
in
the
communities
community
meeting.
We
have
approximately
40
features
for
this
release
its
target
in
106.
You
can
find
them
all
in
the
spreadsheet
that
reflect
the
actual
station
features.
F
B
You
were
breaking
up
a
little
bit.
There
think
the
bullet
points
are
around
40
issues,
suppose
for
the
16
milestone
the
PM
group
met
before
this
meeting.
We
talked
about
most
of
the
issues
for
cigs
storage
and
we
talked
about
some
of
them
for
vegas
and
sig
scheduling.
We
will
be
doing
another
meeting
monday
at
nine
a.m.
pacific
time
to
talk
about
Austria
life
cycle
and
any
remaining
takes.
A
B
We
are
looking
at
it
now
we
are,
we
have
features
filled.
We
have
release
branch.
Thank
you
again
to
Anthony.
We
have
docs
bugs.
We
are
looking
for
the
a
trusting
person
lead
manual,
upgrade
lead
automatic,
upgrade,
lead
and
secondary
positions
for
all
for
all
roles
and
we're
also
looking
for
a
secondary
release
manager
who
I
guess
is
de
facto
me
or
spod.
B
So
if
you
are
interested
in
joining,
please
email,
kubinashi
skype
in
relief
at
the
google
com
or
crew
benetti's
hack
and
release
at
kora
wescom,
it
will
ping
us
on
slack
if
you,
if
you
would
like
to
volunteer-
and
we
all
the
world,
are
super
important
and
which
we
do
not
fail
them,
it
will
be
much
more
difficult
to
get
the
release
out.
So,
if
you
are
excited
for
release,
we
sell,
especially
with
manual
upgrade
testing,
that's
something
we
always
need
help
and.
B
All
I
want
to
like
black
out
these
phone
numbers,
but
there's
a
there's.
A
spreadsheet
for
volunteering.
I
won't
like
that
includes
custom.
Private
formation,
I
think
it's
already
been
linked,
but
that's
okay,
so
yeah.
So
if
you're,
actually
in
selling
it
as
well,
pls
GG
choked
and
we
will
get
you
set
up
arm
can
go
to
one
of
five.
Sorry
I.
Do
you
hear?
Yes.
G
B
Fantastic
I
did
see
that
there
is
going
to
be
no
one
for
release
updates
this
time
around.
So
we
can
go
into
your
demo
Brennan
if
you
are
ready
to
present
cool.
A
H
H
Was
never
of
in
person
so
on,
and
actually
this
is
great
for
go
honestly
like
the
go.
Laying
integration
is
actually
pretty
sweet,
but
that's
that
so,
but
I've
been
working
on
some
ideas
around
how
we
can
do
better
integration
between
the
development
environment
and
urban
Eddie's
in
particular,
because
I
don't
think
that
it
really
is
a
cloud
development
environment.
That's
particularly
good,
and
so
that's
this
is
that
so
we
started
out.
H
All
right-
and
so
it's
in
the
background,
it's
building
it
up
the
docker
image.
You
can
see
that
there's
the
dockerfile
over
here
it's
pushed
it.
It's
created
a
deployment,
some
sort
of
like
wow,
that's
kind
of
interesting.
What
is
this
deployment
thing
that
I've
heard
of
it's
like
I,
actually
say?
Cooper,
Nettie's,
described
I'm
and
also
deploying
and
node
simple,
which
is
the
deployment
that
it
is
created
and
now
down
here.
H
I've
actually
got
the
output
of
the
described
so
I'm,
integrating
the
experience
there,
but
I
may
actually
want
to
move
beyond
that
and
actually
start
thinking
about
the
crib
entities.
Api
objects
themselves,
so
I
can
actually
a
community's
load
and
I
want
to
load
deployments
new
simple
and
what
this
is
actually
going
to
do.
H
Is
it's
going
to
go
off
and
it
actually
went
to
the
server
pulled
down
the
json
object,
and
now,
in
my
editor
I
have
the
complete
json
object,
but
of
course
this
is
kind
of
complicated,
and
so
I
don't
really
necessarily
understand
exactly
what's
going
on
so
I
can
actually
say
coumarin,
it
is
explained
this
is
integrating
I
just
saved
here.
This
is
integrating
the
explain:
they've
got
its
president
cube
control,
so
we
can
actually
go
take
well.
What
does
this
do
over
here
and
I
hover
over?
H
I
get
a
tooltip
that
shows
me
what
this
field
does
or
the
spec.
If
I
hover
over,
I
get
a
tooltip
that
describes
the
spec
object
to
me,
and
so
now
I
may
want
to
do
something
like
scale.
This
thing
up,
they'll
take
replicas
and
we'll
say
I
want
that
to
be
three
replicas
he'll
say
we
know
you
supply
and
at
this
point
it
kicks
up
a
diff
for
me
right.
So
it's
showing
me
it's
gone
to
the
server
its
pulled
down
the
server
side
version
of
the
object
over
here.
H
It's
got
my
version
of
the
object
over
here,
and
so
it's
showing
me
the
diff
of
what
I'm
actually
going
to
apply
when
I
do
the
supply,
and
so,
if
I
choose
to
apply
this
change,
you
can
say.
Yes,
let
me
apply
that
and
it's
configured
and
it's
free.
It's
updated
that
so
three
replicas
are
starting
up
and
now,
if
I
want
to
understand,
better
I
can
actually
say
cumin.
It
is
get.
A
H
You'll
notice
this
time
around
I
didn't
it
didn't
prompt
me
for
what
objects
I
wanted
to
get,
and
that's
because
I'm
contextualized
here
and
this
object,
and
it
shows
that
and
I
can
see
the
three
replicas
that
are
now
up
and
running
and
I
wanted
to
show
the
one
of
the
other
recent
integrations
with
this
I'll
do
that
to
the
trash
which
is
actually,
we
now
have
live,
debugging
support,
and
so,
if
I
actually
do
Cooper
Nettie's
debug,
it's
actually
building
it's
going
to
build
the
image
again
build
and
push
and
I
say
what
I
want
to
run,
because
obviously
the
command
that
I
give
by
default
in
the
image
isn't
going
to
be
the
command
that
I
need
for
debugging.
H
So
I
can
say
no
debug
start
yes,
okay,
okay
here
and
it's
going
to
wait
for
that
command.
So
now,
if
you
see
in
the
background
here,
I'm
actually
here
on
the
debugger
and
so
I
have
a
service.
That's
up
and
running
that
exposes
this,
and
if
I
hit
reload
over
here
you're
going
to
see.
Well
it
can't
it
can't
contact.
It
doesn't
know
what
about
this
thing
and
that's
because
I
haven't
started
the
program
right,
it's
in
the
debugger,
but
it's
stopped.
H
H
That's
running
out
on
my
crew,
pernetti's
cluster
brought
back
to
to
my
my
debugger
here
and
then
the
final
piece
of
the
integration
that
you
have
here
is:
if
we
go
back,
we
can
actually
also
expose
this
as
a
service.
So
I
can
actually
just
say
internet
is
exposed.
Gonna.
Ask
me:
oh
alright.
I
need
to
reopen
that
this.
A
H
That
deployment
that
I
had
now
fixed,
you
say
suppose
we
go
ahead
and
the
service
has
been
exposed
and
now
there's
actually
a
service
that's
been
defined.
I
can
externalize
that
service.
If
I
want
to,
I
could
pull
down
that
object,
so
this
is
all
out
there.
You
can
actually
go
and
install
this
into
extension.
Now
it's
out
on
github
as
well,
if
you're
interested
in
hacking
on
this
with
me,
that's
about
all
there
is
to
it.
So
thanks!
So
much
can
you
come
over.
B
B
C
Sure
so,
for
contra
backs.
On
monday,
we
launched
the
approvers
mechanism.
I
think
there's
been
an
issue
open,
Brian
went
back
contracted
since
last,
maybe
two
years
ago,
2015
to
get
approvers
working
or
to
expand
the
owners
files
and
with
a
lot
of
help
from
Caleb
from
an
salon
for
Eeyore
all
over
the
place.
We
drove
its
completion.
The
names
I'm
missing,
but
it
was
a
huge
effort.
The
way
that
it
works,
I
think,
is
detailed
in
the
link
in
the
agenda,
so
each
file
in
a
PR
will
need
to
be
approved.
C
Their
approver
spouts
throughout
the
directory.
I've
sent
the
discussion
out.
But
if
you
have
questions
reach
out
to
the
contributor
experience
working
group
or
reach
out
to
me
directly
and
definitely
with
feedback
as
well,
I've
been
getting
feedback
pretty
quickly
and
trying
to
keep
it
really
tight
loop
and
fixing
things
or
updating
features,
as
people
use
it
more
and
more.
So
thanks
everybody
for
that.
C
Now
that
that's
kind
of
released
it's
obviously
we
work
in
progress
with
some
of
the
other
things
that
we'd
like
to
focus
on
is
making
sure
that
it
has
the
intended
effect,
and
one
of
the
metrics
that
we're
looking
to
track
is
how
many
people
are
reviewing
code.
Are
we
expending
the
reviewer
pool
and
overall
we'd
like
to
I,
guess
speed
up
throughput
right?
C
The
overall
goal
is
to
get
you
know,
as
the
number
of
contributors
grow,
to
make
sure
that
the
velocity
of
projects
increasing
as
well
so
there's
things
out
there,
their
number
dashboards
that
exists
already.
So
some
of
the
work
is
going
to
be
around
unifying
the
dashboards
there.
Things
like
governador,
which
can
show
individual
contributors,
how
many
PRS
they
have
open.
A
Devon
spec
just
a
moment
to
approvers
and
owners,
and
all
of
that
tooling.
This
was
a
huge
manual
effort
to
try
to
figure
out
who
the
best
choices
were
for
those
files,
but
that
doesn't
mean
it
is
complete,
so
please
put
prsn.
They
can
then
be
discussed.
So
if
you're
missing
someone
in
an
approvers
list-
or
you
find
it
a
good
list,
it
doesn't
have
enough
people
or
things
aren't
happening.
This
is
an
iterative
process.
This
isn't
canonically
ridden
from
above.
These
are
the
only
people
that
can
be
there.
C
Yeah
exactly
was
a
process
that
was
taking
a
really
long
time,
so
it's
meant
to
be
an
iterative
process.
So
we
got
one
iteration
out.
You
know
add
people
to
the
perverse.
Can
you
see
the
documentation
on
the
approved
list?
I'm,
not
sure
what
you
mean
by
documentation,
there's
a
link
to
the
discussion
and
to
the
email
that
got
sent
out.
If
you
want
to
see
the
approvers,
the
owners
files
itself
I
think
there
was
a
PR
I'll
look
for
or
an
issue
that
was
open.
A
B
B
Yes,
to
advance
to
the
hundreds
question:
there
are
several
threads
in
a
contributor
experience,
but
will
be
try
to
collect
in
there.
For
you,
there
are
lots
of
other
questions.
We
can
go
to
the
beans
go
to
tick
off
and
there
a
representative
from
sig
off.
You
would
like
to
give
an
update
or
what
y'all
workin
on
for
16
yeah.
N
Just
broad
permissions
to
a
few
users,
including
the
service,
account,
if
most
Muslims
do
so
in
16.
What
we're
doing
to
enable
our
back
our
defining
default
roles
for
all
of
the
control,
plane,
components
and
all
of
the
controllers,
as
well
as
for
sort
of
end
user
they're
facing
roles
and
severed
documentation
polls
open
for
that.
N
So
this
is
what
those
roles
will
look
like
if
a
cluster
admin,
which
is
a
super
user
role,
then
admin
edit
and
view
which
are
appropriate
to
give
to
a
user
that
you
want
to
be
scoped
within
a
namespace.
So
an
admin
would
let
a
user
do
stuff
with
a
dining
space,
including
grant
permissions
to
other
users
within
their
name
space.
N
An
edit
user
could
just
have
write
permissions
to
things
like
replica
sets
and
pods,
but
couldn't
actually
manage
permissions
in
the
name
space
and
then
a
view
role
would
have
read-only
permissions,
so
so
we're
working
to
define
these
roles
as
well
as
all
of
the
roles
for
the
various
control
plant
components.
The
second
thing
we're
doing
in
16
is
scale
testing,
so
cube
mark
is
actually
running
with
our
back
on
now
and
we've
gotten
some
good
feedback
there.
N
N
In
3
now
are
a
couple
of
command-line
helpers
for
granting
permissions,
either
with
an
item
space
or
across
all
names
faces.
So
you
can
try
these
out
and
give
me
back,
but
it
makes
it
really
easy
to
say:
I
want
to
give
a
particular
role
to
a
particular
user
in
a
particular
namespace,
and
that
also
works
for
groups
and
service
accounts.
And
then
the
last
thing
we're
doing
in
16
is
working
to
enable
full
are
back
into
various
deployment
message.
So
right
now
it
is
running
in
CI.
N
So
if
you
use
cube
up
on
GC
from
master,
you
get
our
back
cube
Adam
when
deploying
a
16
build
will
enable
our
back
ji.
Ke
is
working
to
make
our
back
possible
to
enable
in
16
and
I.
Think
I
just
saw
something
going
into
cops
as
well,
so
we're
working
to
get
this
enabled
in
all
the
various
deployment
message
we
have
and
then
just
as
a
the
last
note,
because
our
back
policy
is
restrictive
by
default.
N
It
starts
with
a
restrictive
policy
and
then
lets
the
super
user
who
started
up
the
cluster
kind
of
delegate
out
permissions
as
they
want
that
can
be
disruptive
to
workloads
that
just
sort
of
expected
to
have
full
API
access
by
default
without
doing
anything
special.
And
so
we
have
a
couple
upgrade
strategies
that
we've
recommend.
N
The
first
is,
if
you
are
currently
running
15
in
like
a
really
permissive
mode
and
want
to
migrate
to
more
scoped
permissions,
you
can
actually
run
our
back
and
a
back
together
and
our
back
we'll
get
a
first
shot
at
authorizing
and
if
it
denies
it'll
log
and
then
hand
off
to
the
a
doc
authorizer.
So
you
can
keep
your
existing
policy
in
place
and
not
disrupt
anything
running
on
your
cluster
and
just
look
at
your
server
logs
to
see
what's
being
denied.
N
So,
that's
that's
probably
what
we're
going
to
recommend
for
people
there's
always
an
escape
hatch.
Maybe
you
just
don't
care
you're
running
locally
on
your
machine
and
you
really
don't
care.
You
want
to
give
everyone
Commission
to
do
everything,
but
still
easy
to
do.
You
can
grant
the
super
user
role
to
your
user
and
the
cube
lid
and
all
service
accounts
you
can.
You
can
replicate
that
permissive
policy
very
very
easily
if
you
want
to
so
that
is
a
that
is
what
we
have
coming
up
in
16
many
questions.
B
N
B
N
Actually,
local
cluster
up.
What
we
would
do
instead
is
default
to
making
the
user
that
you
have
a
cube
config
for
the
super
user,
so
that
user
would
have
full
permissions
and
they
could
then
either
run
the
command
that
gives
superuser
permissions
to
all
service
accounts
and
do
whatever
or
they
could
carve
up
permissions.
But
we
want
to
kind
of
follow
the
best
practice
of
starting
with,
like
one
user
having
superuser
permissions,
and
then
that
user
decides
what
they
want
their
cluster.
To
look
like
you.
B
N
Up
cluster
Clayton
owns
it,
that's
funny,
it's
that's
a
good
question:
it's
not
exercised
in
neci
flows;
it
basically
gets
issues
reported
against
it
by
people
who
use
it
and
that's
not
a
great
situation.
So,
if
you'd
like
to
help
or
ideally
like
to
see,
if
enabled
in
a
CI
job
yeah
jump
into
some
of
these
the
issues
and
how
about.
B
All
right
which
say
that
cluster
life
cycle
or
cluster
all
I,
like.
D
That
this
is
this
would
be
developers
and
so
I
I,
don't
know
this
contributor
experience,
but
it's
definitely
got
like
the
point
of
this
cluster
is
to
make
local
development
you
see
along
with
mini
tube,
and
we
should
figure
out
the
way.
My
those
detectives
actually.
O
P
P
How
do
we
disseminate
broad
sweeping
bits
of
information
so
that
both
the
community
at
large
is
is
aware
without
what's
been
having
meaningful
type
of
signal
for
millions
ratio,
because
a
lot
of
times
people
will
turn
down
the
developed
or
turn
down
some
other
mailing
list,
because
they're
not
fully
aware
right
or
there's
so
much
noise
in
some
channels,
so
I
just
want
to
make
sure
that
there
isn't
the
12,000
issues
and
WTF
that
go
on
when
sometimes
these
changes
occur.
Yeah.
N
Just
are
you
still
talking
about
local
cluster?
Are
you
talking
about
the
broader
rollout.
N
N
G
One
thing
to
add
to
get
your
general
awareness
issues,
so
it's
big
question
off,
so
we're
trying
to
figure
out
some
sort
of
advisory
system
and
that's
a
work
in
progress
and
that
will
be
to
hopefully
provide
a
meaningful
way
for
people
who
are
learning
this
impression
to
learn
about
breaking
changes
or
things
that
are
impactful
to
production
clusters.
So
if
we
figure
out
a
good
pattern
for
that,
may
be
something
that
we
can
apply.
Another
loans
as
well,
so
things
that
impact
developers
and
smaller
space.
B
Okay,
cool
well
so
there's
like
a
school
so
long
as
that.
There's
a
new
brand
spanking
new
mailing
list
communities
dev,
announce
hyphens
between
those
words
at
googlegroups.com,
presumably
for
DG
dissemination
of
this
kind
of
information
that
should
have
very
high
signal-to-noise
ratio
should
really
only
be
announcements
that
are
of
this
kind
of
nature.
So.
A
B
B
Alcohol
with
that
offline,
well,
we
can
move
on
to
the
would
be
to
meet
for
our
last
30
minutes
here:
p:
governance
at
md
and
three
branch,
the
proposal
and
Bob
the
notes
on
governance.
Afterwards,
awesome.
G
B
Q
Q
Q
In
progress,
we
have
some
working
on
staple
set,
bug
fixes
and
generally
auditing
the
code
in
found
some
corner
cases
will
be
fixed
in
16
like,
for
example,
if
you
have
two
failures
in
a
five
instance
staple
set,
then
it
can
get
blocked
from
reckon
filing
vs.
We
would
have
worked
with.
There
was
just
one
failure:
we're
moving
all
the
annotations
that
are
needed
by
staples.
Various
people
are
moving
them
to
beta
and
the
containers,
a
ga
which
makes
it
much
prettier
to
write
a
stateful
set.
Q
Several
people
are
working
on
design
and
implementation
for
Damon,
set
up
dates
and
I
expect
that
in
1.6
you'll
be
able
to
update
a
daemon
set
for
Damon
tests
that
don't
require
machine
drains
that
art
can
tolerate
a
priest
down
time.
There's
a
number
of
deployment
improvements
under
review
or
recently
completed
across
all
controllers
were
improving,
making
sure
we
use
controller
F
and
have
sane
adoption
logic,
and
we
avoid
overlapping
controller
problems.
A
B
A
I
E
A
E
E
E
E
There's
the
council
of
elders,
which
I
understand
that
is
still
in
debate,
and
I
think
there
is
a
view
that
that
is
all
sort
of
sort
of
like
a
pnc
jerkin
different.
I
would
hope
that
that
is
a
bit
more
active
in
terms
of
determining
a
collector
and
direction
really
is
a
normal
corporate
matter
that
would
be
CTO
design
architecture
and
I
think,
having
this
converges
really
critical
projects.
E
This
is
just
a
well
known
patterns
is
offered
hearing
of
teams
bare
feet,
oriented
and
to
build
fragile
architectures
to
guess
those
types
of
teams
that
are
exclusively
our
connections.
Nine
areas
tend
to
produce
features.
The
customers
don't
want
enough.
We'll
use
your
experiences
I
think,
especially
for
a
large
open
source
project,
lots
of
parties.
It
is
through
the
there's,
a
third
a
third
license
school
another
license.
Poulos
profits
policy
really
shouldn't
be
involved
in
terms
of
its
decisions.
E
That
really
should
be
fundamentally
empowered
to
set
the
rules
around
transparency,
ml
and
worry
about
your
organization,
which
at
present
that
decision
making
around
that.
If
there
is
not
some
of
the
stories
per
diem
group,
someone
is
coming
from
the
developer.
Experience
group
like
we
go
online,
I
think
in
the
in
the
large
nonprofit
world.
The
notion
of
equality,
for
short,
is
pretty
well
understood,
best
practice.
E
E
It
as
an
extension
from
the
capital
dollars
I
think
they've
been
a
bit
of
crashing
around
on
project
ecology
issues.
Rapid
Cheney
not
can
be
really
necessary.
Everyone
has
an
awesome.
You
read
I'm,
really
down
amazing
around
about
to
be
around
the
super,
slow
and
then
be
treated
like
a
purse
office.
Q
Q
H
That
that
can
you
hear
me,
okay,
I,
think
that
that
is
where
the
community
writ
large
comes
to
play
right
I
mean,
I
think
that
this
is
at
the
end
of
the
day.
This
is
a
democracy
and
if
we
make
enough
people,
mad
they're
just
going
to
fork
and
go
someplace
else
or
they're
going
to
stop
contributing
right.
H
So
I
think
that
there's
a
natural
counterbalance
to
all
of
this,
which
is
that
everybody
is
here
more
or
less
voluntarily,
and-
and
so
I
don't
see
that
as
being
a
concern
Eric
because
I
don't
see,
I
see
this
being
more
about
the
project
wide
policy,
not
the
meta
policy.
If
you
will
right,
not
the
policy
for
the
policy
yeah.
I
I
H
There's
also
a
actually
just
to
comment
on
that
quickly,
like
I.
Actually,
don't
think
that
except
different
governance
is
sufficient
here.
I
know
I
want
to
talk
to
so,
but
I
don't
think
this
stuff
is
I.
Think
everness
here
is
sufficient,
because
it
is
very
kind
of
I
would
say
tactical
at
some
level
it
doesn't
have
a
lot
to
do.
It
has
a
lot
to
do
with
how
you
do
an
individual
thing
or
an
individual
piece
and
various
rotations
and
stuff
like
that,
there's
very
little
of
the
howdy.
H
O
Matt
an
issue
that
I
want
to
bring
up
here,
real,
quick
and
then
I-
think
Craig
has
something
to
arm.
What
is
the
project
here,
because
I
think
one
of
the
things
that
we've
talked
about
again
and
again
is
actually
scoping
down.
The
core
could
be
something
that's
more
tractable
and
essentially
done
right.
If
we
can
do
that,
if
we
can
make
the
scope
of
this
for
the
decisions
that
need
to
be
made
smaller
and
it's
going
to
be
easier
to
set
erase
this
stuff
out
and
actually
have
people
only
across
the
different
components.
This.
O
Know
Clayton
there's
many
some
stuff,
that's
going
to
actually
cut
across
the
larger
project
and
then
there's
gonna
be
some
stuff.
That's
concentrated
on
sort
of
like
sub-project,
and
so
maybe,
if
we
can,
you
know
if
we
can
find
ways
to
to
sort
of
scope.
This
stuff
down
it'll
be
easier
to
to
not
need
these
sort
of
wide.
Think
now,
I'm
not
saying
we
shouldn't
do
it,
but
I
think
maybe
that'll
give
give
some
folks
some
confidence
on
this
side
so
110
station
here
I.
K
K
You
know
some
of
the
shortcomings
and,
if
there's
a
lot
of
terms
of
one
the
thing
that
I
would
implore
quick
to
think
about
us,
because
it's
so
cool
the
definition
of
this
be
possible
is
rather
than
you
know,
trying
to
sort
of
a
priori,
establish
it
as
like.
You
know,
hey,
you
know
we
interested
before
they
get
any
polish.
Their
Charter
is
to
go,
and
we
think
the
way
that
everything
works.
It's
too
upsetting,
really
really
small,
really
short
sound
like
like.
K
Is
it
like
a
small
white
window
on
it
and
see
if
we
can
do
something
that
just
creates
incremental
value
first
as
a
sort
of
dating
process
before
we
try
to
get
married
with
a
new
way
to
think
about
wall
of
governance,
etc?
So
it
is
something
within
this
project
that
we
can
show
up.
As
a
discrete
like
this
is
an
obvious
endpoint.
K
Let's
try
to
get
a
really
elegant,
simple.
You
know,
system
in
place
to
define
quality
ron.
Klain
point
will
give
us
a
step
forward.
Then
you
know
I,
agree
that
you
know
long-term.
We
figure
out
like
much
much
cooler.
Things
like
you
know.
Is
there
a
choice?
Is
there
a
telescope?
Do
we?
We
have
that
sort
of
a
pointed
or
suppose
did
like
how
this
all
sort
of
work,
but
I
just
don't
want
to
get
too
far
ahead
of
ourselves.
K
M
K
K
A
E
E
That
hot
stuff
going
on
and
running
around
the
project
right
now,
this
is
being
made
enough
in
a
less
than
transparent
fashion,
changes
that
are
affecting
the
entire
project
being
rolled
out
without
really
transparency
or
copper
nodes,
and
this
is
really
bad
for
morale.
So
what
I'm,
really
trying
to
do
is
just
put
a
little
bit
of
structure
around
things
that
are
going
on
the
project
right
now,
if
I
were
to
make
a
remember
to
make
a
proposal
an
incremental
proposal
here,
it
would
be
that
I
mean
you
can
implement
it
like
that.
E
E
L
It
to
comment
on
it
actually
I
agree
with
what
Craig
said
and
because
I'm
the
Senate
skip
this
about
transparency
and
communication.
But
you
also
need
to
acknowledge
reality
is
doing
those
decision
and
defining
those
policies
is
really
time
consuming
right
and
if
you
will
look
on,
for
example,
the
complete
experience
working
group,
this
group
exists
for
a
very
long
time
and
not
many
volunteers
to
be
fun
with
it.
I
think
this
is
how
the
committee
tried
to
make
it
transparent
right.
L
So,
if
one
wants
to
know
what
will
the
processes
also
impact
00
touch
anything
that
will
be
the
place
and
I
am
also
concerned
about
creating
a
teams
in
a
position
that
people
that
are
not
in
that
close
group
will
know
how
to
impact
right.
So
talk
about
concurrent
its
transparency
for
the
entire
community
test,
one
and
also
from
what
I've
seen
even
with
the
elders,
all
those
big
things
and
end
governance.
L
If
you
want
to
build,
will
require
a
lot
of
time
from
people
and
I
would
prefer
to
try
it
first
in
a
smaller
policy
creation
and
serious
people
have
the
time
and
the
attention
to
give
it,
because
otherwise
there
would
be
a
policy
or
governance
board
or
whatever,
but
nothing
will
actually
come
out
of
it,
and
that
would
be
my
concern
ago.
I
really
want
to
share
something
cool
I.
H
Think
there's
a
trail
challenge
here.
Is
that
two
things
one
is
that
we
continue
to
make
these
decisions
right.
We
decided
to
add
basil,
which,
basically,
as
far
as
I,
can
tell
just
adds
a
step
to
my
pre-commit
scripts,
we're
deciding
to
break
up
the
source
code
repository
which
has
we've
had
a
big
discussion
about,
and
we
never
reached
a
conclusion
on
right.
We
are.
H
We
make
a
lot
of
these
decisions
in
ad
hoc
ways,
with
very
poor
sight
lines
and
very
bad
accountability,
and-
and
I
think
that
the
challenges
we
try,
we
did,
we
made
it
all
distributed
right.
We
made
a
contributor
working
group
and
we
made
all
of
these
different
little
sub
working
groups
which
works
in
those
scopes
when
the
scope
of
those
groups
is
clearly
understood.
There
are
these
things
that
crosscut
and
then
what
happens?
Is
these
cross-cutting
decisions
are
made
in
these
small
groups
with
relatively
poor
visibility
unless
you're
tightly
in
that
group
I?
H
Think
that's
the
concern.
The
problems
expressing
I
think
Craig's
proposal
to
practice
is
a
decent
one.
You
may
as
well,
but
I
worry
if
it
will
get
in
back.
In
that
same
loop
of
making
one
small
decision
without
getting
a
broader
process
in
place
to
address
all
of
the
future
decisions
that
we
might
make
sure.
K
You're
right
on
me,
the
thing
I
would
argue
is
like
let's
find
a
hot
spot
right
now
as
a
community,
and
if
it's
like
the
you
know,
if
it's
you
know,
I
don't
know
where,
like
what
the
most
impact
like
the
most
long
lead,
impacting
decisions
of
being
made
like
you
have,
people
are
actually
doing
charcoal
network
and
let's
focus
on
like
this.
Well,
let's
have
people
fit
into
that
sig
and
actually
start
to
establish
this
reticence
right
so
we'll
make
a
decision.
We
need
to
get
into
habits.
Metacity
the
harlot
decision
actually
communicates.
K
Well,
we
probably
need
to
have
a
bunch
of
existing
communication
mechanisms.
Let's
try
using
them.
You
know,
let's
make
sure
that
it's
documented
like
way
to
document.
Well,
it's
just
write
it
down.
You
know,
like
you
know
we
can
specify,
could
try
to
create
more
process
more
structure,
but
at
some
point,
if
you're
not
already
adherence
the
person
to
the
obvious
prosthetics
is
like
you
just
you
sort
of
spiraled
out
of
control.
So
my
belief
is
that
we
should
identify
a
hot
spot.
Just
one
much
politically.
K
Let's
say,
like
you
know,
we
as
a
group
are
no
longer
tolerant
of
you
know,
making
body
block
whatever
it
is
right
and
then
there's
like
focus
on
that
and
then
she's
like
teen
house,
it's
dead
in
it.
Let's
make
sure
that
was
the
organizational
structure
make
sense.
Make
sure
that
we
communicate
to
fix
that
the
scope
is
appropriate
and
that
was
the
best
practices
they
need
a
worker
one.
If
it
works.
That's
great,
you
think
it's
another
thing.
I
knew
the
best
practice
we
establish
mistake.
You
know
it
worked
very
well.
What
works?
K
K
K
K
H
Want
to
say
it
it
to
critic
I'm
totally
in
favor
and
I.
Think
that,
but
I
think
we
need
to
make
sure
that
is
something
it
is
a
decision
that
is
above
the
level
of
the
sig,
because
I
think
that
the
sig
stuff
actually
is
working
pretty
well,
I
think
it's
the
stuff
where
it's
like
this
is
community-wide.
H
You
know
this
impact,
everyone
where
we
do
a
poor
job
and
I
so
I
think
if
we'd
identify
one
of
those
and
we
go
through
it
and
we
survey
effectively
at
the
end
like
I'll
happy.
Are
you
as
a
community
with
this
process?
That'll
be
a
great
signal
that
we're
doing
it
right
and
then
so
I
think
that's
a
great
way
to
do
it,
but
we
should
make
sure
that
it's,
not
small
tactical.
It's
it's.
H
A
Think
I'm
some
stuff
for
a
second
and
jumping.
So
as
an
example
where
we
are
having
problems
is
not
this,
we've
done
very
well
with
subsidiarity
if
it
is
scoped
within
your
sig,
figs
are
making
decisions.
That
is
that
is
working
really
cleanly.
Taking
anything
that
we
want
to
change
across
multiple
cigs
or
across
the
whole
project
is
to
pain
point.
They
don't
have
small
small
things
so
much
that
we
contain,
although
to
be
fair,
Erin's
keeps
bringing
up
and
is
awesome,
I
Bible.
The
way
we
implemented,
two-factor
authentication
on
github
worked
really
well.
A
We
talked
about
it
in
the
community
MTG
multiple
weeks,
not
necessarily
discussing
it,
but
making
it
as
a
notice.
It's
coming.
It's
coming.
It's
coming
it's
coming!
For
these
reasons
it
was
documented,
it
was
been
to
the
community
said
mailing
list.
It
was
reminded
that
it
was
going
to
come
in
three
weeks.
Here
is
then
everyone
who
was
going
to
be
affected
immediately
because
they
hadn't
put
on
to
essay,
was
then
contacted
I
think
three
times.
Thank
you.
Thank
you.
A
Thank
you,
Eric
Paris,
for
the
pain
on
that,
and
then
we
still
screwed
up
in
a
slight
way
in
that
we
didn't
notify
all
the
people
into
pernetti's
incubators
that
we're
going
to
be
affected.
So,
but
we've
learned
from
this,
you
know
so
I
think
taking
those
and
trying
to
figure
out
what
the
best
practices
are
around
establishing
a
new
process
which
is
slightly
different,
cuz
I.
Don't
think
anybody
argued
we
didn't
want
to
SA
and
that's
the
difference
here
is
why
that
one
worked
for
me.
Why
that
would
work?
J
I
I'm
really
sorry.
I
have
to
go,
give
a
talk
right
now,
so
I'm
just
going
to
finish
this
I
think
actually
Brendan
if
I
can
propose.
Brendan
and
Craig
are
actually
saying
the
exact
same
thing,
I'm
slightly
different
way:
Craig
is
I
I.
In
my
opinion,
my
vote
is
closer
to
Craig,
like
I,
am
worried
about
instituting
process
without
like
at
least
betting.
It
some
way,
but
Brandon's
point
is
extremely
valid,
like
solving
this
at
a
cig
level
is
not
the
problem
we're
trying
to
solve.
J
We
are
trying
to
solve,
choose
two
or
more
cigs
who
may
have
somewhat
conflict
issues
and
need
a
third
authority
figure
vote
whatever
to
help
resolve
that
issue,
and
so
all
I
would
propose
is
I
think
we
could
maybe
solve
this
by
picking
two
cigs
that
do
have
some
degree
of
overlap
and
and
attempting
a
proposal
rehmat
and
then
I'll
take
off
thinking.
Hexagon
is
making.
E
Most
of
the
greatest
example,
I
can
think
of
this
kind
of
issue
here.
I,
don't
really
think
this
is
crossing
kinds
of
things
is
more
practical
I'd
so,
for
example,
coming
down
the
cross
coming
up
to
this
discussion
around
government.
Stop
we
had
a
debate
as
debate
was
around
and
by
the
way,
this
pattern
is
just
one
example.
E
Adam
was
we're
going
to
put
a
hold
of
chess
discussion
and
content
into
my
words,
google,
to
google
plot
and
a
whole
bunch
of
the
conversation
around
the
community
is
going
to
be
restricted
to
a
subset
of
the
community
around
doc
and
then
eventually
I'm
going
to
get
more
base.
It's
going
to
get
into
the
PRS
actually
in
the
open
for
a
wider
for
a
community
Bob.
E
Well,
I
think
it
happened
in
other
cases.
You
know,
maybe
that
was
a
poor
example,
but
we
didn't
have
a
debate
around
paying
like
the
standard
way
to
do.
This
kind
of
thing
on
the
project
is
to
put
in
this
er.
There's
much
debate
around.
You
know.
Well
the
features
kind
of
suck
all
the
debate
about.
Can
you
reckon
you
not
actually
see
the
private
histories
or
really
providing
transparency
around
common
stream?
And
I'm
really
not
brown
rice
is
exactly
the
point.
E
O
Everybody's
doing
it
the
same
way.
Okay,
so
so
I
mean
sorry
here.
So
I
think
you
know
two
things
real
quick
number
one
Google
people
need
to
start
creating
google
docs
with
an
email
account
now
with
the
google
account,
because
this
I
sure
to
burn
a
DS
tab,
and
then
you
open
it
with
the
wrong
account
is
really
annoying.
If
the
public
is
public,
we
need
to
start
doing.
That.
Second
thing
is
that
there
are
some
things
that
are
just
by
their
nature
towards
doing
a
project
wide
things
and
I
think
I'm.
O
Thinking
like
a
TI
machinery,
it's
just
very
difficult
for
API
machinery
to
do
anything
that
doesn't
impact
a
ton
of
people.
I
think
we
need
to
be
very
aware
of
the
fact
that
that
you
know
those
things
are
by
necessity,
going
to
have
to
actually
move
slower
and
get
more
buy-in
because
they
do
actually
impact
so
many
more
people.
We
can't
turn
those
big
in
terms
of
sort
of
like
cabal's
that
get
to
make
things
you
can
make
decisions
unilaterally,
and
so
it's
just
something
and
that's
not
necessarily
cross
big.
B
K
B
So
I
try
to
stifle
edius
discussion
here,
but
we
are
bumping
up
against
the
end
of
this
meeting
arm.
It
sounds
like
the
things
are
still
yet
unresolved,
so
I
will
propose
talking
about
this
in
a
possibly
age
a
focused
meeting
or
if
we
could
break
it
out
of
a
topic
out
of
working
with
the
contributor
experience
working
late.
If
that
sounds
good
for
people,
okay,
we've
got
a
one
boat
for
a
breakout.
Another
vote
for
breakout,
okay,
great
well,
no.
A
B
Sounds
good,
so
do
we
want
to
Center
the
next
item
setting
something
to
our
new
giving
probably
a
day
or
so
people
to
join
to
baileys
dev
hyphen
announced
and
then,
by
monday
or
tuesday
morning,
pacific
time
I'm
sending
out
a
invite
to
a
separate
breakout
meeting?
Can
we
have
a
volunteer
to
loan
debt
I.