►
From YouTube: 20190507 sig cluster lifecycle
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
Hello
today
is
Tuesday
May
7th
2019.
This
is
the
standard
state
clustered
lifecycle
larger
meeting
before
we
go
into
sort
of
status,
updates
or
feedback
from
the
broader
sub
projects.
I
wanted
to
have
a
brief
discussion
about
coop
comp
is
coming
up
pretty
fast.
First
question,
which
should
be
like
a
non-starter
I
think
would
be
our
folks.
Okay
with
me,
canceling
the
next
meeting,
which
it
which
is
right
during
the
keuken
timeframe,
it's
up
from
Justin's.
A
All
right,
I'm,
just
gonna,
nod
my
head
and
unless
there's
dissent,
I'm
gonna
go
ahead
and
cancel
that
meeting.
The
second
topic,
which
I
discussed
a
couple
of
times
before
we
get
to
the
status
updates,
was
topics
of
discussion
for
coop
con.
Are
there
things
that
folks
would
like
to
have
sort
of
arranged
for
us
to
discuss
if
there's
times
that
folks,
my
feedback
on
before
we
arrive
at
guca
I.
B
A
A
So
fabbrizio
took
that
intro
session
I
did
is
sort
of
broke
it
down
into
slides
sort
of
to
give
a
better
view
of
the
world.
So
there's
one
other
slide,
that's
precursor
that
which
actually
gives
you
a
better
idea.
So
there's
the
provisioners,
which
is
like
cops,
do
porn
other
ones
than
there's.
You
know
it's
an
EDM
idiom
cube
aiding
in
cluster,
add
ons
component
of
a
and
cluster
API.
A
A
A
B
Know
platform
9
has
a
piece
called
node
idiom,
which
is
I,
think
I,
guess
that
would
be
sort
of
part
of
the
image
Stamper
project
which
is
like
you
know
how
the
question
which
image
stamping
is
sort
of
answering
is:
how
do
we
get
an
image?
How
do
we
get
the
from
BIOS,
I
guess
to
cuvee
DM
or
to
xev
idea
right,
I?
Think
in
your
stomach
is
one
way
and
no
did
we
miss
another
way.
So
I
guess
that's.
A
Well,
I
think
that
could
probably
see
the
conversation.
What
I'll
probably
do
is
start
with
this
and
then
talk
up
and
then
have
to
talk
about
some
of
the
other
key
pieces
to
at
good
con
and
see
where
the
discussion
goes.
I'll
try
to
create
a
list
of
salient
points
or
issues.
One
of
the
things
I
think
is
worthy
to
denote.
C
A
C
B
I'm
here
I'm
is
here,
but
we
Haley,
we
did
move
the
the
slots
into
the
alternate
slot
to
sorry
this
time
slot
in
alternate
weeks,
and
we
had
a
meeting
in
that
slot.
Last
Tuesday
the
we
are
working
on
a
for
example,
operators
showing
the
various
approaches
out
there
and
last
week,
I
think
I
presented
one
using
the
sort
of
coop
builder
and
add-ons
and
declarative
pattern
and
Lee
put
up
one
this
sometime
this
week.
I,
don't
remember
exactly
when,
but
after
that
meeting
using
the
operator
CK.
B
B
Everyone
has
identified,
are
back
permissions
as
a
challenge
of
the
operator
model
when
installing
privileged
pieces,
because
the
problem
is
that
to
create
an
are
back
role.
You
the
operator,
therefore
needs
all
the
permissions
of
the
are
back
role
that
is
creating,
and
you
end
up
with
these
operators
with
way
more
permissions
than
they
really
directly
need.
B
It's
not
wrong.
It's
just
sort
of
a
little
bit.
Annoying
and
we're
trying
to
think
of
ways
around
that
but
yeah
anyway,
everyone
has
any
other
approaches
that
they
would
like
to
work
a
proposed
for
add-on
operators,
please
to
send
a
PR
to
the
repo
which
is
community,
SIG's,
slash,
cluster
add-on
operators.
We
have
one
for
the
keyboard
pattern
and
for
the
operator,
STK
pad
Thank,
You,
Lee
and
we'd
love
to
discuss
any
other
options,
we're
so
trying
to
look
at
all
of
them
in
a
very
concrete
way
and
be
sort
of
systematic
about
it.
B
B
Think
so
I
think
that
would
be
great.
We
should
we
should
have
more
consensus,
I
hope.
After
next
week's
meeting,
when
we
talk
about
lease
operator,
SDK
implementation
and
yeah,
then
we
should
just.
We
should
definitely
discuss
that
further
at
cube
con
over
green
cities
cuz
it
formally
because
I
don't
know
how
to
people
will
be
there,
but
we
can
certainly
try
to
get
more.
F
This
update,
it's
a
pretty
brief
one
for
the
past
two
weeks
since
we've
been
working
on
some
other
areas,
we
still
need
to
solidify
the
API
machinery
primitives
for
encoding
and
decoding.
Some
of
the
improvements
that
are
necessary
for
a
component
config
are
still
open
and
need
to
be
rebased.
The
coop
controller
manager
did
not
have
a
cap
that
made
115
for
component
config,
so
that's
d
scoped
for
this
particular
release,
but
Stewart
messaged
me
this
morning
and
he's
working
on
a
ket
for
that,
obviously
any
help
and
feedback
there.
F
It
always
you
know,
moves
that
stuff
through
the
process,
much
quicker.
So
if
anyone
in
these
looking
and
solving
some
thorny
problems
with
nested
config
and
just
in
general,
dealing
with
the
complexity
of
lots
of
controllers
being
hosted
by
single
binary
and
how
to
configure
that,
please
find
that
in
the
enhancements
repo,
we
already
have
an
open
issue
with
lots
of
comments
there
and
then
yeah.
F
We
had
a
discussion
this
morning,
Luba
married
already
mentioned
previously
on
his
call,
but
he's
gonna
be
starting
some
work
on
a
flag,
override
proposal
to
trying
and
get
some
structure
around,
potentially
solving
overriding
config
values,
using
the
command
line
in
a
generic
way.
That
is
sensible
for
users
and
something
we
can
support.
A
F
One
we
do
a
call
every
week
in
the
time
slot
directly
before
this.
It's
typically
about
30
minutes
and
it's
pretty
intimate,
and
we
could
definitely
you
know,
find
something
that
you'd
be
interested
in
working
on
we'd
love
to
help
some
new
contributors
ramp
up
into
dealing
with
some
of
these
problems.
There
are
certainly
lots
of
learning
opportunities
where
we're
all
diving
into
parts
of
the
code
that
we've
never
read
before
so
I.
A
It's
I
think
it's
fundamental.
It's
it's
the
long,
it's
something
that's
long
overdue!
So
if
anyone's
ever
actually
looked
at
the
sheer
number
of
knobs
for
the
API
server
or
the
Google
it
it's
it's
north
of
200
and
that's
that's
an
insane
thing
to
think
about
all
right.
Next
up
are
there
any
questions
for
component
Vic
I'll,
go
slower
good
once
twice
three.
G
You
this
will
unlock
other
like
issues
that
we
have
seen
doing
some
demos
closer
API.
The
best
provider
has
been
released
so
there's
it
0.21
release
which
kind
of
tests
and
bringing
brings
in
more
reliable
creation
for
multi
controlling
clusters
and,
like
it,
also
fixed
an
issue
visit
issue.
We
did
a
blessed
recession,
which
was
not
being
reused,
which
caused
some
pain
points
with
like
machine
deployments
with
more
than
like
two
or
three
nodes.
G
G
Need
to
move
relief,
a
new
one.
Yes,
thank
you
for
the
work
streams
we
have
like
had
like
cannot
be
making
some
progress
on,
like
all
the
work
stream
across
the
work
for
the
data
model.
The
goal
is
to
have
this
week
tomorrow,
I
believe
like
9:00
a.m.
there
is
another
meeting,
so
we
will
go
over
the
proposal,
that's
linked
in
the
doc
and
kind
of
functionality
and.
G
E
G
So,
for
any
kind
of
like
a
more
like
in-depth
proposals
like
with
the
follow
the
new
cap
proposal
process,
I'm,
not
sure
if
this
should
go
in
the
extension
mechanism
itself.
To
be
honest,
the
extension
mechanism
work
stream
goal
is
to
define
what
the
extension
mechanism
is
and
not
in
like
an
unnecessarily
like
where
the
code
lives
or,
like
anything
related
to.
E
Extension
mechanism
because
I
don't
think,
there's
a
need
for
an
extension
mechanism,
because
the
controller
is
relatively
lightweight
from
the
machine
side,
and
so
it
doesn't
make
sense
to
me
to
have
machine
controller
library
that
does
nothing
doing
reconciling
to
just
calling
endpoints
of
whatever
kind
see
it's
not
really
a
controller.
It's
not
doing
anything
functionally
and
if
we
went
with
the
library
model
similar
to
what
we
have
today,
where
you
have
a
reconciled
loop
and
then
you
call,
you
know
various
functions
that
the
individual
actuator
implements.
E
A
Want
a
table
of
this
discussion
right
now,
Mike
like
but
Michael
I'd
like
to
table
this
discussion
right
now,
because
it's
a
very
campy
specific.
This
is
the
broader
signatur
subprojects
report
out.
I.
Think
it's
worthwhile
to
maybe
have
as
a
discussion
point
at
the
end
of
this
meeting,
but
I
want
to
make
sure
we
don't.
A
So
you
can
bring
that
up
with
the
sub-project
donors
for
Cappy
and
I.
Think
that's
fair
and
you
know
we're
here
on
the
call
too
I
think
the
purpose
of
trying
to
have
separate
meetings.
Two
breakout
sessions
was
not
to
conflate
the
original
meeting.
I
do
think
that
they
need
to
be
collapsed.
It
was
basically
to
have
breakout
sessions
to
have
formulated
discussions
and
individual
topic
areas.
E
G
This
is
where,
like
the
proposals,
come
in,
you
like
you're,
free
to
write
your
own
proposal
or
add
your
ideas
to
the
propose
that
we
have
today
we're
trying
to
push
work
forward
that
we're
not
trying
to
like
kind
of
close
out
any
other
ideas.
The
repulsive
can
be
rejected
or
cannot
after
review,
but
if
they
mentioned,
we
should
table
this
discussion
for
tomorrow's
meeting.
Yeah.
A
So
you
have
every
time
along
the
way
to
to
raise
an
issue
or
to
talk
about
it
or
to
come
up
with
a
counter
proposal
to,
and
the
purpose
of
the
sub
project
is
to
determine
the
weight
of
the
larger
group
within
that
group
to
weigh
the
options
that
exist
there
and
I
think
they've
been
fair
right
to.
In
my
opinion,
I
think
it's
been
a
lot
of
meetings,
but
at
the
same
time
the
goal
is
to
try
and
get
broader
engagement
and
focused
discussion
of
given
topic
areas.
Now.
Could
we
refine
the
process?
A
Sure
I
think
the
the
intense
was
you
know
was
in
the
right
place,
but
the
goal
of
those
meetings
is
to
collapse
eventually
to
come
out
with
a
proposal,
and
then
the
meetings
are
done
right.
There
should
not
be
an
ongoing
set
of
work
streams
afterwards
and
Oaks
have
counter
proposals.
You
know
I
recommend
talking
with
those
people
who
are
creating
the
proposals
now.
I
I
Some
of
the
decision
I
could
try
just
assume
implicitly
that
there
are
work
hook
and
what
are
the
proportion
we
trying
to
move
forward
is
to
decouple
that,
because
I
would
like
to
propose
Controller,
Bay
extension
mechanism,
but
that
deploy
zone
will
be
easier
to
the
main.
Is
some
data
model
supports
that
or
not
so
I?
Think
there
is
a
kind
of
I
understand,
EDT
separation,
but
at
the
same
time
is
like.
If
you
separate
data,
then
the
data
model
should
be
agnostic
of
the
essential
mechanisms.
So
how
III
find.
A
Believe
that
there's
overlap,
I,
think
there's.
These
proposals
are
not
are
the
people
that
are
working
in
the
proposals
they're
not
set
in
stone
until
we
actually
have
the
conversation
right
and
what
what
I'm
Maya
my
hope
is
not
I'm
not
running
the
show
here
in
that
particular
one
I'm
kind
of
passively
observing,
but
you
know
I'll
help
step
in
where
people
feel
like
they
need
me
to
step
in.
Is
that
there's
going
to
be
a
long
time
before
or
not
a
long
time?
E
Somewhat
dismissive
so
I
think
this
project
needs
to
focus
on
is
because
everything
stems
from
there.
It
doesn't
make
any
sense
to
discuss
extensions
and
all
these
other
things
until
the
data
model
is
locked
down
now,
in
the
context
of
having
meetings
that
specifically
discuss
that
individual
point
I
think
that's.
A
A
There's
gonna
be
rationalization
that
has
to
occur
and
that's
totally
fair
right,
but
I
think
right
now
too,
it's
not
fair
for
the
other
people
in
this
call
or
have
broader
updates
for
the
entire
set
of
sub
projects
that
are
not
just
can't
be
related
for
us
for
them
to
talk
about
their
their
particular
issues.
Now,
if
you
want
to
bring
it
up
at
the
end,
let's
go
for
it,
let's
kind
of
burn
through
the
rest
of
the
agenda
and
discuss
this
topic,
but
we
should
also
bring
this
up
in
the
kepi
meeting.
F
Yes,
but
the
portion
of
the
call
that
we're
going
in
right
now
is
to
give
people
who
are
helping
leader
contributes
to
the
diverse
group
of
projects
that
are
administered
by
this
sig
to
have
a
chance
to
update
the
community.
Let's
have
them
yeah,
so
you
have
a
discussion
point,
but
that's
not
what
the
purpose
of
that
we
are
going
through
right
now
is.
B
The
audience
tomorrow
is
I
think
the
perfect
audience
for
this
right,
because
there
are
some
people.
It
should
be
in
that
discussion
like
hardik,
for
example,
that
it
aren't
I,
don't
see
on
the
call
today
he
might
be
already
pretty
page
to
everyone,
but
there
are
a
lot
of
people
and
it
does
feel
like
it
is
a
good
topic,
but
it
is
a
great
topic
for
tomorrow
and
I.
Think
if
you
put
it
on
the
agenda
for
tomorrow,
I
think
we
should
headline
with
that
agenda.
Yeah.
I
A
E
H
A
G
A
A
We
don't
think
that
this
capi
is
potentially
a
VIP
consumer,
but
they're
not
it
doesn't
necessarily
fall
underneath
that
umbrella,
so
I'll
paste
a
link
to
that
proposal,
which
is
totally
a
mess
right
now
and
my
plan
is
this
afternoon
to
go
through
it
and
what
with
more
detail
and
hopefully
try
and
have
a
focused
discussion
on
that
particular
one.
This
Friday
it
collapsed
the
camp
by
next
meeting.
So
that
way
we're
we
should
no
longer
having
a
life
cycle
group,
probably
by
one
to
two
weeks
time
frame.
C
Cubed
am
I
right
wrote
some
notes
in
the
gender
doc
reports
appeared
at
the
walls,
multiple
control
plane,
those
to
join
concurrently,
contributed
by
a
sous
developer.
He
also
promised
that
he
is
going
to
write
some
iteration
test
for
that.
We
merge
the
PR
that
allows
now
to
use
wildcards
for
sublimation
sitters.
C
We
also
now
have
experimental
support
for
renewing
certificates
automatically
during
upgrade.
This
was
general
consensus
in
the
cube,
a
diem
office
hours
that
we
might
want
to
do
it,
and
but
this
for
this
to
experimental,
when
it's
kind
unclear
how
we
want
to
proceed.
The
work
for
everyone
beta
to
the
configuration
of
cafe
diem,
has
already
started.
The
first
PR
exposes
the
certificate
key
that
is
used
to
copy
certificates
between
the
primary
control,
plane
and
secondary
control.
Plane
knows
that
joined
the
Questor.
This
is
the
work
here
is
going
to
continue.
C
B
Yes,
we
are,
you
know
we
did
our
like
feature
bankruptcy
thing,
and
so
we
have,
or
we
have
active
Simon
happening
in
the
master
French,
the
I
think
most
exciting.
One
is
I,
think
we
we
were
using
aggregated
API
servers
and
like
the
heavy
API
machinery,
and
now
we
are
using
or
will
shortly
be
using
crts,
which
will
also
it's
a
lot
lighter
weight.
F
B
Yes,
the
we
we
basically
were
I
think
remember.
The
community
produce
like
a
really
good
graph
would
show
that,
like
our
number
of
days
of
lag
was
like
had
gone
up
to
like
basic,
like
approaching
six
months
of
lag
between
releases,
KK
and
Risa,
the
the
corresponding
cops
version,
and
so
what
we
did
was
me
essentially
like
I'm
foreshadowing,
my
keuken
talk.
B
We
essentially
like
said
we
will
not
put
any
significant
new
features
in
cops:
112-111,
112,
113,
114
and
so
master
is
open
for
business
for
115,
but
112
113
114
should
be
much
more
like
just
no
new
features,
just
like
fixes
on
stuff
now
yeah,
so
we're
still
gonna
like
stabilize
them,
but
we
are
not.
We
are
essentially
saying
all
features
jumped
to
the
latest
the
future
release
now,
as
it
flows
to
continually
playing
catch-up,
trying
to
stabilize
and
add
new
features,
which
was
just
super
hard.
J
Three
times
all
right,
except
mini,
cube
sure
we
released
101
last
last
week,
which
is
defaults
to
Kate's
141
we've
got
the
one
one
milestone
planned
for
the
end
of
the
month,
with
our
first
stab
at
language,
localization
and
improved
support
for
image
mirrors
in
China
and
fast
restarts
the
image
mirrors
in
China
I'm.
Still
we
we
have
some
good
PR
s,
we're
still
a
little
worried
about
how
do
we?
J
How
do
we
prevent
poisoned
images
from
being
introduced
since
cue
bad
idiom
pulls
by
tag,
and
so
we
kind
of
have
to
trust
the
tags
rather
than
the
digests.
So
we're
we're
we're
looking
into
ways
to
still
have
a
good
user
experience
there
and
we
have
two
new
maintainer
so
who
will
actually
sorry
yeah?
It
is
114
one
sorry
about
that
two
new
maintainers
Sharif
and
media,
who
will
actually
be
and
keep
Connie
you
if
you
need
to
bring
anyone
on
stage
or
have
any
discussions
so.
J
We
go
yeah,
we've
we've
talked
about
actually
like
hard-coding.
A
map
you
know
for
the
releases
of
like
here
is
the
known
good
Shaw's.
It
gets
a
little
yucky,
but
that's
probably
what
we'll
have
to
do.
I
didn't
realize
that
I
see
somebody's
mentions
that
you
can
do
tag
at
digest.
I
will
have
to
I
didn't
realize.
That
was
a
feature,
so
I
will
check
on
that.
B
J
F
A
One
of
the
things
I
want
to
maybe
broach
is
the
topic
during
coop
con,
which
will
be
you
know
just
to
think
about
put
it
in
the
percolator
and,
let's
not
talk
about
it
now,
but
is
too
is
the
abstraction
layer
for
developer
centric
workflows
to
kind
of
put
not
necessarily
at
all
the
technology,
but
put
to
pull
that
abstraction
layer
for
many
cute
kind
and
dined
into
a
meat
grinder
just
so
that
we
have
like
one
unified
plane
by
which
a
person
can
kick.
The
backend
could
be
whatever
right.
A
A
K
And
unfortunately,
not
not
much
update
I
still
have
the
actions
Ida
action
items
for
myself
and
on
producing
the
release
artifacts.
If
someone
reached
out
to
me
offered
help
but
yeah,
we
need
to
lost
lost,
lost
contact
there
and
hopefully
get
that
get
that
going
and
yeah
I'm
still
still
need
to
work
on
the
the
organization
side
and
and
and
talking
with
testing
for
on
testing.
K
So
yeah,
just
God
got
caught
up
in
some
things
for
the
past
two
weeks,
but
I
am
planning
to
actually
dedicate
most
of
this
afternoon
and
more
time
this
week
today,
DM.
So
that's
the
update.
K
Yeah
but
the
the
idea
is
not
just
for
seda
DM
to
be
the
CLI,
which
is
you
know,
the
current
sort
of
implementation,
but
also
to
incorporate
the
sed
manager
code
and
and
and
ideas
that
that
Justin
has
been
working
on
for
a
while.
So
it's
a
layered
on
top
of
a
CLI.
So
that's
the
that's.
The
road
map
in.
A
B
I've
been
remiss
in
not
doing
my
bit
of
this
project,
so
I
apologize,
but
I.
Think
one
of
the
interesting
things
I
want
to
explore
is
Dinah.
I've
talked
a
lot
about.
This
is
whether
we,
whether
from
an
administer,
is
point
of
view.
It
would
be
better
to
do
it
as
a
library
or
two
which
would
be
Lions
with
natural
inclination,
but
the
alternative
which
is
interesting
is
to
drive
CLI
commands
so
that
we
so
there's
not
a
magic
box.
B
K
K
A
E
Time
sure,
well,
you
know,
I
kind
of
I
was
kind
of
quiet
when
we
kicked
off
all
the
discussions
about
creating
these
sub
meetings.
So
I
was
originally
against,
but
I
figured.
You
know,
I'm
newer
to
the
project
and
everybody
else
I'll
see
how
it
plays
out
before
I
add
my
opinion
to
the
mix,
and
so
so
far
my
takeaways
have
been
that
discussion
is
happening
all
over
the
place
in
various
mediums.
E
E
You
know
getting
quite
burdensome
to
try
to
keep
track
of
the
pieces
that
you
care
about
and
so
I
don't
think
it's
been
particularly
effective
and
I
think
we
deferred
a
lot
on
the
goals
and
non
goals
before
we
kicked
off
these
meetings,
and
so
the
pre
assumed
patterns
that
we
saw
emerging
with
particularly
around
web
hooks
and
assumptions
into
the
data
model
have
now
basically
you
know
come
to
fruition,
which
I
think
you
know.
That
was
a
higher-level
discussion
and
it
feels
like
because
it's
more
fragmented
now
it's
hard
to
coordinate.
You
know
ideas.
E
A
You're,
you
know
the
feedback
is
good
to
hear
I,
don't
think
no,
you
can't
and
as
anyone
who's
been
in
any
relationship
your
nose,
you
can't
invalidate
somebody's
feelings,
so
your
feelings
are
always
earth
to
hurt
and
I
understand
it.
The
I
think
we
need
to
do
a
better
job
of
getting
things
into
brass
tacks:
water
that
people
can
actually
consume.
I.
Think
that's
what
the
next
step
that
people
are
trying
to
get
towards.
A
So
I
think
what
we
can
do
at
least
going
forwards
is
to
make
sure
that
people
have
buy-in
for
the
caps
that
are
created
and
we
have
the
discussions
within
the
main
meeting.
We
kind
of
sort
of
collapsed
the
work
streams
over
time,
because
the
work
streams
are
just
a
conversation
point
for
us
for
those
groups
to
have
you
know,
ideas
and
discussions
about
different
options
that
exist,
but
at
the
end
of
the
day
you
know
the
it
has
to
collapse
down
to
the
main
capi
meeting
and
people
have
to
have
discussion
well.
E
E
You
know
or
are
all
the
different
things
proposed
in
those
documents
going
to
become
caps
or
does
anything
from
those
documents
have
to
become
caps,
and
so
at
this
point
it
seems
like
we're
working
on
an
assumption
that,
whatever
the
consensus
is
in
that
document,
the
Google
Doc
that's
going
to
become
the
cap,
and
then
it
seems
like
that's
no
I'm,
not
saying
this
is
gonna
happen.
But
if
that
is
the
case,
then
the
cap
is
basically
just
formalized.
E
Anything
is
particularly
suitable
and
I
think
we're
trading
one
problem
for
another,
but
if
I
wasn't
able
to
participate
in
the
proposal
document,
which
others
may
not
be
because
the
meetings
are
taking
place
at
various
times
and
discussions
in
various
locations
that
that
we
would
move
forward
with
something
that
really
nobody
or
a
minority
of
people
are
actually
happy
with
and
basically
for
the
sake
of
moving
forward.
So.
A
In
your
kind,
I
was
like
fair
up
to
a
point,
but
in
your
conversation
piece
you
have
a
bunch
of
at
hominins
right,
so
I
can
call
you
out
on
that,
because
it's
recorded
so
I
think
like
I
mentioned
before
people
have
to
come
back
together
and
inside
of
any
proposal.
There
should
be
the
options
listed,
a
versus
B
versus
C
versus
D.
That's
actually
listed
out
the
side
every
single
kept,
so
they
should
actually
enumerate
the
state
space
of
different
post
options
that
exist.
A
But
you
know,
and
like
I
mentioned
before,
you
should
be
part
of
when
we
collapsed
back
together
and
in
the
maned
capi
meeting.
That
folks
should
be
able
to
discuss
what
you
know.
What
are
the?
What
are
the
benefits
in
the
traffic's
and
I
guarantee
you
not
everyone's
going
to
be
happy
and
that's
going
to
be
part
of
the
trade
offs
that
we
have
to
make
and
we're
not
going
to
come
up
with
some
grand
design
that
will
be
meet
everybody's
needs
for
all
time.
A
A
The
most
the
proposal
should
be
of
a
recommendation
that
the
subcommittee
to
size,
but
it
should,
in
numerate
all
of
the
different
options
that
they
looked
at
and
talked
about
the
positives
and
negatives
that
they
saw
for
option
a
versus
B
versus
C,
and
it
should
be
very
to-the-point,
matter-of-fact,
be
like
this.
This
one
has
the
most
positives
versus
the
most
negatives
that
we
saw
and
if
they're
missing
something
call
them
out.
E
Well,
I
think
it
seems,
then,
that
that
proposal
doc
is
really
like
the
source
of
like
truth,
as
far
as
like
the
cab,
and
so
it
does
seem
like
it's
like
a
one-to-one
ratio
there
like,
if
you're
getting
one,
accept
a
proposal
and
we're
gonna
build
a
cap
off
of
whatever
that
accept.
The
proposal
was,
and
then
we're
gonna
vote
up
or
down
on
that
one
particular
one
Michael.
F
H
I
can
cover
that,
so
this
is
part
of
the
extension
mechanism
it
was.
It
was
a
document
just
to
list
out
the
pros
and
cons
of
various
ideas
that
we
had
and
the
web
the
web
hook
/g
RPC
kind
of
had
the
most
points
added
to
it.
So
I
thought
the
proposal
from
there
would
be
would
be
appropriate.
I
think.
I
It
is
something
is
that
I
think
the
root
of
this
problem
is
that
I
recall
the
initial
me
dick
off
what
is
were
stream,
something
that
was
actually
over.
Suppose,
yes
and
I
mentioned
that
at
some
point,
we
would
like
to
explore
different
station
mechanisms
again
just
to
to
come
back
to
my
previous
command.
Is
that
I
think
that
we
are
you're
in
this
dilemma
is
just
because
we
are
trying
to
feed
the
data
model,
maybe
in
consciously
blessing
that
this
has
happening,
that
this
kind
of
made
power
to
implement
more
than
one?
I
Essentially,
if
we,
if
we
try
to
keep
the
data
modeling
with
brother,
it's
an
compromised,
but
probably,
as
you
said
in
this
point,
probably
these
food
coffees
had
the
most
people
interested
in
and
I
have
my
own
proposal,
which
that's
not
been
getting
more
structure.
Iii
admit
that
it's
happening,
but
what
I
would
like
to
have
the
chance
to
implement
it,
but
is
that
a
model
does
a
field?
I
will
never
be
able
to
do
that,
so
I
will
never
be
to
concern.
I
A
Your
point,
okay,
I,
think
it's
important
to
denote
that
like
this
is
not
it.
This
is
alpha
2
right.
This
is
not
GA
right.
So,
if
you
have
a
better
idea
code
talks
or
in
the
communities
community
since
epoch,
you
know
some
people
in
the
community
came
with
code
and
examples
and
that
that
code
talked
for
itself
right.
If
you
have
an
idea,
there's
nothing
that
prevents
you
from
creating
a
POC
to
show
as
an
example.
A
I
Agree
that
that's
my
poem,
but
it's
a
data
model,
a
very
something
that
came
potentially
block
and
that
to
help
you
get
the
data
model
because
he
said
that
a
model
goes
in
one
direction.
This
probably
almost
impossible
to
implement
an
alternative
model.
Then
I
had
the
double
work
of
changing
the
data
model,
which
is
going
to
be
probably
harder
and
then
prove
my
point.
I
I
agree
with
you,
I
mean
I
ain't
noted
that
that's
what
I
had
not
been
pushing
more
my
idea,
because
I
understand
what
you
said,
but
my
only
point
is
in
the
in
the
content
of
the
data
model.
That's
the
only
thing
that
you
probably
prevent
other
activity
models.
If
we
are
looking
for
either
saying
that
is
happening
I
just
saying,
we
should
probably
be
careful
not
to
probably
consciously
bias
the
one
essential
model
and
prevent
no.
No,
my
purpose
was
just
making
tough
hard
implement
another
model.
I
G
I
G
I
think
I
think
that's
fair
feedback
and
I'll
could
release
the
proposal
a
little
more
so
that
we
don't
make
assumption
I.
Think
the
trick
for
me
is
like
when
we
put
something
in
the
in
a
proposal.
It's
like
a
document
living
document
that
like
can
be
changed
at
any
at
any
point
in
time
and
nothing
it's
set
in
stone
and
it's
not
even
in
a
pure
yet
so
it
shouldn't
be
taken
as
like.
The
final
were
the
like
the
source
of
truth
of
anything.
G
F
A
Think
it's
good
feedback
I.
Think
part
of
this,
too
is
I,
appreciate,
Michael
and
Pablo.
Raising
the
signal,
I
think
it's
good
signal
to
get
feedback
on
I
want
people
to
feel
empowered
that
they
can.
They
can
talk
about
hard
topics.
I.
Do
also
think
that
you
know
agenda
squashing.
This
isn't
necessarily
the
best
way
of
going
about
doing
it,
but
your
the
agenda
is
open.
So
you
can
add
your
points
at
the
group
topics
as
part
of
the
agenda.
A
So
I
hear
you
loud
and
clear,
I
think
there's
a
meeting
tomorrow,
because
we're
gonna
run
out
of
time
here
there
is
a
meeting
tomorrow
for
Cappy
I.
Think
that's
probably
a
better
place
for
some
of
this
I
do
think
that
people
are
trying
to
do
what's
in
the
best
interest
for
the
sub-project
I
haven't
seen,
people
try
to
be
malice,
I
do
think
they.