►
From YouTube: SIG App Delivery Meeting 2019-09-25
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
A
A
A
C
C
C
C
B
B
B
All
right,
let's
get
this
show
on
the
road
just
before
we
get
started
just
let
you
know
that
this
call
is
recorded.
So
anything
you
say
everyone
will
be
able
to
hear
and
if
you
don't
like
that,
don't
say
anything,
but
today
is
what
September
25th
or
is
it's
1/5
yeah
today
is
September
25th,
and
this
is
the
cig,
app
delivery,
call
I'm,
Bryan,
Liles
and
today
my
other
co-chairs
here
with
me.
B
B
A
Sir,
the
first
item
that
recently
the
cig
in
virginal
as
a
big
scenario,
application
delivery
is
basically
for
expand
or
kinetic
deerforce.
They
will
exchange
all
I
items
or
terminologies.
The
Stig
is
talking
about
and
we
have
already
realized
the
documentation
for
step
around
the
baby.
We
have
aware
they
received
a
lot
of
feedbacks
from
the
community
and
have
addressed
all
the
issues
and
then
so
now,
if
you
read
this
documentation,
you
will
see
several
definitions
like
what
is
a
plication
voltage
application
instances
extra.
He
knows
that.
A
Let
me
talk
about
application
or
any
kind
of
items
here,
but
they're
not
trying
to
defy
a
general
purpose
definition
for
that
terminology.
We
are
just
talking
about
okay.
What
is
the
application
we're
talking
about
in
this
sink?
So
the
purpose
of
this
documentation
is
not
like
a
and
this
mission
is
not
like
e
one
paper
or
something
it's
like
a
tool
or
dictionary.
A
So
if
anyone
working
on
or
working
this
sake,
okay,
if
you
are
confused
about
several
technologists,
so
the
first
place
you
may
want
to
look
at
is
if
it's
already
filing
this
documentation.
So
I
won't
give
you
example
by
start
from
application
so,
but
about
applications
here
and
there
in
this
SIG.
So
you
may
be
very
curious
that
okay,
what
aid
application
so
the
application
talking
this
take
is
more
like
something
about
what
we
want
to
deploy.
A
So
it's
basically
a
collection
of
decorative
configuration
and
we're
honorable
units
which
describes
for
the
platform
how
to
deploy
start
already
and
in
retire
the
company
six
artifacts.
That's
all
so,
there's
some,
but
somebody
that
that
is
asking
that.
Okay,
are
you
talking
about
runnable
units?
So
do
we
have
to
assume
that
the
runnable
units
should
be
micro,
services
or
not?
Okay?
A
So
from
my
point
by
point
of
view
that
it
can
be
micro
services,
but
it
don't
have
to
be
because
micro
services
is
a
very
good
start
if
you
want
to
do
CRO
native
application
delivery,
but
it's
not
the
requirement,
because
we
can
also
start
from
more
traditional
workers
like
the
virtual
machines
or
just
the
java
web
applications
without
container
that's
fine,
the
dot
that
will
not
stop.
That
will
still
use
a
quality
way
to
delivers
application,
because
our
expand
is
more
little
better.
So
that
is
okay.
What
we're
talking
about
the
application?
A
We
also
have
comes
bounding
concepts
like
application
instances,
it's
more
like
the
runtime
concept,
because
the
application
we're
talking
about
is
actually
static,
description
of
the
application,
but
how
only
puts
checked
or
how
already
deployed
we
would
have
application
instances.
The
best
example
of
application
instances
is
like
okay.
If
you
have
somebody
something
like
kubernetes
deployment,
many
replicas,
it's
three,
then
you
will
have
three
application
instances
that
which
are
or
equivalent-
but
please
note
that
it's
not
actually
staying.
We
must
have
three
parts
or
three
containers
workers
on
the
application
instances
label.
A
Another
example
is
HDD
cluster,
which
is
managed
about
operator
in
size
of
five
that
you
will
five.
You
will
get
five
application
instances
in
that
case,
but
in
that
case
the
five
application
instance
are
not
equivalent
to
each
other.
There
are
five
individual
study
instance,
that's
all,
and
we
also
have
concept
like
platform.
A
The
platform
is
what
we
you
to
run
this
application,
which
is
underlying
system
that
provides
the
functionality,
is
to
manage
applications
through
its
lifecycle,
the
best
lump,
of
course
in
the
kubernetes,
but
again
in
this
league,
but
are
saying
you
have
to
use
kubernetes
or
everything
must
be
deployed
on
where
it's
not
that
case,
the
only
firkin
Sanko
native
application
drivers.
So
if
you're
using
directly
the
class,
the
class
services
or
cloud
infrastructure,
that's
fine,
o
or
you
may
be
using
stuff
like
splits
from
Target.
A
D
I
interact
with
the
question:
yes,
would
you
prefer
me
to
keep
them
till
the
end
of
my
I
was
actually
wanting
to
ask
about
your
definition
of
an
application,
so
typically
applications
are
made
up
of
many
micro
services,
or
at
least
more
than
one.
You
seem
to
be
implying
that
that
an
application
instance
was
kind
of
a
DM,
whereas
in
my
mind
an
application
is,
is
a
collection
of
lots
of
different.
You
know
there
may
be
micro
services,
as
you
say.
D
D
A
The
application
application
instance,
it
means
action
means
okay,
if
you
have
one
application
and
they
come
running
so
maybe
some
case,
it
only
had
one
application.
Instance,
that's
fine
that
it's
not
same
issued
already.
We
should
always
have
a
lot
of
instances
or
replicas.
That's
not
again.
It's
actually
application.
Application
instance
is
only
the
description
of
your
of
your
run
of
your
application.
You
know
running
stations,
so
it
is
running
and
it
will
have
stable
or
only
one
or
any
number
of
application
instances
that
that
is
the
image
of
the
application
instances.
E
F
A
A
Imma
just
skip
the
rest
of
the
concept,
because
it's
very
straightforward
to
go
to
the
layers
of
application
flavory,
because
this
is
actually
the
guidelines
ones
we
use
if
we
want
to
evaluate
a
open
source
project
about
application
delivery,
or
you
know
the
whole
lifecycle
of
computing
application.
How
to
deliver
applications
include
anyway,
so
and
I
actually
jawed
a
various
more
a
very
simple
picture
to
show
the
the
concept
here.
A
So
we
basically
have
four
layers
if,
during
the
whole
lifecycle
of
delivery
application
and
the
one
to
three
layers
are
actually
main
concern
of
this
sake.
The
reason
we
are
gonna
join
this
picture,
because
actually
we
can
go
to
the
layers
one
by
one,
so
the
first
layer,
if
you
want
to
clear
application,
it's
actually
about
how
to
describe
the
application.
For
example
yeah
somebody
is
using
here
somebody
you
can
customize
somebody
just
using
the
demo
Klaus
and
so
something
you
see
nap.
A
That's
all
fine
because
they
are
all
applications
described
in
here,
but
as
you
can
see
that
maybe
those
project
has
work
and
also
functionalities
in
other
areas.
So
we
will
not
use
specific
example
here
we
will
not
say
Heil
means
application,
definition
or
customize
advocate
in
Tunisia.
It's
not
like
that.
It's
some
there
are
a
lot
of
project
they
make.
The
action
have
functionalities
that
actually
covers
many
aspects
in
this
picture.
So
that's
why
we
want
to
makes
it
clear
that
what
is
a
specific
functionality
to
honestly
example
in
helm,
it
has
something
coding
up.
A
The
chart
is
a
perfect
example
for
an
application
description
and
also
another
perfect,
vampire,
benezia
Moffat's.
It's
described
what
portfolio
mingle
application
would
easy
or
complication
for
application,
but
also
we
notice
there
are
technology
that
is
used
to
do,
parameter
or
configuration,
for
example,
customized.
We
also
have
several
DSL
suppose
they
have
several.
For
example,
templating
make
them.
A
Is
also
in
the
scope
of
application
declaration
and
the
application
layer
actually
will
give
us
the
exactly
configuration
or
a
description
of
your
Abdullah
application,
the
object
at
layer.
We
will
have
second
layer
which
needs
application
to
coil
around
that.
As
long
as
we
have
an
application,
we
will
believe
our
installation
through
the
platform,
but
we
won't
you.
We
want
to
do
that
with
strategy.
The
second
layer
actually
defies
how
we
do
the
application
lifecycle
management,
how
to
bootstrap
the
application,
how
to
roll
them
advocate
application.
A
They
actually
have
functionalities
in
this
layer
and
the
traditional
CD
system,
auto
focus
of
these
layers
Aska,
as
we
should
also
pay
attention
to
that
reaction,
have
the
layer
of
three
ops
with
a
layer.
Two.
So
I
will
explain
this
a
little
bit
here,
because
it's
maybe
kind
of
different
from
what
they
were
thinking
so
in
layer
two,
we
are
actually
focused
on
the
application
layer
will
route,
but
I
think
you
guys
might
notice
that
we
also
have
something
like
the
application
instance.
A
If
level
route,
for
example,
you
have
application
that
application
is
using
a
claim
that
it
will
usually
created
implement
at
workloads.
That
means
it
will
encourage
deployment
to
start
of
these
application
instances.
So,
first
of
all,
we
have
the
road
strategy
fold
application
level
like
an
arrow.
That
means
I
want
to
created
the
application
first
and
then
create
a
new
version
of
an
occasion,
instances
as
narrow
instance
and
then
trying
to
do
a/b
testing
to
do
benchmarking
and
then
promote
the
all
the
instances
to
production
women
and
then
delete
the
owner
of
the
application.
A
D
H
F
E
E
D
E
You
might
just
argue
about
it,
how
you
slice
and
dice
it,
but,
although
I
think
it's
a
good
first
structure
that
we
can
work
with,
it
also
separates
us
a
bit
from
like
something
like
cig,
apps
and
the
kubernetes
working
groups
and
has
a
nice
definition
in
me.
This
is
this
layers
are
a
good
starting
point.
I
think
they
might
need
some
time
tweaking,
maybe
but
most
of
the
naming
sites,
but
I
find
them
useful
and
I'd
like
to
get
that
feedback
from
others.
Whether
you
like
generally
think
this
is
useful.
D
E
Yeah
I
think
we
can
do
that,
so
I
don't
see
initiative.
I
think
my
only
comment
is
I
would
really
really
give
them
layers
actual
in
their
names,
which
we
have
our
where,
like
application
definition
deployment,
all-out
instance,
automation,
platform
related
out
platform
per
se
and
you're,
just
like
the
zero.
E
Data
layer,
zero
piece
and
under
development
yeah
we
can
put
this
one
on
top,
so
it
modified.
But
for
me
this
is
a
good
starting
point
like
to
put.
This
may
be
a
bit
more
front
and
center
like
this
is
how
we
see
the
world,
that's
where
I
think
where
we
could
have
all
in
agreement.
This
is
helpful
because
it's
the
first
real
structuring
of
the
work
that
we
are
doing
here.
A
D
A
A
A
See
the
example
here,
I
think
it's
actually
there
the
clear
boundary
between
these
two
layers
yeah.
But
that
item
means
that
the
terminology
is
here
like
application
or
application
instances
are
very
confusing.
But
if
you
reading
these
example,
you
I
think
they
get
a
clear
idea
about
that,
and
it's
always
will
come
that
if
anybody
can
compose
new
nating
for
the
terminologies
here.
A
E
It
might
also
help
us
with
with
the
initial
steps
around
the
landscape,
because
now
we
could
actually
try
to
fit
tools
into
the
different
layers
in
which
they
should
cover,
with
which
the
cover
obviously
sound,
well
spend
maybe
multiples,
but
that
would
also
be
a
good
test
for
the
tools
that
we
have.
Whether
they're
the
layer
definition
here
helps
us.
A
E
A
A
E
E
But
that
would
be
the
next
step
for
me
to
put
some
tools
in
there
and
see
how
how
this
works
out,
because
then
I
think
we
also
closer
to
the
landscape
definition
like
we.
Some
tools
might
do
it,
but
they're,
not
a
finding
that
that
we
should
have,
as
as
part
of
this
eventually
is
to
give
some
to
provide
some
proposal
to
the
CAF.
As
we
see
something
like
fundamentally
missing,
it
should
be
there,
and
this
should
then
also
become
obvious.
Daddy.
I
E
A
Yeah,
another
motivation
that
we
want.
You
I
think
you
fiber
layers,
because
I
think
it
is
actually
you
can
see
the
change
that
people
are
trying
to
now
use
an
operator
to
believer
adapt
their
applications.
If
you
are
used
to
deliver
application,
will
notice
that
okay,
you
actually
have
a
layer,
3
component,
which
is
actually
a
operator,
but
you
still
need
to
have
the
layer,
true
project,
to
handle
the
application
role
for
you
program
how
to
roll
the
operator
in
the
narrowly
right
that
is
actually
in
two
separate
concerns
there.
A
E
And
if
example,
other
projects
we
have
to
CID
would
also
be
part
of
the
application
definition.
To
some
extent,
you
can
see
that
there
there
are
dependencies
which
I
think
are
fine,
but
that's
why
I
think
it's
a
good
idea
to
define
for
like
each
of
those
not
for
each,
but
at
least
for
some
of
the
projects
that
we
see.
Are
there
like
helm
and
a
couple
of
others
what
we
see
them
doing
at
the
different
layers,
yeah.
A
E
That
from
the
overseas,
the
one
where
they
shouldn't
be
really
doing
a
lot
yeah
well,
the
question
was
an
operator
is
actually
as
they
used
it
done
as
the
kubernetes
api,
whether
they
are
actually
extensions
to
the
platform
or
not.
What
did
I
implement
the
custom
see
light
is
an
extension
to
the
platform
world
or
not
I.
Think.
A
The
point
is
that
we
will
further
so
okay,
the
functionality
could
be
provided
by
platform
is
its
McClung
sense,
but
the
functionality
blown
into
each
layers.
There
is
like
the
the
networking
model.
Actually,
all
the
components
are
in
linux,
kernel
right,
but
the
different
components
may
working
on
the
deeper
layers
of
the
networking
model.
This
also
same
case
for
our
layers
here,
so
the
functionality
is,
could
I
think
most
of
the
functionalities
are
provided
by
the
platform
or
plugins
or
platform
right,
but
notes
functionality
is
actually
working
on
different
layers.
A
There
is
logical
layer,
it's
not
the
physical
area.
You
know
you
have.
You
must
have
several
projects
working,
it's
not
like
that.
So
even
you
are
using
kubernetes
itself.
You
have
no
other
reporters
at
all,
but
the
crouppen
is
just
provide
different
functionalities
in
deep
layers.
So
that
is
I.
Think
the
point
of
this
diagram.
A
So
that,
depending
on
what
you're,
using
those
functionality
to
do,
wadcutter
purpose,
for
example,
if
you
are
just
talking
about
in
a
birth
policy
right,
it's
more
like
belonging
to
the
platform
later.
But
if
you're,
using,
for
example,
in
the
affinity
functionality
to
do
wrong
strategies,
then
Road
strategy
itself
faintly
plumbing
to
a
narrative.
But
it's
not
saying
that
not
affinity
is
a
layer
to
fraction
I,
think
no
definitive
steal
or
functionality
from
from
the
platform
layer.
A
E
E
It
might
even
be
a
mystic
level
of
abstraction
because
actually
the
higher
layers,
you
certainly
need
to
know
a
lot
from
the
deeper
layers
and
one
example.
There
is
the
little
dicey,
specific
deploy
to
multiple
environments.
If
you
suddenly
have
environment
specifics
like
a
very
high
level
like
the
application
definition
there,
you
might
create
dependencies
that
you
even
don't
want
to
be
there
necessarily
so
I
might
have
like
my
I'm.
Another
developer
of
duplication,
read
event,
decided
individual
components
and
so
forth,
but
I'm,
not
the
one
responsible
necessary
for
networking
securities.
E
I
might
not
even
be
able
to
define
this
if
I
already
have
to
define
it
at
my
levels.
I
have
trouble
because
I'm
might
not
even
be
my
responsibility
and
nor
do
I
have
the
knowledge
to
even
define
these
things.
If,
however,
a
certain
tools
which
requires
me
to
define
them
at
a
higher
level,
then
I
would
be,
then
it
would
be
in
trouble,
maybe
if
it
makes
any
sense.
You.
D
Know
we've
come
across
similar
similar
concerns
in
the
multi
cluster
environment.
There
are
often-
and
this
is
relates
to
policies
in
general
I-
think
there
are
policies
that
are
implemented
by
the
application
definition
layer.
So
things
like
you
know
my
application
has
to
be.
You
know
financial
standard,
XYZ
compliant
and
that's
a
that's
actually,
typically
in
the
application
definition.
They
know
that
that
this
is
a
credit
card
application
and
it
needs
to
be
you
know
sarbanes-oxley,
compliant
or
whatever
that
may
be,
but
then
you
know
further
down
the
infrastructure.
E
Yeah
I
think
it's
even
easier
examples,
for
me
is
it's
the
application
definition
I
might
just
defined
which
services
I'm
using
but
I
might
not
necessarily
care
about
replica
accounts
because
I
don't
know
which
of
which
environment
it's
deployed
to
so
replica
accounts
might
come
in
later,
because,
depending
whether
this
is
a
test
environment
or
production,
where
I
meant
replica
counts
will
be
different.
Yeah
yeah
I
wouldn't
even
define
this,
but
so,
if
you
look
at
it,
it's
kind
of
also
separated.
E
I
Different
sorts
of
configuration
like
if
we
take
number
of
replicas
as
an
example
at
different
layers,
so
you
might
start
out
specifying
a
replica
count
for
digital
environments
at
the
application,
deploy
and
roll
out
layer
for
each
environment
that
you
deploy
to.
But
then
you
might
switch
to
implementing
it
at
the
application
instance.
Automation,
operation
layer,
if
you
switch
to
horizontal
plot
Auto,
scalars
and
kubernetes,
for
example,
and
you
get
that
benefit
out
of
the
capabilities
of
the
platform
or
of
higher-level
API,
is
built
on
the
platform
same
thing
for
a
blue-green
rollout
strategies.
E
Yeah
but,
for
example,
with
Bluegreen
I
think
it's
a
great
example
here,
because
there's
no
definition
right
now
how
it
would
specify
that
something
is
even
a
Bluegreen
deployment.
So
if
you
would
write
it
down
in
our
charts
and
maybe
if
your
routes
and
stuff,
you
would
know
that
it's
a
blue
queen,
but
it's
nowhere
explicitly
defined.
But
this
is
currently
a
degree
performance
right.
But
the
green
part
I
think
that
you
would
have
to
like
appear
that
meta-level
to
even
understand
what
these
things
things.
Things
mean.
I.
I
Think
it
actually
might
make
sense
to
have
a
application
deployment,
workflow
layer,
that's
a
level
up
from
declarative
application
definition,
and
then
that's
where
you
would
put
your
today.
You
might
have
a
Bluegreen
deployment
strategy
say
to
find
in
spinnaker,
which
would
be
orchestrating
multiple
instances
that
are
declaratively
described
by
a
helmet
or
by
manifests
and
in
the
future.
What
I'm
saying
is
that
responsibilities?
D
D
Be
I
think
lay
it
the
way
it
is
to
meal.
It
makes
more
sense
than
what
you're
describing
but
I
think
it'd
be
worth
making
sure
that
we
know
what
order
those
layers
are
in,
so
that
people
don't
get
confused.
One
could
probably
argue
it
both
ways,
but
the
way
it
is
diagrammed
at
the
moment
seems
to
make
sense
to
me
flipping
it
around
makes
it
more
confusing.
I
guess.
I
A
My
personal
idea
that
you
know
the
timing
of
the
functionalities
happened,
if
not
actually
a
moment
diagram
trying
to
show
the
readers,
because
yeah
of
admission
that
activation
I
don't
make.
It
may
have
been
happening
in
the
workflow
I'm,
speeding
her
at
the
very
last
stage,
but
I
think
it's
more
like.
We
were
trying
to
illustrate
that
to
reader
that
what
you
need
to
have,
where
you
take
back.
For
example,
you
didn't
have
application
descriptor
before
you
deploy
application,
then.
G
D
There's
an
interesting
question
in
my
mind
that
this
raises-
and
it
probably
has
bearing
on
the
previous
question
around
what
order
the
layers
are
in
one-
could
adopt
the
approach
that
we
would
like
these
layers
to
be
kind
of
interchangeable.
If
I
can,
oh,
maybe
that's
not
the
right
word,
but
but
an
application
descriptor
might
be
independent
of
whether
helm
is
going
to
deploy
it
or
not.
D
Maybe
the
application
definition
is
kind
of
contained
in
the
spinnaker
definition
at
the
moment,
the
spinnaker
config
and
another
spinning
cage,
but
by
any
stretch
of
the
imagination
but
I'm
guessing
here,
and
maybe
that's
where
some
of
the
confusions
coming
in
sense.
Is
that
worth
talking
about
whether
we
think
these
layers
should
be
kind
of
that
the
higher
layers
should
be
independent
of
the
imitations
of
the
lower
layers
or.
A
H
Yeah
I
would
say
that
it's
a
detail
of
like
because
you
can
build
things
that
are
very
composable
and
you
can
build
things
that
are
very
tightly
coupled
than
there
are
pros
and
cons
to
both
approaches
so
having
something
that
does
end
to
end
one
thing
in
a
very
opinionated
way
and
provides
a
high
level
abstraction
like
Cloud
Foundry
is
one
approach
having
something
that's
nearly
the
opposite
approach,
there's
very
like
LED
and
very,
like
separate
concerns
like
kubernetes
bill
out
of
bits
and
Hellman.
The
other
things
like
both
of
those
do
a
valid.
H
G
H
Those
you
might
say
well,
no
like
it's
much
more
componentized,
yeah
III.
Don't
think
that's
a
flaw,
I
think,
but
it
is
worth
you
know
it's
worth:
observing
I,
don't
I,
don't
I.
Also,
don't
think
we
should
say
categorically.
These
are
always
separate
because
it's
an
implementation
detail
of
whatever
that.
A
G
E
H
E
Even
say
that,
like
with
the
spinnaker
example,
okay,
which
is
CDF,
but
but
we
didn't
care
about
it,
CVS
or
not,
where
is
the
explicit
definition
and
what
does
Mike's?
The
definition
of
a
Bluegreen
deployment
looks
like
and
who
has
specified
this
usually
and
can
I
use
a
higher
level
definitions
work
with
this?
Where
do
I
have
to
rec?
A
One
thing:
actually
you
you
actually.
This
is
a
very
important
point
because
from
my
phone
we
knew
that,
for
example,
the
the
policy
or
technician
of
these
challenges
of
the
of
the
lower
layers
should
be
part
of
the
a
ossification
example
your
blue-green
component,
shallot
or
ease.
The
definition
of
that
part
should
be
part
of
that
patient
equation.
A
B
E
Is
after
the
cube
run,
proposals
I
have
to
admit
a
happen.
It
didn't
have
any
times
yet
to
write
them
and
I
have
to
do
it
right
now,
Allison
to
send
out
the
note,
but
one
will
be
like
fully
general
group
introduction
and
the
bigger
discussion
is
I
think
believe.
What
do
we
want?
The
second
one
to
be
about
once
you've
really
been
producing
to
the
work
of
the
group
and
what
we
should
and
what
it's
doing
to
the
people
are
up
to
spade.
E
Stay
up
to
speed,
understand
the
Charter
and
so
forth,
and
then
the
second
one
could
be
more
like
of
a
workshop
and
I.
Don't
as
I
haven't
written
it
because
I
was
opening
up
to
the
group
would
like
the
whole
application
definition.
The
dictionary
maybe
be
something
to
discuss
with
the
wider
audience
and
to
put
this
into
the
second
session
I.
B
H
Mean
if
we've
got
a
sense
of
the
number
of
people
we
might
have
in
the
room
like
something
with
like
walls
and
post-its
will
be
amazing,
like
in
terms
of
we
have
this
model.
Where
do
you
think
think
things
fit
and
like
put
put
your
favorite
projects
on
the
board
type
things
like
gangs
like,
but
it
depends
on
how
many
people
we
think
might
like
what
size
room
we've
got
and
how
many
people
who
might
get
coming
on.
B
E
H
E
But
doable
agreement
with
seed,
okay,
how
projects
sit
in
there
and
they
have
the
dictionary
and
the
discussion
is
the
working
session
and
the
first
one
is
just
in
production
to
recharge
things.
We've
work
on.
Why
do
we
exist
general
questions,
so
I
will
then
really
promise
to
write
this
up.
Luckily,
you
have
to
spend
some
time
at
the
airport
tomorrow
and
this
number
could
he
do
it,
but
on
Sunday,
I
didn't
find
time
to
and
I'll
share
it
then,
on
the
mailing
list.
B
All
right
so
right
now
we're
are
we
at
the
last
bullet
point,
because
what
I
wanted
to
get
to,
because
it's
a
kind
of
important
one
for
this
meeting
for
this
slide
itself.
So
Quinton
raises
that
there
are
a
lot
of
CMC
of
cig
meanings
right
now
and
in
the
in
the
back-channel
worked
out
that
one
possibility.
One
thing
we
could
do
is
we
could
actually
change
weeks,
keep
the
same
time
but
move
it
to
the
first
and
the
third
week
rather
than
the
second
and
the
fourth
week,
I.
A
B
E
G
G
B
E
B
Right
well,
yeah
today
was
a
pretty
lively
discussion.
I
am
so
happy
to
see
that
there's
so
much
interaction
around
around
this
dictionary
and
and
and
just
get
it
all
right
and
clear.
So
until
next
time.
Actually
we
need
to
figure
out
what
the
next
date
will
be,
but
we'll
do
that
offline,
so
I
think.