►
From YouTube: Cloud Foundry for Kubernetes SIG [January 2020]
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
B
Before
going
into
the
actual
contents,
I
think
one
of
the
suggestions
from
last
time
or
maybe
like
before.
We
even
start
that
Chris.
Could
you
start
the
recording?
Oh,
you
already
did
it's
very
recording
yep
and
what
we
can
edit
all
the
intro
chat
is,
is
recorded
too
so
yeah
before
going
into
the
actual
topic.
I
think
one
of
the
suggestions
from
last
time
was
to
see
if
we
kind
of
time
box
discussions
a
little
bit
I
think
there
was
a
suggestion
to
say
so
far.
B
C
Universe
in
the
sense
of
initially
when
we
started
we
nee,
we
kind
of
had
to
everything
in
the
arena
team
is
no
one
else
was
going
to
do
it.
We
had
to
do
the
failure
than
we
have
to
do
the
log
similar
that
kind
of
stuff
in
the
networking,
and
that
that
you
know
didn't
seem
very
scalable
long
term.
So
we
tried
to
stop
doing
those
things.
You
know
the
other
teams
with
more
context
could
start
doing
them
and
that's
that's
what
we've
been
in
the
universe,
but
I
think
that
the
I
mean
we
mean.
C
Ebert
is
a
thing
like
that.
That
is
now
a
thing.
I
things
we've
got
ready
what
I'm
figuring
out,
how
to
move
and
keeps
yet
we've
got
the
networking
people
etc.
So
I
think
now
that
we're
really
talking
about
is
how
do
we
get
like
a
cf-19
5,
as
in
we
get
small
to
bilities
native
containers
that
are
hopefully
a
puddle
applied
batch
F
away
from
adding
to
improve
Buster,
and
you
have
you
can
see
it
pushing
this.
B
C
C
We
need
to
decide
the
topic
a
bit
earlier,
but
the
thought
was
that
could
be
kind
of
get
like
a
combined
roadmap
for
all
the
different
themes,
so
we
would
have
kind
of
a
tiara
an
overall
idea
of
when
when,
when
you
know,
we
expected
the
warden
of
all
the
things
integrated
when
we
expect
such
cool
thing
when
we
expect,
when
each
team
expects
features
and
cetera
et
cetera,
I
think
I
mean
to
be
honest,
a
lot
of
that
currently
happens
inside
pivotal.
The
nature
of
a
lot
of
it.
C
Maybe
the
other
conversation
happens
very
tangentially
related,
but
here's
how
we
can
get
more
of
a
open
source
ecosystem
around
some
of
this
selling
stuff.
How
do
we
open
up
the
roadmap?
How
do
we
open
up
the
pieces
that
we're
doing?
How
do
we
build
some
kind
of
independent
excitement
around
some
of
these
things
for
being
spent,
and
that's
did
their
things,
and
people
should
be
excited
about?
How
do
we
do
that?
One
of
the
things
we
should
be
talking
about
I
think
those
would
be
conversations
I.
A
Fully
support
the
idea
of
having
a
written
down
roadmap
with
all
of
the
teams
declaring
their
intent
and
their
and
estimating
a
timeline
on
delivery
of
the
things
they're
working
on,
because
part
of
the
problem
we've
had
at
Sousa
is
that
we
thought
we
knew
what
everyone's
roadmap
was.
We
based
our
own
product,
roam
ads
on
those
things,
and
now
it
turns
out
there's
like
very
different
ideas
of
how
we're
gonna
do
this
thing,
as,
as
you
might
have
ascertained
from
my
posts
and
my
previous
rants.
A
We've
got
to
come
to
some
agreement
on
on
how
much
of
that
you're
going
to
use
where
cube
CF
fits
in
the
efforts
that
are
going
on
in
relevant,
and
we
have
to
have
that
discussion
here.
I
know:
there's
been
some
informal.
Can
discussions
with
SCI
between
our
team
has
we
need
to
surface
this
issue
because
it's
it's
we're
gonna
end
up
in
two
different
places.
If
we
don't
figure
it
out
now
let.
C
Is
important
have
roadmap
and
the
evil
and
an
idea
of
where
we're
all
headed
at
the
same
time?
That's
you
know.
We
know.
Maps
are
gonna
change
like
what
we
don't
want
to
do
is
write
a
road
map
that
we
don't
change,
sometimes
and
something
to
give
a
change,
and
that
should
be
okay
right,
I,
don't
think
that
where
we
mess
I
do
simply
messed
up
as
a
community
a
lot
I,
don't
think
that
the
way
we
messed
up
was
changing
the
road
map,
I
think
sometimes
I
mean
I.
Think
I.
C
Think
UCF
was
exactly
the
right
thing
to
build.
Given
the
assumptions
that
existed,
you
know
when
we
started
building
it.
Those
are
something's
changed
and
it
would
suck
to
keep
along
the
line.
Just
because
we
had
a
road
map,
you
see
what
I
mean,
what
we're
I
think
we
completely
messed
up.
Is
we
didn't
communicate
at
the
community
at
all
about
those
assumptions
changing
and
well,
if
I'm,
honest
I,
don't
think
UCF
is
the
long-term
correct
direction.
C
C
C
D
Jen
I
can
add
something
like
I
totally
agree
with
choose
that
yeah.
It's
like
teams
and
everything,
but
I
think
that
those
wand
roles.
It
is
missing.
That
was
a
Diego
transition.
That
Samba
needs
to
look
overall
program
and
talked
to
some
teams.
We
can
say:
oh
yeah,
please
teams
communicate
with
each
other.
There
should
be
I
know
some
program
manager,
product
manager,
someone
who
will
go
and
pincer
people
from
other
teams.
Like
said,
ok,
you
have
dependency
on
some
as
a
team.
Let's
do
better
than
just
might
be
sometimes
mentions
I.
D
Would
he
trouble
issues?
What
was
a
chart?
I?
Let
us
get
to
requirements
from
one
product
to
another
product
earlier
then
ask
p.m.
of
the
project
to
do
this,
because
the
end
of
the
project
is
always
overloaded
and
it's
that
might
be
one
of
the
debye
idea
talk
because
on
a
rainy
site
we
don't
have
I,
know
I
didn't
feel
the
strong
communication
with
keep
CF
course.
Our
communication
about
keeps
here,
I'd
I,
think
it
was
even
less
on
the
copy
and
relevance
I'd
like
it
started,
but
it's
going
on
p.m.
D
level
should
be
on
some
higher
level
and
that
what
prevents
it
from
doing
some
product
I
know
that
on
pivotal
site
or
be
aware
site
that
definitely
sound
like
people
who
product
Eisen
and
they
know
what's
happening,
but
they
working
more
on
closest
spot
and
notes.
Is
that
much
an
open
source
part
with
all
people
right
now?
That's.
A
Right
exactly
where
I
I
feel
the
cell
down,
there
was
project
management.
There
was
or
program
management.
Whatever
you
want
to
call
the
overarching
management
that
happened
within
pivotal
and
we
didn't
we
don't.
We
it'll
doughnut
have
a
clarity
on
where
those
efforts
are
headed,
so
I
hereby
nominate
daniel
Jones
to
be
the
PM
for
CF
on
communities.
A
So
we're
we
are
far
enough
down
a
path
that
we
are
gonna
continue
working
with
cube,
CF
again,
not
married
to
the
repo
name,
but
we've
done
this
work
on
the
operator.
We
we
got
consensus
from
SAV
and
IBM
and
Sousa
early
on
that
this
end
and
validated
that
against
the
Cloud
Foundry
community,
that
this
is
a
direction
that
we
want
to
go
and
we're
going
to
develop
these
technology,
and
so
we
have
something
that
works
with
the
CF
operator.
It
works.
It's
passing
cats
with
cube
CF.
The
missing
component
I
think
is
that
cube.
A
Cf
is
not
yet
an
upstream
artifact.
We
plan
to
rectify
that
very
soon.
At
that
point,
what
I
would
like
is
at
least
an
effort
from
this
group
to
understand
what
the
work
that's
been
done
and
how
it
works
and
I
need
very
specific
technical
objections
to
problems.
I
need
to,
because
we've
talked
philosophically
about
oh
well,
we've
changed
their
idea
of
how
this
should
be.
Maybe
we
should
be
using
a
different
template.
Er
I
need
to
know
what's
wrong
with
what
we
did
before.
A
We
change
off
it
and
start
on
a
new
thing
because
and
here's
my
it's
not
some
emotional
investment
with
cube
CF,
because
I
have
a
little
bit
of
that.
But
I
think
I
can
see
that
what
I'm
worried
about
is
a
timeline
I'm
worried
about
having
a
deliverable
and
having
a
path
towards
a
deliverable
that
it
doesn't
prevent
working
releases
now
so
I
think
we
have
a
path,
and
you
know
Vlad
and
Jeff.
A
So
I
think
we
have
a
path
from
where
we
are
right
now,
mostly
the
Susan
team,
but
those
people
who
are
involved
in
cube,
CF
and
CF
operator
to
get
us
to
where
we
need
to
be
in
six
months
time
in
a
year's
time,
and
that
was
one
of
the
reasons
we
took
s
CF
and
started
to
to
change
it.
Do
instead
of
starting
from
scratch,
we
took
what
we
had
done
already
and
figured
out
a
path
to
get
to
a
truly
more
cube,
idiomatic
way
of
producing
cloud
foundry
on
kubernetes
I.
A
Don't
think
that
it's
broadly
understood
that
that's
our
intention
or
I
mean
I,
know:
I've
said
that
but
I
don't
think,
we've
gotten
into
the
technical
details
of
how
we
get
there
from
here.
Basically,
how
we
meet
the
component
teams
when
they
are
ready
to
release,
cube
native
and
cube
idiomatic
components,
and
we
have
that
discussion
here
or
should
we
plan
a
document
or
like
how
should
that
discussion
go.
C
C
E
C
E
C
I
think
loves
and
happy
I
mean
I
think
would
be
good,
I.
Think
I
think
I'd
really
like
to
get
a
document
that
one
said
where'd
we
want
to
end
up
and
to
said
I'm
not
sure
I
won't
Duncan,
which
is
like
YPG,
has
done.
I
I
understand,
what's
really
saying,
but
I
think
what
we
want
to
say
is
is
where
we
want
to
end
up,
and
here
is
how
we
plan
to
get
there
and
then
we
can
all
discuss
yeah
positively.
How
we
do
that
and
except
again.
C
D
And
I
want
to
read
again
that
we
need
some
like
responsible,
your
son,
who
will
go
and
talk
to
all
other
teams
and
I'm
not
sure
that
from
VMware
Site
we
can
provide
this.
Be
us
and
you
too
I
caused
a
noise.
We
have
little
bit
less
people,
so
I,
don't
think
we
have
3
p.m.
right
now.
So
maybe
it's
some
other
company
should
get
this
person
who
will
go
by
Kazakh
teams
might
be
go
like
say:
ok,
let's
dark!
D
A
No
I
I
joke
a
semi
jokingly,
suggested
Daniel,
but
Berendt
has
actually
been
the
quarterback
for
the
early
phases
of
these
discussions
and
I.
Just
wonder
if
that's
too
much
low
workload
to
throw
at
at
you
parent
or
if
there's
interest
there
in
in
project
managing
the
larger
thing
or
or
what
that
would
mean
well.
B
E
I
think
there's
also
something
of
a
different
skill
set.
There
I
mean
we
have
I
think
what
Alex
is
referring
to
is
how
we
had
a
lot
of
what
we
call
technical
program
management
investment
on
things
like
CFC
are
and
Kubo,
as
well
as
being
coordinated
on
the
pivotal
side,
VMware
side
with
PKS
and
so
I
think,
there's
a
possibility
that
we
might
be
able
to
allocate
on
the
VMware
side,
some
of
the
folks
that
work
in
that
technical
program
management
practice
to
help
coordinate
some
of
this
effort,
but
I
also
don't
control
those
allocations.
E
G
It's
just
going
to
say
that
that's
something
that
we
can
can
advocate
for.
Ultimately,
this
will
go
up
to
T'pol
Ozone,
who
sits
on
the
board
and
is
very
supportive
of
this
being
successful,
and
it's
it's
still
early
in
the
year
as
far
as
us,
having
been
acquired
by
VMware
and
making
these
sorts
of
requests.
But
that's
something
that
I
can
talk
with
Paul
about.
G
A
I
know,
but
I
think
we
need
to
talk
to
Paul
like
we
need
to
have
a
direct
conversation
with
board
members
and
stuff,
maybe
involved.
If
he's
on
the
board,
maybe
Thomas
could
could
be
the
liaison,
because
you
understand
our
concern
right.
We
built
something
based
on
assumptions
we
had
about
how
this
project
would
go.
A
These
assumptions
have
changed,
we're
looking
at
a
lot
of
a
lot
of
engineering
investment
that
has
gone
in
on
our
side
and
and
it's
basically
going
to
be
ignored
unless
unless
we
can
find
a
way
to
integrate
it,
unless
we
can
find
a
way
for
for
this
work
to
be
useful
to
the
pivots
that
are
working
on
the
court
on
the
on
the
component
teams,
this
the
the
work
will
just
be
left
and
it'll
be
a
weird
side.
Project
I
can't
have
that.
C
Well,
I
mean
I,
don't
let
me
just
say
some
things.
Firstly,
I.
Don't
think
that's
really
like
I
mean
I,
think
we're
gonna,
try
and
reuse
like
in
all
cases,
we're
gonna,
try
and
reuse
as
many
of
the
pieces
as
we
can.
Whatever
happens
like
I,
don't
think
it's
like,
let's
throw
all
the
way
thing,
but
also
should
always
be
throwing
things
away.
C
We're
like
this
whole
project,
let's
just
like
say
out
loud
like
this
whole
project,
it's
much
more
difficult
than
it
should
be
because
we
held
on
to
Diego
for
n
months
longer
than
we
should
be,
and
we
can
disagree
about
how
what
the
magnitude
of
n
was
right.
It
wasn't
zero
right
and
culturally.
What
we
should
never
do
is
feel
we
spent
a
lot
of
engineering
F
on
that.
Therefore,
we
should
continue
spending
engineering
F
on
that.
If
there.
C
A
project
where
we
shouldn't
do
that.
It's
this
one,
no,
no
and
and
I
think
that's.
Why
we're
like
at
least
pleasant?
It's.
Why
I'm
a
little
kind
of
like
them
more
kind
of
nervous
about
this
kind
of
thing,
because
I
think,
if
we
hold
on
to
some
of
the
kind
of
things
like
boxing
cetera,
it
does
stand
in
the
way
of
building
a
thing:
TL,
TL,
DR
I'd,
much
rather
waste
the
effort
of
cube,
C,
F
and
maybe
FF
of
the
whole
thing.
C
C
That
is
the
star
by
saying
here's
what
we
want
to
get
to
right
and
then
knowing
that
we
want
to
get
to
not
from
releasing
to
like
what
we
have,
because
what
should
it
look
like
what
we're
building
and
then
figure
out
where
the
pizzas
can
come
from
to
build
that
in
the
most
expedient
way
possible.
That's
what
we
should
do
and
then
maybe
use
as
much
as
you
can
well.
C
A
I
want
I
want
to
take
issue,
I
mean
cuz,
I
you're,
preaching
to
the
choir
in
one
way
and
I.
Don't
think
you
understand
that
you're
preaching
to
the
choir
and
I
want
to
take
issue
with
what
Shannon
has
said
that
I,
don't
we
don't
understand,
sunk
cost
fallacy?
Some
of
our
team
has
thrown
out
an
entire
Cloud
Foundry
distribution
three
times
I'm
on
this,
and
it's
getting
to
be
a
little
tyrant
like
we
have
thrown
out
so
much
code.
That
has
not
made
it
upstream,
but
I'm
really
really
loath
to
do
that
again.
A
So
that's
why
I'm
trying
to
find
a
path
where
we
can
and
I
don't
mind
getting
rid
of
the
stuff
I
just
don't
I
mind
getting
rid
of
the
stuff
before
anyone's
tried
it
or
anyone
who
used
it
or
anyone's
really
realized
that
it
works.
That's
what
I'm
getting
at
here!
So
it's
like
eat
your
vegetables.
Just
try
it!
You
know
just
just
try
it
and
if
you
don't
like
the
taste
of
it
at
least
tell
us
what
taste
is
wrong
and
see.
Maybe
why
we
put
those
spices
in
like
there
is.
A
There
are
reasons
for
all
of
the
weirdness
that
you
might
see
and
see
if
operator
and
I'm
I'm
projecting.
Maybe
nobody
thinks
see
if
operators
weird
but
I
suspect
some
people
do
then
find
it
overly
complex.
I
want
to
know
what
the
parts
are
that
are
particularly
objection.
Well,
we
never
we've
gotten
into
a
lot
of
philosophical
discussions
here.
H
Think,
like
the
see,
if
operator
is
kind
of
a
wash
director
replacement,
although
it
can
delegate
most
of
the
work
to
communities,
I
think
it
naturally
doesn't
provide
that
many
assets.
If
you
decide
to
leave
wash
behind
right,
even
things
like
what
was
billed
around
stateful
sets.
It's
like
most
of
the
CF
components
shouldn't
be
stateful
sets
yeah
yeah.
There
are
just
naturally
not
that
many
acids
that
you
can
like
rescue
over
into
a
world
that
has
abandoned
right,
I.
C
Really
want
to
say
this
right,
I'll
feel
bad.
If
you
get
anything
without
saying
this,
I
want
to
say
this
out
loud.
There
is
not
a
criticism
in
my
head
about
boobs,
yet
it
is
exactly
what
you
would
build
if
you
start
from
the
correct
assumptions
that
were
the
correct
assumptions
when
we
started
building
it,
that's
right,
I,
don't
have
quit.
I
just
think
the
assumptions
behind
it
specifically,
but
it's
that
we
can't
move
the
things
directly
off.
C
Watch
to
something
natives
are
now
me
right,
like
most
of
the
things
should
be
deployments
and
say
four
sets
and
config
maps
and
I'm,
not
even
sure
we
just
have
ytt
to
be
doing.
Templating
I
actually
think
we
should
probably
have
an
operator,
but
I
think
the
operator
should
start
from
that
simple
thing
and
be
a
non
required
thing.
On
top
of
it.
I
don't
think
yeah,
like
I,
don't
think,
operates
with
the
priming
way.
It's
not
promoting
teenagers,
not
parents
on
the
sto
or
anyway,
we
were
tell
me
any
other
piece
of.
C
There
are
operators
for
those
things
and
we
should
have
an
operator
and,
in
my
opinion,
the
coop
CF
team
are
by
far
the
most
qualified
team
which
we
like
that
operator.
There's
many
of
the
pieces
from
on
cube
CF,
but
nobody
on
any
team
could
be
right.
18
Bosch
manifests
in
2024,
Venezia
and
no
baby
on
any
team.
It
should
need
the
operate
that
different
to
deploy
that
component.
But
those
are
my
maintenance
for
what
it's
just
a.
F
C
F
The
thing
about
cube
CF
is
that
it's
not
tied
to
Bosh,
it's
just
a
home
track
right
now
it
has
a
deployment,
it's
a
deployment
manifest
in
it,
but
that's
just
a
llamo
file.
You
were
placing
the
database
now,
so
it's
no
longer
about
Felice
and
Irina
they're.
Both
in
progress
team
is
working
on
ireenie
and
we're
working
on
the
database.
And
after
that
we
move
on
to
something
else.
So.
C
A
A
Why
don't
we
call
it
that
a
do
not
mean
because
I
think
there's
also
interest
in
because
there's
this
big
community
of
releases
of
Bosch
things
that
are
now
Sara,
Lee,
CFTR
and
right
now,
the
what
we
currently
call
CF
operator
is
required
to
properly
run
cube
CF.
But
we
we
intended
not
to
be
the
case
in
the
future.
We
intend
a
much
smaller
operator
to
be
in
the
mix
that
is
just
providing
the
intelligence
that
is
required
to
babysit
a
Cloud
Foundry.
C
I
mean
I
think
there
should
be
a
bus
operator
as
in
the
thing
that
just
deploys
generic
cost
manifests
and
that's
what
we
have
right
now
and
I
think
some
of
the
hesitation
people
have
about
that
being
the
generic
CF
operator.
If
we
want
the
last
very
CF
focused
operator,
that
is
actually
an
operator
thoughts
yet
which
made
well
emerge
from
what
we
already
have
whatever
but
needed,
and
you
know
from
a
design
point
of
view
would
be
bad
enough
to
buy
I.
Think
we
all
seems
you
don't
both
of
them.
A
I
Mean
we
choose
Cloud
Foundry
more
as
the
community
are
not
so
much
as
the
software,
so
it's
an
operator
for
the
Cloud
Foundry
community
and
not
so
much
for
fondly
the
application.
One
time
sure
we
could
rename
to
borscht
and
we
do
have
some
pieces
which
are
not
wash
on
specific,
which
are
not
really
on
the
for
now,
but
most
of
it
deals
with
force
because,
as
you
said,
the
assumption
was
there's
a
published
API,
there's
a
publish
set
of
a
claret
of
manifests
in
the
community,
and
this
is
false
right.
I
So
the
decision
to
drop
everything
Bosch
like
the
build
system,
Monty
the
manifest
and
everything
this
display.
This
destroys
a
huge
part
of
the
tour
chain
I
for
everybody,
especially
so
for
the
operator
which
was
built
to
be
compatible
to
that
tool
chain
right.
Most
of
it
is
just
translating
it
and
chiku
Benitez
that
are
the
parts
which
can
go
away
in
the
future.
Once
we
reach
zero
Bosch,
which
might
be
like
never,
maybe
there
will
always
be
a
little
bit
of
Porsche
on
the
Cloud
Foundry
application
run
time.
I
E
Think,
as
we've
been
looking
at
the
systems
that
are
getting
replaced
by
parts
of
the
kubernetes
ecosystem,
whether
that's
core
kubernetes
itself
or
things
like
SEO
I
mean
we're
seeing
a
much
less
complicated
system
sitting
on
top
of
those
building
blocks
within
the
community.
That's
why
we've
been
much
more
enthusiastic
about
kind
of
starting
again
from
scratch,
just
saying
like
what?
E
But
they
would
run
just
fine,
as
deployments
with
some
like
out
of
bands,
job
coordinating
database,
migrations
or
something
like
that,
and
so,
like
there's
I,
think,
there's
there's
a
lot
of
constraints
that
Bosch
has
imposed
that
have
gone
into
those
component
operations
that
are
then
getting
reflected
in
CF
operator,
Bashaw
operator,
because
it's
starting
from
those
artifacts
and
getting
them
to
try
to
work
the
same
way.
Yeah
and
so
I.
E
There's
been
some
discussion
on
the
relevant
proposal
for
integration
about.
Should
this
just
use,
cube,
CF
and
I
think
it
might
be
more
valuable
to
have
that
as
a
separate
kind
of
nurturing
arena
for
those
component
teams
to
try
and
just
build
this
up
from
scratch
again,
and
we
can
then
try
and,
like
I,
think
it'd
be
productive
to
share
the
learnings
in
the
community
with
both
approaches,
because
we'll
learn
more
about
the
CF
operator,
components
and
there's
lower
building
blocks.
E
That
might
be
actual
pieces
of
your
reducible
complexity
that
we
have
to
deal
with
in
the
system
and
those
could
inform
how
much
additional
complexity
has
to
go
into
that
more
direct
approach
and
then,
as
we're
bootstrapping
like
one
of
the
main
problems
that
a
lot
of
the
component
teams
have
been
having
as
they've
been
trying
to
work
more
directly
with
kubernetes
packaging
and
deployments.
Is
that
it's
been
hard
for
them
even
to
get
a
minimally
complete
system
up
and
running
for
integration
just
so
they
can
iterate
fast.
E
This
is
something
they've
been
struggling
with
for
months
and
if
my
gut
says
that
doesn't
come
from
cube,
CF
immediately
like
that,
we
can
get
something
smaller
off
the
ground.
With
these
relevant
efforts
and
the
proposal
that
psy
has
shared,
and
as
that
that
keeps
iterating,
we
can
work
with
the
cube
CF
project
to
incorporate
those
components
as
they
get
matured
into
that
yeah.
F
H
For
what
it's
worth,
we
we've
done:
experiments
thrown
out
ireenie
part
of
cube
CF
and
replaced
it
with
home
we've
thrown
out
even
like
the
controller
stuff
and
at
least
got
like
a
working
CF
login
with
actually
the
copy
communities
release,
replacing
parts
from
from
cube
CF,
and
we
I
think
just
today.
Somebody
finish
to
actually
attach
CF
kubernetes
networking
to
a
CF
operator
based
thing
and
you
could,
like
you,
had
both
routes,
but
basically
we
kept
the
go
router
in
because
it
also
does
the
system
component
routing.
H
A
And
likewise,
where
we're
gonna
be
doing
some
of
that
as
well,
that's
great
to
hear
that
you've
done
it
independently.
Okay,
so
maybe
it's
not
as
screaming
and
emergency
as
I
thought
it
was,
and
it's
not.
There
are
places
where
the
existing
projects
can
touch
each
other
and
try
things
out.
I
felt
like
I
just
needed
to
vent
there.
So
thank
you
for
tolerating
that
in
the
meeting
I
don't
want
to
derail
the
conversation
anymore,
because
I
know
there
was
other
issues
that
we
wanted
to
talk
about.
Yeah.
E
One
other
thing
that
I
wanted
to
say:
I
I
could
potentially
see
longer-lived
shim
roll
4cf
operator
in
a
cfar
distribution
depending
on
whether
any
of
these
component
or
subsystem
changes
would
end
up
introducing
significant
behavioral
changes.
E
So
one
example
might
be
transitioning
from
go
router
to
you
sto,
so
Guerrera
certainly
has
some
features
that
aren't
like
we
haven't
figured
out
how
to
re-implement
an
ST
or
that
might
not
make
it
about
services,
maybe
being
one
example,
and
if,
if
you
want
to
keep
that
ingress
routing
here
for
application
workloads,
maybe
it's
going
to
be
more
valuable,
longer
term
to
keep
the
go.
Router
and
I.
Don't
think
that
there
are
current
plans
to
repackage
that
frigates
distribution,
maybe
that
would
be
relatively
simple
to
do.
G
I
don't
know:
Eric
is
closest
to
working
with
many
of
the
teams
and
fewer
so
we're
doing
the
kids
native
packaging.
So
I
don't
want
to
speak
too
much
about
that,
but
it
will
say
like
everyone's
working
on
this.
This
is
kind
of
their
each
each
of
the
project.
Teams.
I
think
this
is
their
highest
priority.
To
figure
this
out,
so
I
I
have
high
confidence
that
we'll
do
this.
G
Have-Have
have
something
here
like
in
the
next
month
or
two
to
to
show
and
then
maybe
by
mid
year,
it'll
be
more
robust
but,
like
that's
that's
all
very
variable
rate
as
Eric
was
describing
like
it's
not
going
to
have
full
functionality
natively,
because
things
like
route
services,
things
like
container
to
container
networking,
etc.
So,
hey
I,
think
for
that
web,
app,
12
factor
where
that
we
can
tell
a
pretty
convincing
story,
hopefully
in
the
next
couple
of
months
and
more
robustly
by
mid-year
I,
think,
that's
our
hope
and.
A
This
allows
us
to
take
a
take,
a
slightly
different
approach
where
we
evaluate
each
component
for
its
ability
to
pass
the
tests,
its
feature
parity
with
the
Bosch
component,
and
then
we
decide
to
call
something
production
ready
when
our
QA
says
its
production
ready
it's
on
a
component
by
component
basis,
which
is
why
we're
going
down
this
path.
To
that
end,
I,
don't
know
whether
we
should
segue
to
talk
about
ireenie
feature
parity
with
diego
from
a
performance
perspective.
A
A
That's
an
excellent
suggestion:
I'll
take
that
back
and
and
see
if
we
can
find
somebody
that's
appropriate.
Okay,
I
would
I
would
say
that
that
and
we
would
also
consider
putting
someone
forward
for
for
not
only
release
integration,
but
also
for
project
management
or
program
management.
If,
if,
if
there
was
a
need
for
that
again,
I'd
have
to
talk
about
resourcing
with
you
know:
Jeff
and
people
up
his
team,
but
it's
it's
something
we
could
consider
I.
H
Think
at
least
this
the
program
management
part
would
add,
for
everybody
outside
of
the
amber
alert
politicals
some
amount
of
confidence
on
some
amount
of,
let's
say,
transparency,
to
see
what's
happening,
because
my
feeling
is
right
now
the
only
people
who
can
really
cool
enough
for
like
well
enough
understand,
what's
going
on
in
the
individual
teams,
are
working
with
those
teams
turrent
right
from
the
outside.
It's
pretty
hard
to
get
a
good
understanding,
even
when
looking
into
tracker
like,
what's
really
happening,
which
kind
of
decisions
are
being
done
in
terms
of
hey.
H
This
is
some
kind
of
functionality
we
are
currently
not
targeting
or
not
looking
at
in
order
to
make
things
easier
and
so
on
so
yeah
I'm
not
sure
like
what
the
solution
to
this
is.
But
at
least
that's
my
feeling
that
we
are
a
bit
like
on
the
outside
and
it's
terribly
hard
to
to
figure
out.
What's
going,
and
maybe
it's
this
meeting,
maybe
it's
a
different
meeting.
Maybe
it's
a
dashboard,
maybe
it's
something
else.
Maybe
it's
the
first
I
don't
know.
F
F
F
A
But
so
we
do
need
to
find
somebody
and
it
doesn't
have
to
be
from
Sousa,
but
I
get
us
point
that
maybe
one
of
the
other
partner
companies
as
well
I
have
to
take
this
back.
I
mean
this
is
we
do
have
a
couple
people
that
might
be
suitable
in
terms
of
what
their
current
workload
is
with
cloud
application
platform
I'd
really
have
to
have
a
chat
with
Jeff
offline,
and
then
up
is
chaining
up
my
chain
and
and
figure
out.
So
I
can't
tell
you
right
now.
B
Like
Marco,
try
to
like
talk
about
the
timeline
aspect
or
bring
bring
up
again,
Daniels
questions
around
timelines
like
and
and
initially
we
discussed
this
this
document.
So
so
is
that
also
part
of
such
a
document
to
kind
of
try
and
document
those
timeline
assumptions
kind
of
knowing
that,
obviously
it's
not
something.
That's
that's
carved
in
stone
either,
but
like
having
that
timeline,
do
you
want
things
I.
G
Always
like
to
start
with
some
clear
milestones
and
then
like
that
that
people
generally
understand
as
what
those
milestones
are
and
then
and
then
from
there
were
work
through
them.
The
timelines
are
general
projected
hopes
and
dreams
and
and
kind
of
things.
So
if
we
can
somehow
collaborate
on
what
are
some
some
milestones,
people
are
interested
in
or
seem
reasonable,
I
think
that
would
be
a
good
place
to
start
and
then
and
then
let
the
lens
of
timelines.
There.
B
On
that
particular
topic,
so
Troy
is
that,
like
starting
that
document,
something
that
you
would
be
taking
as
an
action
item
or
when
I
have
such
a
document,
but
okay
I,
guess
we
didn't
really
say
how
we
come
to
you
like,
not
actually
who's.
Writing
it,
but
like
sometimes
it
just
helps
to
say
here
is
that
empty
page
everybody
please
contribute
their
thoughts,
ideas.
A
G
A
G
A
G
B
It
doesn't
seem
to
be
the
case,
so
yes
thank
you
Matt
for
now
and
we'll
talk
again
in
two
weeks
and
probably
Jules.
Maybe
like
you,
you
could
take
a
look
at
like
the
the
upcoming
agenda
vote,
because
the
likelihood
is
high
that
somebody
wants
to
see
and
irineu
damo.
At
least
that
was
like
also
highly
highly
voted
for.