►
Description
wasmCloud is a platform for writing portable business logic that can run anywhere from the edge to the cloud, that boasts a secure-by-default, boilerplate-free developer experience with rapid feedback loop.
A
A
B
Thank
you,
sorry
yeah.
We
initially
started
off
with
a
rust-based
operator
and
we're
wondering
about
whether
that
was
the
best
approach
to
go
or
not.
So
I
created
a
poc
to
well
a
scaffolding
out,
I
suppose,
to
actually
go
through
and
look
at
whether
which
way
we
really
wanted
to
go
in
terms
of
direction.
B
B
B
So
when
I
look
at
the
test,
this
is
kind
of
how
this
is
not
a
complete
implementation
of
the
oem
spec
and
what
david
and
I
you
settled
on-
was
actually
to
use
the
orem
spec
in
total,
as
opposed
to
or
it's
an
entirety,
as
opposed
to
picking
and
choosing
pieces.
So
I
started
off
the
path
of
picking
and
choosing
pieces
because
I
wasn't
sure
whether
we
wanted
to
track
all
the
drift
that
would
happen
in
the
spec,
whereas
we
only
use
a
subset.
B
B
I
am
able
to
create
the
application,
and
I
and
then
I
eventually
come
back
and
actually
look
that
it's
actually
created
and
that's
really
the
gist
of
it
and
then
for
the
actual
controller
or
the
reconcile
loop
is
really
very
simple.
It
just
goes
through
verifies
that
you
have
a
paid
load
and
then
runs
it.
So
if
I
go
and
run
this
test,
which
is
in
here
we'll
actually
just
see
a
complete
dump
of.
B
So
that's
really
what
we
settled
on
in
terms
of
what
to
actually
do
and
then
the
pr
actually
takes
both
concepts
and
merges
it
together.
So
this
is
where
we
are
right
now
in
terms
of
what
we've
settled
on
in
terms
of
using
om
or
not
using
om.
So
here
in
this
deck,
we
have
chosen
to
use
the
complete
specification
as
opposed
to
using
a
subset.
B
So
some
of
the
comments
that
I
actually
had
was
what
pattern
we
really
wanted
to
follow,
because
we
want
to
make
sure
that
we
are
able
to
add
and
remove
the
resource
as
we
go
along,
and
that
requires
a
finalizer-
and
I
guess
that's
not
in
here
yet
in
terms
of
code
and
we'll
have
to
follow
up
on
exactly
what
david
and
are
you
actually
thinking
about.
There.
C
B
If
you
don't
say
that
you
have
a
finalizer
which
essentially
says
that
the
operator
has
something
on
the
x
on
the
outside,
that
it
needs
to
monitor
and
or
that
it
really
depends
on
to
do
its
work.
And
then
that's
what
its
external
world
view
is
and
in
fact,
and
that's
what
it
really
needs
to
reconcile.
D
B
D
C
C
That's
that
you
need
in
order
to
delete
the
application,
which
would
be
the
name
and
the
version
that
sounds
perfect
yeah,
I
think
that's
great
and
then
there'll
be
other
there'll.
Be
some
other
functions
that
you
can
access
through.
Nas
topics
like
you'll
be
able
to
query
the
materialized
view
that
has
the
observed,
lattice
state,
so
you'll
be
able
to
get
the
low-level
state
as
well
as
high-level
state,
where
you
know
we'll
be
monitoring
the
status
of
individual
spread
scalars
and
things
like
that.
B
C
B
So
one
option
here
is
to
actually
do
like
when
you
do
a
result,
you
can
actually
give
it
a
time
period
that
basically
reaches
the
requisite
reconciler
and
if
that's
the
case,
you
can
basically
it'll
run
through
the
whole
loop
over
and
over
again
I
mean
I've
done
it.
You
do
it
in
like
five
seconds
when
we're
checking
certain
things
or
30
seconds
or
whatever,
but
I
think
I
was
actually
checking
for
certain
expirations.
It
was
probably
more
like
24
hours,
but
I'm
not
sure
yeah.
C
B
Yet
been
implemented,
yeah
point
taken,
but
it
looks
like
right
now.
We
are
very
much
going
towards
the
idea
of
essentially
reading
a
config
file
right,
it's
it's
a
post
event,
yeah,
but
essentially
being
a
big
file
where
everything
is
declared
up
front,
and
I
was
wondering
whether
that
was
gonna
be
more
configurable
in
the
future,
like
looking
at
something
like
xds,
which
is
envoy
and
grpc.
Now
that
implements
or
doing
service
mesh
stuff.
But
it
is
a
protocol
so
to
speak.
B
C
Well,
I
mean
you
know,
there's
certainly
nothing
off
the
table
right
now.
The
the
plan
is
to
essentially
get
the
the
mvp
running.
So
if
the
mvp,
if
the
the
shortest
path
to
being
able
to
get
something
in
people's
hands
that
they
can
actually
play
with
and
test,
is
just
having
everything
as
a
simple
single
declarative,
yaml
file,
then
we
can
start
there
and
then
move
to
something
more
complex
as
necessary.
C
C
C
Yeah,
that's
the
trick,
is
I
don't
want
to
the
lattice
controller
needs
to
remain
agnostic,
so
it
can't.
It
can't
do
things
that
are
specific
to
kubernetes
or
specific
to
envoy,
but
we
can
certainly
take
inspiration
from
how
other
products
are
doing
things
so
yeah.
You
can
share
a
link
to
that.
Yeah.
A
Along
the
lines
of
some
of
the
decisions
we're
making
out,
I
can
already
hear
a
road
map
kind
of
emerging
here
on
you
know
what
is
the
mvp
for
this
look
like
so
my
first
question
is:
is
it
worthwhile
for
us
to
maybe
define
a
couple
of
you
know
what
would
be
in
scope
for
mvp
versus
what
might
be?
You
know
like
a
later
release
to
include
xds
and
you
know
secret
store.
I
don't
know
you
know
something
along
those
lines
as
we
move
forward.
A
Does
that
make
sense
for
us
to
kind
of
define
that
a
little
bit
and
make
sure
that
we
can
scope
the
work
or
you
know,
is
the
work
just
going
to
happen
when
the
work
happens,
and
I
guess
the
tracking
the
work
is
a
little
less
critical
and
maybe
steve
if
you're
yeah
you're
on,
if
you
don't
mind,
maybe
speaking
to
that
a
little
bit
you're
sort
of
our
you
know,
manager
on
on
the
on
the
cosmonaut
side.
Here.
E
A
Yeah,
maybe
it
makes
sense
kevin
to
make
sure
that
your
mvp
lines
up
with
the
operator
mvp
from
a
feature
set.
It
may
be
worthwhile
less
spending
a
little
time
here
or
someplace
else.
Just
lining
up
the
notes
or-
or
this
is
a
question-
do
you
feel
that
the
rfc
process
you
guys
have
kind
of
done-
has
enough
detail
here
that
you
guys
are
going
to
line
up
and
on
the
feature
sets.
C
Well,
I
think
that's
actually.
What
we're
doing
right
now
is
having
these
discussions
in
the
community
meeting
and
in
slack
and
on
pull
requests
and
so
on.
Where
you
know
we
will
get
the
the
bare
minimum
functioning
in
order
to
in
order
to
start
being
able
to
have
a
fully
integrated.
You
know
test
between
the
kubernetes
operator
and
the
lattice
controller,
and
then
you
know
once
once
the
mvp
for
both
of
those
is
functioning.
A
B
A
So
story
you,
you
know
your
team
with
the
great
demo
you
guys
have
been
working
on.
I
think
really
prioritized
this
work
on
the
roadmap
a
little
bit.
You
know
what
do
you
do
does
where
we
are
now
sound
like
we're
aligned
with
you
know
something
that's
going
to
plug
into
the
demo
well,
but
you
know
to
support
that
work
going
forward
first
and
then
my
second
ques.
Let
me
stop
here
and
ask
my
second
question.
D
I
mean
no,
I
think
it
definitely
does
in
that.
In
that
demo
we
had
a
script
loop
running
in
the
background
was
that
was
so,
as
we
brought
up
a
new
host,
it
knew
what
to
do
yeah
this.
This
will
definitely
fix
that
and
I
think
we're
reasonably
well
aligned.
Now.
I
think
the
missing
bit
for
david
and
aisha
this
morning
was
like
how
do
you
represent
to
create
a
creation
of
the
resource
and
the
deletion
of
the
results,
but
I
think
the
dot
dell
and
dot
poll
dot
put
is
great.
A
Okay
and
then
then,
my
next
question
is,
and
I'm
definitely
not
trying
to
boil
the
ocean
here,
it's
really
just
more
of
a
fit
the
turtles
in
my
head.
I
it
would
kuvela,
then
be
able
to
drive
the
kubernetes
blossom
cloud
operator,
which
could
then
drive
the
lattice
controller,
which
would
then
drive
the
lattice
to
drag.
Wasn't
cloud.
Is
that
the
you
know,
coupe
vela
feels
like
broad
it
can
deploy
to
kubernetes
to
lambda
the
all
the
different
things
here
is
that
how
this
would
fit
together?
D
I
think
in
the
future
we
need
we
need
to
make
sure
that
we
don't
clash
with
kuvela,
but
I
think
we've
done
that
with
a
different
group
version
and
kind
for
now.
I
think
we're
planning
to
be
completely
independent
of
it.
Apart
from,
I
believe
they
are
using
a
part
of
kubella
as
a
library
to
do
the
oam
validation.
D
I
think
it
makes
sense
to
I
mean
like
that
that
that
product
came
out
of
the
cost
plain
sort
of
like
family
and
oem,
as
I
think
it's
all
part
of
the
same
thing.
So
I
think
it
makes
sense
to
use
reuse
that
code
and
I
think
that
the
version
v1
alpha
one,
which
is
what
they're
on
at
the
moment
you
know
if
they
change
that
spec
they'll
move
to
v1r
for
two,
I
guess
or
v1
beta
one
or
whatever.
So
we
should
be
protected
from
that
yeah.
A
I
I
like
it
for
a
couple
reasons
I
like,
if
we're,
making
this
investment
and
there's
a
broader
vision
in
the
in
the
overall
kubernetes
ecosystem,
that,
if
we're
importing,
you
know
that
piece
of
the
oem
model
from
them
that
we'd
hopefully
be
compatible
with
their
implementation,
so
that
we
would
plug
in
as
a
potential
provider
into
the
greater
eucharist.
For
me,
like
this
whole,
the
whole
push
to
kubernetes.
A
A
And
what
we're
trying
to
align
to
is
this
greater
cube
velostory
for
the
org,
then
you
know,
like
maybe
part
of
the
medium
term
roadmap,
should
be
to
officially
plug
in
as
a
provider
there
and
just
suggesting
that
not
trying
to
say
that's
a
requirement
or
more
just
trying
to
understand.
If
that
should
be
a
requirement
is,
is
perhaps
where
I'm
going
with
it.
D
Yeah,
I
mean
the
operator's
like
two
bits:
isn't
it
it's
like
watching
the
application
model
manifest
and
then
sending
it
to
the
letters
controller?
I
think
the
the
first
first
bit
kid
valid.
Does
that
already?
If
it
doesn't
do
the
second
bit
and
we
want
to
be
independent
of
cabela
in
case
you
don't
want
to
run
cooper,
don't
tie
into
that.
So
I
think
the
operator
needs
to
stand
a
lot
stand
alone
on
its
own,
but
it's
possible
that
it
could
also
act
as
a
like
back
end
for
kuvela.
A
E
I
wanted
to
follow
up
on
junita's
comment.
Just
you
said
that
you
wanted
the
opportunity
to
give
feedback
on
on
the
spec
and
that's
absolutely
the
way.
The
way
we
work.
So
you
the
reason
you
haven't
seen
a
spec
isn't
because
we're
gonna
like
spec
it
and
build
it
and
release
it
all
at
once.
It's
just
because
in
the
timeline
we're
not
up
to
spec
yet,
but
the
spec
doesn't
mean
it's
done
and
we
definitely
want
to
get
feedback
from
from
you
and
people
in
the
community,
so
you'll
definitely
see
it
and
get.
B
A
A
One
is
gonna,
be
a
webassembly
101
and
the
second
one
is
going
to
be
somewhat
of
a
overview
discussion
panel
around
the
challenges
facing
early
developers
featuring
our
red
badger
friends.
So
super
excited
about
those
community
events.
A
couple
other
little
notes.
If
no
one
saw
the
ideas
lab
team
just
released
their
wongi
stuff
here,
which
is
the
web
assembly
gateway
interface,
so
somebody
made
the
joke
about
wah
pearl
being
next
to
see
this
come
out,
but
I
think
what's
really
interesting
about.
A
A
So
it's
been
out
for
a
while,
but
to
see
them
actually
do
a
pr
and
some
external
momentum
on
it
means
to
me
that
they
want
to
start
doing
some
testing.
So
I
think
that'll
be
interesting
to
see
what
comes
to
fruition
there
and
then
wasm
day,
we're
just
full
steam
ahead.
Registrations
are
picking
up.
We
have
a
a
handful
of
reporters
like
a
gaggle
of
reporters,
have
all
kind
of
come
together
and
we've
got
some
good
high
profile
registrations
there.
A
A
E
Thank
you
hello.
So
we
are
testing.
Is
the
remaining
work,
we're
functionally
complete,
we're
doing
a
documentation
walkthrough?
I
posted
a
link
yesterday
on
the
on
the
slack
channel
with
draft
documentation
and
there's
a
couple
updates.
We
need
to
do
to
that.
We're
we're
pushing
releases
to
docker
hub
and
create.io.
E
A
I
tried
to
like
give
up
something
and
it
just
felt
like
I
was
in,
like
the
you
know
like
it
was
just
terrible.
I'm
just
I
mean
it
was
it
wasn't
even
close
to
god.
I
couldn't
find
a
a
reasonable
gift
like
for
the
in
the
club
or
anything
like
that
that
I
thought
we
could
hijack.
The
best
thing
I
found
was
a
sparkly,
a
50
cent
sign,
but
it
wasn't
even
that
great.
Maybe
the
price
is
right.
Oh
that's
what
we
should
do.
A
You
pull
the
pole
down
and
it
spins
the
wheel
and
then
there's
like
you
know
the
land
on
a
dollar.
We
could
land
on
50
cents.
That's
the
gift!
I
gotta
get.
Okay,
all
right
solve
my
own
problem.
With
that
one
open
floor,
anything
else.
Anybody
wants
to
bring
up
or
enough
we'll
finish
up
a
little
early
today.