►
From YouTube: wasmCloud Community Meeting - 14 Jun 2023
Description
Welcome to the wasmCloud community! Tune in live where we discuss the latest developments in the wasmCloud ecosystem, WebAssembly standards, and break out sweet demos.
Agendas for wasmCloud community meetings can be found at: https://wasmcloud.com/community
A
Yeah,
hello,
everyone
welcome
to
the
wasm
cloud
community
meeting.
Hopefully
my
connection
is
okay
enough
to
announce
this.
Just
gonna
check
on
a
couple
of
things:
I'm
actually
at
a
I'm
actually
at
a
conference
right
now,
not
at
home,
so
excited
to
be
tuning
into
this.
A
But
we've
got
a
a
pretty
wash
loaded
agenda
today,
we're
going
to
be
talking
about
two
of
the
new
experimental
flags
that
are
coming
in
and
if
you
weren't
in
the
last
week's
meeting,
we
can
also
talk
a
little
bit
about
the
experimental
flag
in
wash
and
then
we're
going
to
talk
about
the
watch.
Geoda
18
release
which
just
finally
landed
yesterday
and
as
a
whole
a
whole
Suite
of
new
features
that
we're
going
to
talk
about
and
then
I've
just
kind
of
left.
Some
time
at
the
end.
B
Says
wash
spy
on
the
agenda
so
I'm
doing
both
pppu
okay.
Let
me
go
ahead
and
share.
A
Yeah
I
know
that
we
talked
about
Washed
by
the
last
week,
but
I
love
it
so
much
that
I'm
gonna
be
happy
to
see
it
again.
I
actually
just
got
a
chance
to
use
it
yesterday
and
I
was
freaking
out
about
it.
Okay,
all
right
go.
B
B
Behold,
I
have
things
spun
up
in
case
you
didn't
know
what
am's
running
to,
but
I
was
not
spinning
this
up
with
a
damn
at
the
time,
so
I
just
spun
it
all
up,
but
with
the
Amazon
in
here
as
well.
This
is
the
new
Washington
0.18.
It
was
just
released.
Yesterday,
thanks
Brooks
and
in
case
you
didn't
see
this
last
week
we
have
our
Ecker
and
KV
counter
that
are
running
and
I'm
going
to
show
two
different
things
that
are
that
are
going
on
here.
B
B
The
watch
experimental
flag
is
something
we
started
introducing
for
things,
we're
not
entirely
sure
good
ideas
yet,
but
we
want
to
get
out
there
for
people
to
try,
and
so
you
just
have
to
put
put
on
if
you
try
to
run
this
command
without
doing
the
wash
experimental
equals
true
you're,
going
to
end
up
with
it.
Just
saying
sorry,
you
need
to
set
the
flag,
so
what
I'm
going
to
do
is
I'm
going
to
enable
watch
capture
so
now
I
have
what
it's
actually
doing
underneath
the
hood
in
case
you're
curious.
B
Is
it's
creating
a
capture
stream
right
here?
It
actually
look.
It
looks
like
it's.
It
was
already
there
from
before,
and
this
sets
up
basically
a
sliding
window.
So
the
idea
behind
watch
capture
is
that
if
you're
familiar
with
gaming,
you
often
have
a
button
that
you
can
press
that
can
capture
the
last
like
30
seconds
or
60
seconds
of
video
video
that
you
did
so
you
can
share
that
video
or
that
clip.
B
So
you
know
you
get
that
nice
head
shot
playing
Halo
or
whatever,
and
you
you
grab
that
video.
That's
what
generally
happens
so
the
idea
was
well.
Why
can't
we
do
that
with
for,
like
debugging
purposes
with
Watson
cloud,
so
this
sets
up
something
by
default
as
a
sliding
window
of
an
hour,
and
it
captures
all
the
different
indications
that
you
set
up.
You
could
also
set
it
up
on
your
own
to
capture
your
own
invocations
if
you
wanted,
but
that
captures
every
event.
B
That's
going
on
we'll
come
back
to
that
once
I
demo
Washed
by
one
more
time
so
I'm
gonna
go
ahead
and
watch
Spy
on
KB
counter,
so
I
don't
have
to
specify
an
actor
ID.
B
It
already
knows
that,
like
hey,
I,
have
a
KV
counter
actor
running,
and
so
now,
if
I
come
and
I
Purl,
the
API
counter
right
here
so
I'm,
just
I'm,
just
calling
those
things
and
you'll
see
that
we
see
from
the
HTTP
server
to
the
KB
counter
from
the
KB
counter
to
the
rediskey
value
store,
and
we
can
see
the
body
that's
going
back
and
forth
here,
so
I'm
able
to
watch
what's
going
on.
B
It
helps
with
some
debugging
eventually
we'll
want
to
add
in
some
stuff
where
it
can
even
like
capture
the
responses,
because
this
is
a
request
response
pattern.
But
this
gets
us
started.
So
that's
what
wash
spy
does,
but
the
watch
capture,
which
is
the
other
one
we've
been
talking
about,
is
I'll,
go
ahead
and
hit
the
echo
actor
as
well.
B
Just
a
couple
times
so
I
have
it
and
then
you
know,
let's
say:
I
had
a
bug
with
this
and
I
want
to
capture
what
just
happened:
I
can
go
watch,
capture,
oops
and
because
it's
enabled
it'll
go
ahead
and
grab
it.
It'll
take
a
second
while
it
goes
and
grabs
all
the
messages
and
it
outputs
it
to
a
tarball,
and
we
call
it
like
just
a
wash
capture
file.
It's
just
a
tarball,
and
so
now
I
have
that
like
locally
I
have
this
watch
capture
tarball
and
there's
this
nice
command?
B
We
have
called
replay
where
I
can
take
this
this
capture
file
and
it
will
actually
play
through
all
the
events.
There's
also
a
capture
of
the
host
inventory.
In
there
we
don't,
we
don't
expose
that
you
can
unpack
and
see
the
Json
for
it,
but
this
just
replayed
all
those
messages
from
the
from
the
last
capture
window,
but
what's
nice
is
I,
can
actually
filter
based
on
like
what's
going
on.
So
if
I
want
to
filter
like
all
HTTP
requests,
I
can
say
provider
id
and
now
I'm
only
getting
HTTP
server
stuff.
B
Now,
let's
say
I'm
doing
Let's.
We
have
the
counter
I
believe
yeah.
So
let's
say
we
just
want
to
do
one
with
the
like
to
the
all
HTTP
requests
to
the
KV
counter.
I
can
come
back
here
and
say
you
know
I'm
going
to
do
this
provider
id
and
that's
actor
ready.
B
And
so
now,
I'm
getting
just
the
HTTP
request
to
the
actor
in
the
provider
id
so
I
can
also
go
through
and
say
you
know,
I
I
actually
just
want
to
check.
The
database
calls
so
I'm
going
to
come
over
here
and
I
want
to
check
everything
that
has
to
do
with
the
KV
redis
provider,
and
so
then
I'm
able
to
filter
it
down,
and
so
I
can
also,
if
I
want
to
do
this.
Basically
Interactive.
B
This
isn't
this
isn't
perfect,
we're
hoping
to
add
even
more
cool
features
to
this
in
the
future
like
make
debugging
really
easy,
but
at
this
at
this
stage
you
can
actually
like
go
through
and
filter
through
each
like
message
or
each
thing.
That's
going
on
to
to
see
each
thing
as
it
happened
so
basically
I'm
able
to
entirely
like
to
slide
like
you,
take
the
sliding
window
and
capture
whatever
I
just
found
and
Discover
it
like,
if
I
discovered
a
bug
and
save
it
for
later
debugging
purposes
so
down
here.
B
C
B
Actually,
just
going
to
do
and
delete
that
stream,
and
then
it's
not
there
anywhere.
So
there
you
go,
that's
wash
capture
and
wash
spy
any
questions.
A
And
tailor
those
two
well
we'll
talk
about
it
in
just
a
second
in
the
in
the
release
notes,
but
those
two
actually
did
make
it
into
0.18,
and
so
you
can
actually
run
them
now.
As
long
as
you
install
the
latest
version
and
add
that
experimental
flag,
you
get
a
warning,
if
you,
if
you
don't
add
experiment
it'll
tell
you
what
to
do,
but
they
are
out
now.
So
everybody
is
welcome
to
try
them
out
and
let
us
know.
A
All
right
next
on
the
agenda
is
a
demo
of
wash
Dev
another
exciting
experimental
feature
that
will
be
in
shortly.
Victor
go
ahead
and
take
it
away.
D
Yeah
thanks
Brooks
yeah,
so
we've
been
working
on
a
feature
called
wash
Dev
as
well
sort
of
in
that
same
vein
of
sort
of
debuggability
and
a
better
development
process.
So
I'll
share
my
screen
here
and
I'll
I'll
give
a
little
demo
of
what
it
can
do.
D
Okay,
so
so
here
we've
got
we're
inside
the
examples.
Examples
repo,
so
just
looking
at
the
the
basic
Echo
actor,
which
is
pretty
simple
right,
so
it
just
you
know,
talks
to
an
HTTP
server
provider
and
you
know
Echoes,
whatever
request
it
gets
back
to
whoever
sent
to
whoever
sent
it
right,
the
usual
Echo
and
so
here
and
just
a
little
demo
of
what
it
looks
like
when
you
forget
to
put
the
experimental
flag
on.
D
So
if
you
type
wash
Dev
you'll
get
you
know
a
nice
little
warning
that
tells
you
that
it's
still
experimental,
of
course
this
will
go
away
if
we
decide
to
stabilize,
but
this
is
exactly
what
what
Taylor
described
earlier.
So
then,
let's
go.
D
And
we'll
run
it,
and
so
what
happens
is
it
starts
off
and
realizes
there's
no
wash
okay
host
running
real,
quick.
So
what
it
does
is,
it
starts
a
awesome.
Cloud
instance
waits
for
it
to
become
reachable
and
then
once
it's
reachable,
you
can
see
the
logs
of
a
wasn't
Cloud
host
right
here.
So
there's
multiple
ways
that
you
can
do
this,
and
so
here
so
here
it
found
the
host
eventually
right
once
you
can
see
the
logs
from
the
host
right
above
here
and
then
what
it's
done
is
actually
started.
D
The
project
build
so
this
project
on
the
left
here.
The
code
here
on
the
left
is
being
built
continuously
by
the
process
in
the
right.
So
it
depends
on
the
configuration
you
want
to
use
so
like
sometimes
maybe
you
want
to
run
the
Watson
Cloud
instance
as
a
sub
process,
or
maybe
you
want
to
run
it
detached
like
there
are.
D
There
are
differences
in
sort
of
trade-offs
for
both
of
those
approaches,
but
those
are
all
configurable
and
and
basically
this
process,
Will
Wait,
regardless
of
how
you've
launched
a
watch,
whether
you're
detachable,
whether
you
haven't
the
process,
will
wait
until
you
make
a
change
to
any
of
the
files
in
the
repo
and
or
I
should
say
repo,
but
I
really
mean
the
folder
you're
in
right,
like
you
could
be
in
a
you
know,
a
product,
that's
not
necessarily
A
git
repo,
whatever,
and
it
makes
sure
to
reload
your
actor,
the
actor
that's
being
developed
when
any
file
changes
happen.
D
So
to
show
you
that
the
host
is
actually
running
here
on
the
on
the
right
side.
What
I'll
do
is
just
jump
to
my
browser,
real,
quick
right
now,
it's
not
load,
it's
not
pointed
at
anything,
but
we're
going
to
go
to
4000,
which
is
the
usual
port
for
awesome
cloud
and
we'll
see
that
the
echoactor
is
actually
running.
So
at
this
point
it's
already
actually
built
the
local
actor
I
had
pushed
it
up.
It's
running
in
awesome,
Cloud
I
know
that
echoactor
needs
an
HTTP
server.
D
So
what
I'm
going
to
do
is
I'm
going
to
run
one
real
quick.
So
what
I
did
was
I
started
a
provider
from
the
registry
and
use
just
using
the
default
options
right,
so
the
default
options
by
default
you'll
have
port
8000.
D
as
the
place
where
requests
will
go
to
so
I'll
refresh
that
here
and
we
can
see
if
I
zoom
all
the
way
in
that
we've
got
the
request
that
was
sent
over.
D
Right,
if
we
add
to
the
query
string,
we
get
we
get
the
query
string
back
so
Echo
is
working.
All
of
this
is
sort
of
normal.
At
this
point
right,
this
is
what
you
could
have
done
manually
quite
easily.
So
what
I'll
do
here
is
I'll
jump
back
to
the
terminal
and
we're
going
to
just
make
a
little
change
here.
D
So
I'm
going
to
add
this
and
save
it,
and
you
see
over
to
the
right
that
we've
already
started
in
well
finished,
compiling
this
this
project,
so
I'm
going
to
go
back
to
the
browser
and
I'm
going
to
just
simply
refresh
right,
and
then
here
you
see
that
extra
value
right.
So
that's
the
functionality
of
sort
of
getting
an
actor
rebuilt
locally
as
quickly
as
possible
and
having
it
be
testable
right
in
the
browser.
D
So
that's
basically
the
the
idea
behind
washdev
and
a
quick
demo
of
how
it
works
again.
This
is
still
experimental,
but
it's
you
know
a
feature
that
we
seems
likely
to
stabilize,
especially
if
we
get
good
feedback
and
that's
one
thing
that
I
think
everyone
can
help
with
is
just
trying
it
out
and
seeing
seeing
it.
You
know
what
options
they'd
like
or
you
know
how
much
this
tool
really
accelerates
your
workflow
and
that's
what
we'd
like
it
to
do
so
Rexall
pass
it
back
to
you.
A
Victor
that
I
haven't
seen
it
working
in
the
end.
That's
really
sweet.
I
have
a
couple
of
questions.
A
Oh
yeah
I,
I'm
I'll
open
up
the
floor
just
so
that
I
don't
monopolize
the
whole
time.
Anybody
have
any
questions,
comments
for
Victor.
A
Okay,
then
I
will
go
forward
with
with
my
questions,
so
this
is
really
sweet
for
developing
on
an
actor.
The
the
first
thing
that
I
wanted
to
ask
is
how,
because
I
know
you
have
it
all
how
to
you:
have
it
all
locally?
How
does
it
actually
start
the
actor
after
it
builds
so
like
we,
we
build
the
web
assembly
module
and
you
launch
the
host
kind
of
at
the
same
time.
How
is
that
is
that
pushing
to
a
local
registry?
How
is
that
getting
there.
D
Yeah
so
there's
some
functionality
from
for
starting
an
actor
actually
from
file.
So
basically
what
you
can
do
is
use
a
local
file
path,
kind
of
like
as
in
the
form
of
a
URI,
and
so
that's
how
watchdev
works
right
now,
it's
possible
to
do
other
things,
but
right
now
it
uses
the
local
file
that
you
you
build
after
you,
I
should
say,
build
and
sign
right,
because
the
the
artifacts
are
signed,
as
is
usual.
A
Okay,
so
you're
using
that,
and
then
you
said
that,
basically,
what
it
does
is
it
watches
that
file
locally?
So
as
soon
as
you
make
is
it
does
it
watch
all
of
the
files
in
the
project
like
react,
or
is
it
triggered
kind
of
when
you,
when
you
run
build?
How
does
that
work.
D
Yeah
so
it'll
watch
files
in
the
folder
and
then
based
on
those
kick
off
a
rebuild
which
will
change
the
the
generated
file
right,
there's
actually
a
a
little
bit
of
of
logic
in
there
to
turn
off
the
watching
right
during
the
building,
because
otherwise
you'd
get
into
a
loop
right,
because
the
wash
the
the
build
process
generates
a
wasn't
file.
And
if
you
wear
a
trigger
on
the
wasem
file
generated,
you
would
start
trying
to
build
again
and
never
end
so
I
ran
into
that
manually.
D
So
I
can
tell
you,
that's
the
that's
how
that
works,
but
but
yeah.
So
it's
basically
just
triggered
on
files
inside
the
inside
the
folder.
That's
running.
A
Cool
cool
awesome,
I
I,
really
like
how
it's
really
focused
on
like
developing
a
single
actor
like
it'll,
still
start
that
up
for
you
and
then
you're
free
to
kind
of
go
in.
Maybe
you
could
use
Autumn
for
this,
but
just
generally
like
put
the
providers
to
getting
links
together
just
to
like
get
that
initial
environment
up.
A
My
other
question
was:
can
you
do
this
with
a
currently
running
host?
So
that,
like
you,
could
do
this
with
multiple
actors
like
basically
have
a
wash
Dev?
That's
not
responsible
for
running
a
host
just
responsible
for
watching
an
actor
because
I
work.
D
Yeah,
so
it
actually
tries
to
just
detect
if
there's
a
host
running
right,
so
I
I
didn't
have
a
host
running
when
I
started
it,
which
is
why
it
sort
of
printed
out
that
message
that
was
like
couldn't
find
a
host.
So
gonna
start
one
but
yeah.
It
definitely
works.
If
you
just
decide
to
run
a
host
yourself
and
like
when
you
want
to
manage.
Let's
say:
yeah
see
that
was.
D
Use
case
and
I
I
do
like
the
idea
of
using
Watermark
Madame.
However,
we
like
to
say
it
as
an
orchestration
layer,
because,
obviously
the
there's
a
little
bit
of
friction
where
I
had
to
go
in
there
and
start
the
HTTP
server
provider,
but
that
could
obviously
be
done.
There's
lots
of
ways.
We
can
do
that
right.
We
can.
If
we
know
what
kind
of
thing
we're
running,
we
can
actually
just
try
and
start
it
from
from
inside
wash
but
obviously
I.
D
That's
pretty
cool
and
there's
also
the
idea
of
passing
with
Dam
file
and
then
actually
just
modifying
it
on
the
fly
right
so
like
you
can
actually
pick
spots
of
that
with
damn
file
too
pull
out
and
pull
that
and
through
and
throw
back
in
change
right
because
we're
because
it's
it's
yaml,
so
it's
all
programmatically,
you
know
easy
to
easy
to
edit,
but
all
that's
in
the
future,
and
so
right
now
for
sun
actor
is
great
and
then
we'll
see.
What's
next.
A
Awesome
I
I
can
really
see.
I
can
really
see
the
potential
for
this
one
to
be
a
quickly
unmarked,
experimental
one,
because
I
want
to
use
it
for,
for
everything.
I'm
really
I.
A
Think
probably
some
of
the
things
that
we're
going
to
need
to
think
about
will
be
the
interaction
with
wadham
and
how
we
want
to
handle
that
because
it
would
be
nice,
get
a
clean
slate
to
go
in
an
actor
repo
and
do
something
just
say,
wash
Dev
and
it
launches
the
host
it
launches
the
actor
and
watches
it
and
it
launches
all
the
providers
and
links
and
and
everything
that
you
need.
So
basically,
you
can
just
like
get
started
right
away
and
you
don't
have
to.
A
If
you
you
know,
pick
up
and
put
down
the
developer
Loop,
you
don't
have
to
restart
those
providers.
That'll
be
really
interesting,
but
but,
like
you
said,
you
know,
we
we're
totally
going
to
continue
to
add
to
this
and
go
from
there.
A
D
Even
in
the
like,
there
are
some
intermediary
steps
that
are
probably
pretty
easy,
like
just
detecting.
If
there's
a
you
know,
the
HTTP
server
capability
is,
you
know,
is
there
then
we
know
that
we
should
at
least
have
one
HTTP
server
and
we
can
do
at
least
you
know
the
bare
minimum,
which
is
just
starting
to
starting
the
provider
manually
but
yeah.
It's
pretty.
It's
pretty
awesome.
A
Super
cool
anybody
else
have
any
other
questions
for
for
Victor
on
watchdev.
A
A
Speaking
of
that,
wash
0.18
was
kind
of
something
that
we
added
a
good
bit
of
features
into,
which
is
why
I
want
to
talk
about
it
here,
but
I
think
I'd
like
to
get
0.19
out
a
little
bit
quicker
on
the
the
Cadence
from
from
0.17.,
so
I
think
that
that
naturally
leads
as
a
good
segue
into
the
next
part
of
the
call
which
is
just
talking
about
0.18.
A
So
in
the
wash
repository
now,
you
can
now
see
that
we
have
0.18
out.
This
came
with
wash
lib
0.9,
which
includes
most
of
the
functionality
that
we
added
to
this
release.
It's
just
this
is
the
CLI
and
I
noticed
something
in
our
release.
A
Notes,
actually
that
there's
a
lot
of
dependabot
output
now
in
our
generated
release,
notes
so
I
think
what
we're
going
to
be
looking
into,
and
this
may
be
a
good
topic
for
the
next
Community
call-
is
a
strategy
for
generating
these
release
notes
in
a
meaningful
way,
because
I
I
kind
of
separated
them
out
manually.
But
there
are
just
as
many
depend
about
bumps
as
there
are
features
in
this
one.
A
For
for
this,
one
I
want
to
call
out
some
of
the
some
of
the
bigger
features
that
are
coming
in
this,
including
the
flattening
of
some
of
our
more
verbose
commands
from
like
watch,
control,
link
into
wash
link,
wash
control,
start
and
stop
and
to
watch
start
and
stop
some
various
improvements
from
Roman
and
other
people
in
the
wasm
cloud
Community.
Just
like
fixing
some
of
our
build
and
user
experience
things
just
miscellaneous
things
around
the
wash
CLI,
including
wadham,
by
default.
A
Additionally,
obviously,
what
Taylor
added
around
washbuy
comes
in
with
the
experimental
flag
and
wash
capture
both
of
those
have
made
it
into
this
release
and
last
but
not
least,
really
a
thin
wrapper
by
me
around
the
great
work
that
Roman
put
together
in
the
Watson
Cloud
OTP
repository
as
wash
burritos,
which
we
keep
saying
that
and
it's
so
funny
if
it's
set
out
of
context.
A
Really
all
it
means
is
that
the
wasm
cloud
host
that's
wash
is
going
to
be
running,
is
more
static,
more
platform
independent,
so
on
Mac
and
windows
it
doesn't
have
any
platform
dependencies
and
on
Linux.
A
big
deal
is
the
fact
that
you
don't
have
to
install
openssl
anymore
in
order
to
run
Watson
cloud
is
kind
of
it's
statically
built
as
a
dependency,
which
is
awesome.
I
also
want
to
give
a
shout
out
to
Lee
who
got
their
first
contribution
in
second
and
third.
A
It
looks
like
in
this
release
it's
pretty
cool,
that
they
GitHub
kind
of
close
this
out,
but
thank
you
Lee.
Some
really
awesome
improvements
around
the
around
the
dev
container
here,
which
I
personally
have
been
using
more
and
more
often
for
just
little
small
fixes
where
I
just
want
something
to
launch,
especially
if
I'm
on
my
other
machine
so
definitely
go.
A
Give
this
one
a
try
today,
it's
available
on
all
of
our
package
managers
and
all
that
stuff,
so
you
should
be
able
to
upgrade
right
into
it
and
what
we're
going
to
be
doing
with
this
release.
We're
also
going
to
be
updating
some
of
our
introductory
tutorials
and
some
of
the
content
in
our
documentation,
because
waddam
is
generally
available
and
being
used
to
run
applications.
It
really
makes
some
of
the
things
in
our
introductory
tutorials,
like
instead
of
clicking
around
in
the
awesome
Cloud
dashboard,
just
applying
a
manifest
a
lot
easier.
A
There's
a
link
at
the
bottom
of
our
webpage,
we'd
love
to
have
you
in
the
in
the
community
all
right!
That's
all
that
I
wanted
to
talk
about
around
wash
0.18.
Anybody
have
any
questions
around
some
of
the
new
features
or
what
we've
put
in
in
the
last
release.
A
Awesome
well
I've
carved
out
a
little
bit
of
time
at
the
end
of
this
end
of
this
call,
I
know
that
we
I
wanted
the
focus
to
be
on
the
demos,
because
some
of
the
new
experimental
things
are
really
exciting
and
I
carved
out
a
little
bit
of
time
for
talking
about
the
wash
developer.
Iteration
Loop
I
honestly
feel,
like
I
I,
believe
I
queued
that
up
before
I
knew
that
we
would
be
demoing
all
the
cool
things
that
we've
been
working
on
on
the
developer.
A
Iteration
Loop
and
I
I
think
I've
gotten
to
talk
about
a
lot
of
the
the
agenda
items
that
that
I
wanted
to
talk
about
there.
Washite
Dev
really
puts
the
cherry
on
the
cake
or
whatever.
For
me,
oh
God,
I'm
I
can't
do
the
metaphors
today,
like
old
bros.
A
A
So
anyways
I'm
still
happy
to
have
a
discussion
if
there
are
more
points
that
people
want
to
talk
about
around
the
wash
iteration
Loop
Jordan
you've
been
you've
been
a
little
quiet
and
I
hope.
You
don't
mind
that
I'm
calling
you
out
here
I
kind
of
want
to
hear
your
thoughts
on
some
of
this
new
stuff,
because
well,
you've
done
a
lot
of
wasm
cloud.
Development
you've
been
through
the
developer,
iteration
Loop
a
good
bit,
especially
with
with
writing,
go
and
kind
of
just
want
to
hear
your
thoughts
on.
A
D
I
can
speak
from
two
hats
from
the
the
in
the
second
of
which
is
is
not
as
positive
from
the
developer's
perspective.
I
love
it
right.
I've
been
washed
spying
Forever
by
Matt's
sub
arrow
and
doing
it
the
hard
way
right,
so
it
it
definitely
makes
life
a
lot
easier.
D
Dev
mode
seems
cool,
I'll
have
to
play
with
it.
I,
don't
really
know
what
how
it'll
change
my
my
Loop,
but
it
seems
fun,
yeah
and
then,
and
then
replay
is
just
fun.
I,
don't
know,
I,
don't
know
what
I'm
gonna
do
with
it,
but
I'm
gonna
play
with
it
see
we'll
do
something
with
it.
D
My
problem
that
I'm,
having
comes
from
having
audited
many
companies
from
a
security
perspective
for
a
long
time
wash,
is
becoming
very
powerful
and
wash
drops
keys
in
a
lot
of
places
and
washes
very,
very
easy
to
misconfigure
making
wash
a
very
big
security
concern.
From
my
perspective,
not
only
can
It
change
out
actors
and
providers,
it
can
now
just
like
spy
and
drop
artifacts
that
can
then
just
be
egressed
off
the
machine.
D
Granted
I've
read
Taylor's
documentation
on
like
how
to
properly
lock
down.
In
Keys,
but
I
spent
a
lot
of
time
in
N
keys
and
they're
also
very
easy
to
mess
up
and
either
in
after
about
three
or
four
times
you
haven't
added
the
right
subject.
You
just
start
adding
carrots
and
arrows
and
stars
because
it
gets
frustrating.
D
D
But
those
are
really
my
big
concerns
from
a
developer's
perspective,
it's
great
from
a
security
perspective.
It
scares
me
a
lot.
A
I
think
that
makes
a
lot
of
sense
and
we've
had
a
few
inklings
of
a
discussion,
but
not
really
a
full
one
around
both
wash
and
the
waslam
dashboard.
If
they're
meant
to
be
developer
tools,
production
infrastructure
like
provisioning
tools,
production
interaction
tools
or
or
all
of
them
or
or
none
of
the
above.
If
that
makes
sense,
because
I
see
the
distinction
that
you're
you're
calling
out
washes
really
growing,
very
well
as
a
developer
tool,
and
the
power
is
probably
more
than
you
would
want.
D
C
B
B
So
when
you
get
like
your
credentials,
even
if
you
are
using
watch
which
you
technically
can
right
right
against
it,
you
can't
do
certain
things
like
outside
of
that
that
cluster,
and
so
like
it's
it's
kind
of
an
interesting
conversation,
because
we
want
to
keep
that
flexibility
around
for
people,
because
there
could
be
various
ways
you
want
to
distribute
this,
like
it's
very
different
from
a
team
that
just
wants
to
run
like
a
single
thing
somewhere,
where
someone
wants
to
rent
something
more
distributed.
So
I
I
think
I'm
trying
to
think
of
ways.
B
We
can
maybe
like
incorporate
that
and
I'm
I
know.
The
watch
lockdown
was
like
a
like
semi
like,
but
I.
Think
it's
a
really
good.
Like
I,
don't
know
if
we
caught
wash
lockdown
or
or
wash
I'm,
trying
to
think
of
a
funny
another
funny
name
for
it,
but
like
something
like
that
can
maybe
help
like
the
configuration.
So
it's
not
so
like
giving
you
the
the
foot
gun
to
shoot
yourself
with,
like
you
were
saying,
Jordan.
D
D
How
are
we
telling
people
to
do
it
in
production
right
because,
like
wash
up
scares
me
because
then
it
drops
keys
and
if
you
forget
to
clean
up
keys
or
if
you
forget,
to
copy
them
to
somewhere
secure,
then
you're
kind
of
like
Fubar,
so
I
spent
a
lot
of
time
last
two
weeks,
actually
just
deploying
things
from
the
host
binary
to
see
what
that
feels
like
and
yeah.
Obviously
it's
nowhere
near
as
nice
as
wash
up.
D
However,
it's
working,
but
nonetheless
I
get
what
you're
saying
Taylor
I
think
so,
like
washes
connectivity,
I,
don't
know,
even
if
there
was
like
some
sort
of
way
like
time
out
the
creds
right,
because
I
know
like
in
Keys
has
a
JWT
attached.
I,
don't
know
if
we're
using
that
part
of
it,
but
like
maybe
like
wash
anything
every
once
in
a
while
required
some
sort
of
re-authentication
I
know
that
kind
of
defeats,
the
purpose
of
decentralized
authentication.
But
it's
scary
man.
D
I'm
I'm
wondering
what's
the
what
kind
of
threat
model
are
you
thinking
about
here?
Are
you
thinking
about
someone
running
locally
and
doing
something
wrong
or,
like
is?
Are
you
thinking
of
running
awesome
if
you're
running
awesome
cloud
and
like
a
sort
of
distributed
situation
like
on
a
computer?
That's
not
yours,
essentially
right
just
like
not
the
one
in
front
of
you
like
a
remote
one
because,
like
when
I
think
of
the
first
case,
I.
D
Think
of
something
like
maybe
because
you
know
how
like
brew,
has
Brew
doctor
right
and
you
can
kind
of
see
a
like
a
readout
of
like.
Oh,
you
have
these
files
on
the
disk
they're,
not
right
because,
like
on
the
same
machine,
there's
there's
a
limit
to
how
much
you
can
make
how
secure
you
can
make
things
right.
It's
just
like
you've
got
file
level
security.
D
You've
got
encryption
at
rest,
but
that's
like
that's
all
you
have,
because
you
know
someone's
on
your
machine
they're
on
your
on
your
machine,
but
so
so
I
I'm
curious
because
it
like
it
changes
how
how
we
could
approach
it.
D
So
one
of
the
things
that
sticks
out
my
mind
is:
is
the
host
file
or
what
do
we
call
it
the
underscore
host
off,
because
they
it
drops
everywhere
if
I
had
a
dollar
for
everyone,
I
found
in
a
rogue
spot
on
my
machine
because
we
just
drop
them
everywhere
with
with
like
a
seed
in
it,
I
have
a
lot
of
dollars.
D
So
that's
one
and
I.
Don't
really
know
how
we
like
doctor
that
because
it
can
be
anywhere-
and
my
second
thought
is
if
the
if
wash
up
is
the
prescribed
method,
we
record
you
know
the
the
team
recommends
for
production,
which
I
don't
know
it
is,
but
if
it
is
right
and
then
there's
some
like
yeah,
someone
forgets
to
you
know
clean
up
the
wash
Keys
folder
right,
there's
a
lot
of
secrets
in
there
and
and
the
threat
model
I'm
thinking
of
is
there's
only.
D
You
only
have
to
pop
one
thing:
you
don't
even
have
to
escalate
right,
because
all
this
runs
in
and
you
know
non-root
user
place,
so
it
there
are
that's
kind
of
what
I'm
getting
at
I'm,
not
Brooks
put
me
on
the
spot.
So
my
brain
doesn't
really
work
when
I
do
that.
C
D
That
that's
kind
of
what
I'm
thinking
right,
having
audited
more
networks
and
I
care
to
admit.
In
my.
E
D
So
yeah,
however,
from
a
developer's
perspective,
I
love
all
this
I'm
wondering
if,
like
so
there's,
there
are.
Obviously
there
are
levels
right
to
to
like
isolation.
How
you
do
it
right
and
you've
got
like
containers
as
a
first
step,
then,
like
virtual
machines,
is
the
second
step
right
like
if
you
want
to
really
protect
the
process.
There's
only
so
many
like
ways,
you
can
do
it
right.
D
There's
like
you
know,
then
you've
got
like
the
old
school,
like
user
level,
sort
of
resource
setting
and
stuff
like
that,
I'm
wondering
if
that
so
there's
there's
the
one
problem
of
just
like
of
configuration,
could
be
everywhere
and
shouldn't
ideally
be
everywhere
right,
and
maybe
we
shouldn't
automatically
create
it
right,
like
in
a
production
environment
like
maybe
just
fail
immediately,
rather
than
dropping
in
on
disk
or
pull
it
from
somewhere
secure
right.
Those
are
all
like.
Those
are
all
features.
D
So,
like
that's
one
thing,
that's
like
protecting
the
startup
process
or
or
sort
of
is
less
particular.
Maybe
it's
more
like
locking
down
or
sort
of
making
the
startup
process
a
little
bit
more
secure
and
then
there's
the
problem
of
actual
process
isolation.
So
it's
like
I'm
I'm,
trying
I
have
these.
These
sounds
separate
like
separate,
separate
points,
because
I
mean
all
that's
relevant,
but
it's
like
they
sound
like
separate
points
and
I'm
thinking.
D
It
sounds
like
the
process.
Isolation
is,
is
one
solution
and
then
changing
the
startup
process
and
making
that
a
little
bit
more
consistent
and,
like
you
know,
maybe
not
putting
creds
in
too
many
places
or
like
consolidated
configuration
and
and
secrets,
is
another
thing
does
that
is
that?
Do
you
think
that,
like
sort
of
accurately
represents
what
you
were
getting
at.
C
A
You're
right,
I
think
there's
the
like
Jordan
is
definitely
concerned
with
the
writing
out
secrets
to
disc
which
I
I
think
we
certainly
should
audit,
especially
around
wash
up
our
mechanism
now
of
host
config
is
it's.
It
was
nice
for
launching
multiple
hosts
before
we
had
watched
up,
but
now
we
have
it
and
I
think
we
should
probably
remove
that
mechanism
by
default.
A
It
could
be
something
you
opt
into,
maybe
around
wash
Keys
getting
generated
when
you
build
actors
and
forgetting
that
they
exist
I
think
that
would
be
a
good
thing
to
take
a
look
at
so
there's
like
the
local
configures.
You
said:
there's
actual
process,
isolation
around
wasm
Cloud,
which
is
I,
think
probably
the
most
in-depth
one
that
we
haven't
put
as
much
effort
into
and
then
actually
locking
down
a
Nats
connection.
A
I
think
that
we
have
some
basic
documentation
around
like
what
you
could
do
to
start.
Taylor
wrote
some
of
that
for
wadham,
but
that
is
ultimately
the
most
important
thing
around
locking
down
a
wasm
cloud
lattice,
it's
around
the
connection
to
Nats,
because
all
the
action
that
a
host
takes
are
in
response
to
control
interface
commands
if
you're,
starting
and
stopping
hosts
or
sorry
starting
and
stopping
actors
providers.
All
those
things,
that's
all
done
via
an
app's
message
to
the
control
interface
and
all
of
the
compute.
A
All
the
RPC
that
happens
in
a
lattice
is
done
over
the
RPC
connection.
Those
are
actually
separate
within
a
lattice,
the
the
different
connections,
so
you
can
have
a
lockdown
connection
for
the
control
interface
and
for
RPC
and
they
can
be
separate.
So
even
if
you
were
to
leak
a
key
for
the
control
interface,
somebody
could
you
know,
stop
all
your
actors,
for
example,
but
they
couldn't
Forge
in
our
PC
call
and
shoot
it
into
the
lattice
or
you
know.
A
That's
that's
kind
of
the
interesting
thing
around
it
and
I
think
it
and
I
wish
Kevin
was
here
because
I
think
he
would
have
a
good
time
answering
this
one,
but
I
think
some,
some
really
in-depth
documentation,
or
maybe
even
a
blog
post
around
like
Nats,
is
the
key
to
security
and
wasm
cloud
and
not
just
networking
would
be
really
interesting
because
it
is
very,
it's
very
locked
down
there
and
it's
difficult
to
see
from
the
default
perspective
because
we
set
up
like
here's,
a
Nat
server
and
here's
all
your
connections
go
wild,
so
it'd
be
good
to
detail
that.
B
I
do
think
there
is
room
for
like
what
I
was
saying
in
the
chat,
so
for
those
who
are
not
in
the
zoom
meeting
and
watching
this
live
stream
or
recorded
later
it'd
be
really
nice
if
we
could
like
have
like
custom
key
storage
stuff
so
by
default.
It'll
save
to
disk,
like
in
the
FI
in
like
the
washer
like
we're
doing
right
now
for
Keys,
which
I
think
we
are
saving
those
with
the
proper
permissions
at
least
they're
all
600,
that
kind
of
stuff,
but
like
it'd,
be
really
nice.
B
If
we
had
like
a
limitation
for
Vault
and
for
like
password
managers
and
or
credential
managers
on
the
system,
I'm
kind
of
like
what
like
Docker
does
at
this
point,
you
know
when
you
do
like
a
Docker
login
and
have
like
keys
and
stuff
stored
in
there
by
default
could
be
interesting,
but
also
even
if
we
just
have
those
as
options,
people
could
do,
and
it
would
be
really
good,
especially
when
people
start
using
and
deploying
bigger
production
environments
of
of
awesome
Cloud,
because
then
they
can
have
the
the
ability
to
hook
it
into
whatever,
like
kind
of
secret
vaults
types
of
things.
C
B
E
A
No,
oh.
A
Trying
to
think
around
actually
Justin,
would
you
mind
expanding
a
little
bit
more
on
that
on,
like
the
what
parts
of
like
the
auditability
an
Enterprise
customer
would
be
really
into
I'd
like
to
keep
that
going
just
a
little
short
on
ideas.
So.
E
Say
I
as
a
developer
in
my
distributed,
team
am
working
on
creating
functions
and
or
capabilities
when
I
create
a
key,
or
my
local
file
system
creates
a
key
that
key.
You
know
similar
to
how
we're
doing
like
multiple
signs
right
on
different
keys,
so
multiple
sources
of
signatures
on
that
key
I
can
then
use
those
signatures
to
trace
back
okay,
who
created
this.
You
know
it's
like
a
get
blame
for
the
development
work
that
was
done
in
my
you
know
three
node
cluster,
that's
running
across
the
us
or
something
it's
a
it's.
A
It's
all
there,
it's
a
little
complex
to
Mint
all
of
your
keys
from
some
scratch.
This
would
totally
be
something
you'd
be
interested
in
doing
if
you're
running
a
production
system
in
wasm
Cloud
or
with
wasmcloud
Taylor
I'll.
Let
you
go
ahead
and.
B
I
know:
we've
been
keeping
an
eye
on
as
maintainers
the
new
stuff
coming
out
in
gnats
2.10,
around
being
able
to
kind
of
like
issue
credentials,
Dynamic
credentials
on
a
full
on
the
Fly
based
on
things
so
like
there's
gonna
be.
Hopefully
that
story
is
going
to
be
easier,
because
that
is
one
of
the
hardest
things
about
the
this,
like
brick,
said
the
distributed
key
model.
B
There
is
very
powerful
because
you're
able
to
like
tie
things
to
very
specific
identities
and
users
and
and
really
dial
down
like
what
each
thing
has
access
to,
but
it
is
a
little
bit
more
complex
to
Mint
those
credentials
and
then
then,
like
you,
have
the
the
actual,
like
this
thing
of
like
being
able
to
Mint
like
a
temporary
credential
or
a
thing
based
on
like
another
type
of
auth
service.
That
would
make
this
make
it
really
nice.
B
So
that's
what
we're
looking
for
yeah
and
like
Brooks
is
going
on
in
the
in
the
thing
they're
like
swappable
and
key
back
into
the
keychain
access.
That's
with
all
the
kinds
of
things
I
would
love
to
see
with
this
we're
always
open.
If
someone
like
was
really
itching
to
try
this
we'd
love
to
see
something
like
that,
that'd
be
really
cool
as
like,
even
just
an
example
PR.
B
Even
if
you
don't
think
you
want
it
merged
because
you're
like
I'm
just
toying
around
that's
the
kind
of
thing
we
we'd
love
to
see
and
if
not
I
think
like
I
said
as
soon
as
we
have
someone
trying
to
actually
pilot
this
out
somewhere,
I
would
love
to
like
get
together
with
whoever's
doing
it
out
there
in
the
community
who's
starting
to
like
try
to
do
this
and
think
about
the
key
management.
B
So
we
can
start
to
maybe
design
a
solution
around
around
that
that
kind
of
idea,
because
the
the
security
model
was
in
Cloud's,
really
good
I
think
we've
talked
about
now,
Jordan
brought
up
the.
It
could
probably
be
some
better
security
in
the
like
tooling
side
of
it
that's
generating
the
stuff,
but
it
also
has
led
to
this
idea
of
I
think
we
can
make
the
management
of
them
a
lot
easier,
so
yeah
totally
down
to
do
that
and
open
for
ideas,
as
as
we
continue
to
improve
and
start
running.
A
A
I
think
it
could
be
worth
I'd
like
to
hear
a
little
bit
more
of
Jordan's
perspective
on
the
other
end
I'm,
assuming
he
like
a
bear
chewed
on
his
ethernet
cord
or
something,
and
he
didn't
just
Rage
Quit
from
something
that
Victor
said
that
we
said
around
security.
A
So
I'd
like
to
hear
more
of
his
perspective,
so
it
might
be
worth
queuing
this
up
again
for
next
week
and-
and
you
know
getting
some
more
ideas
together,
especially
if
somebody
reaches
out
on
the
key
management
space
on
slack
and
that'd,
be
neat
Victor
go
ahead.
D
Yeah
I
wanted
to
just
say
real
quick
I,
I,
assume
I
I'm,
not
speaking
for
anyone
else,
but
I
assume.
Other
people
agree
as
well
on
the
cosmotic
side
that
bad
news
is
also
good
news
right,
I
feel
like
Jordan,
was
a
little
hesitant
to
like
shared
like
something
he
didn't
like,
but
I
think
that
feedback's
often
extremely
useful.
So
if
you're
feeling
pain,
please
tell
us
because
then
we
can
stop
it.
D
We
can,
or
at
least
do
something
about
it
and
like
the
the
odd
like,
for
example,
the
auditability
call
out,
which
is
which
is
which
I
see
or
I'm
interpreting
kind
of
like
the
supply
chain,
security
right,
that's
a
that's
been
a
bit
of
a
like.
You
know:
that's
been
growingly
like
being
implemented
more
and
more
recently.
D
A
Yeah
these
these
discussions
are
that's
a
great
call
out
Victor
some
of
these
discussions,
where
you
know
it
can
be
a
good
thing
or
a
bad
thing
that
starts
at
or
some
of
the
most
valuable
discussions.
We
on
the
maintainer
side
try
to
queue
up
cool
things.
If
you
have
something
that
you'd
like
to
talk
about
like
what
Justin
brought
up,
which
I
think
we
could
talk
about
for
30
minutes
on,
a
community
call
always
feel
free
to
suggest
them.
For
the
next
community
meeting,
they're
always
open
agendas.
We
can.
A
We
can
always
modify
that
and
talk
about
things
that
people
want
to
chat
about,
especially
when
it
comes
in
contact,
Watson,
Fab,
all
right.
Everyone
I
think
we've
got
just
a
couple
of
minutes
left
in
the
community
call.
Does
anyone
have
anything
else
that
they'd
like
to
bring
up
in
terms
of
the
wasm
cloud
ecosystem,
or
maybe
a
broader
webassembly
land?
I
know
that
we've
got
a
lot
of
updates,
but
just
want
to
give
the
last
couple
of
minutes
for
for
open
time.
There.
A
All
right,
everyone
well
I,
think
we'll
go
ahead
and
call
it
there.
I
will
actually
be
back
in
my
normal
location
next
week,
and
so
I
will
see
all
of
you
for
the
Watsonville
community
meeting
on
the.