►
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.
https://wasmcloud.com
A
All
right
welcome
everybody
to
awesome,
Cloud
Wednesday
for
August
31st.
We
actually
have
a
new
member
joining
the
call
today
and
I'd
love
for
them
to
introduce
themselves
so
Carolyn.
Would
you
like
to
go.
B
Yeah,
of
course,
hello,
Hello
nice
to
meet
you.
My
name
is
Caroline
tarber,
M.A
marketing,
communications
consultant
working
with
Liam
and
team
in
a
run-up
to
the
launch,
cosmetic,
GA
and
and
to
help
generally
and
kind
of
we're
positioning
the
company
as
we
start
to
grow
so
yeah,
it's
really
nice
to
meet
you
all
and
anything
I
can
be
of
assistance
with
please
Shout.
A
All
right,
thank
you,
Caroline
for
for
doing
an
introduction,
and
now
that
we've
done
that
we
have
a
very,
very
cool,
very
exciting
demo.
Today,
that's
kind
of
building
off
of
some
things
that
Jordan
showed
last
week.
He
went
together
and
well.
I
won't
spoil
the
demo
Jordan,
take
it
away.
C
Okie
dokie
one
sec.
C
As
it
goes,
my
history
does
not
have
all
the
things
I
would
say
as
well.
There.
It
is
okay,
so
yeah
I
have
done
such
that
was
me,
focusing
so
yeah
I
played
with
what
I
had
been
doing
with
the
the
go
providers.
Last
week,
where
I
had
like
the
toy
HTTP
server
example
and
I
wanted
to
really
I.
Don't
know,
I
wanted
to
do
something
more
more
fun
with
it.
So
I
don't
know.
C
If
you
have
or
haven't
heard
of
light
stream,
it
is
more
or
less
a
tool
that
has
made
it
so
that,
like
SQL
light
databases
are
slightly
more
production,
capable
tools
for
us
to
use
it
essentially
monitors
and
streams.
Your
updates
to
your
sqlite
database
locally,
to
an
S3
of
some
sort,
so
I
think
that's
really
cool.
C
So
essentially,
I
I
took
the
walls
and
Cloud
SQL
TV
contract,
wrapped
it
around
sqlite,
and
you
know
the
function
the
functionality
needed
for
that
and
what
we
got
was
a
provider
that
will
satisfy
the
SQL
DP
contract
and
stream.
The
updates
real
time
to
an
S3
bucket
I'm,
using
just
for
this
I'm
using
the
mid
IO,
which
is
the
S3
whatever
it
works,
so
I've
gone
ahead
and
put
the
code
up.
C
I
haven't
done
that
with
the
HTTP
server
yet
from
last
week,
but
a
lot
of
the
things
that
I
a
lot
of
the
hardships
I
ran
into
was
really
around
like
once,
you
understand
kind
of
the
flow
of
it
all
right,
a
provider.
You
know
you
start
a
provider,
it
comes
in
and
you
know
it
starts
and
it
starts
listening
for
link
definitions
and
as
soon
as
it
gets
a
link
definition,
it
reads
the
stuff
on
the
con.
C
You
know
that
you
put
in
the
values
and
then,
if
it's
you
know
if
it's
validates
it
starts
the
provider,
and
then
we
just
sit
here
and
we
start
listening
for
for
requests
from
our
actor
and
once
you
understand
the
flow,
it
actually
gets
pretty
easy.
We
just
we.
This
provider
action
is
just
a
channel
listening
for
things
off
the
mats,
the
appropriate
next
topic,
and
then
we
just
start
looking.
Listen
there.
You
know
sorting
by
the
the
operations
that
satisfy
the
that
are
defined
by
their
wasn't
Cloud
contract.
C
You
know
and
I
believe
the
SQL
DB
has
an
execute
operation,
a
query
operation
and
a
ping
operation
that
is
not
currently
at
limit
yeah
and
the
real
difficult
piece
around.
All
this
was
actually
learning
the
encoding
right.
The
encoding
wasn't
probably
uses
message
packed
and
seaboor
in
some
places.
C
I
don't
really
know
what
Seaborn
coding
is,
but
there
was
enough
commenting
in
the
code
to
help
me
figure
it
out,
and
while
you
won't
see
any
of
that
in
the
provided
directly,
that's
because
I
I
wrote
a
little
the
rapper
library
that
does
most
the
encoding
for
me.
So
this
is
all
pulling.
You
know
doing
all
that
for
us,
and
it
is
straight
out
of
the
wasn't
file
time
to
get
a
message
back
in
wasn't
Cloud
time
to
go,
see
where
so,
there's
nothing
special.
C
This
code
can
be
done
directly
in
the
provider.
I
just
found
myself
copying
pasting
it
a
lot,
so
I
pulled
it
out
and
now
essentially
what
we
have.
We
have
this
toy
actor
and
all
it
does
is
listen
for
HTTP
requests.
If
it's
query
it
does
a
select
star
statement,
that's
hard
coded
into
the
actor
and
if
it
is
at
slash
execute
it
looks
for
the
the
name
query
tag
and
whoever
the
name
is
that
it
puts
it
in
the
database.
C
So
if
we
sit
here
and
then
you
sit
here
and
if
we-
those
are
my
three
dogs-
but
if
we
went
over
here
and
said,
Brook
Brooks
is
my
new
friend.
It
doesn't
say:
okay,
but
come
back
over
here
now
now
we
have
Brooks
as
a
friend
and-
and
you
know
we
can
keep
doing
this.
Liam
he's
been
nice
to
me.
He
can
be
my
friend
and
we
and
we
have
Liam
and
the
the
cool
piece
about
this
is
now
we
get
to
how
do
I
change
screens.
C
So
we
get
to
you,
know
I'm
in
here,
there's
nothing
in
nothing
in
this
directory
right
and
we
get
to
use
the
power
of
light
stream
to
restore
out
of
our
S3
bucket.
And
now,
if
we
do
trains,
you
select
star
from
friends,
we
have
our
database
with
Liam
and
Brooks
in
it,
and
all
of
that
replication
was
done
with
the
provider
and
I.
C
Just
think
this
is
really
cool,
because
now
I
can
just
give
you
know
Brooks
or
whomever
a
key
to
access
this
S3
and
they
have
the
database
that
they
can
use
locally
or
I'm
still
learning
like
stream.
I
know
it's
supposed
to
be
able
to
allow
us
to
do
some
really
cool
things
with
SQL
to
make
it
a
production
work.
You
know
using
locks
or
whatever
it
makes
it
shareable
I,
don't
know.
I
still
have
a
lot
of
learning
to
do,
but
but
yeah
that's
it
and
it
can
be
yeah.
C
It
can
be
found
here.
It
is
not
feature
complete.
It
only
works
on
text
fields
at
this
point,
but
I'm
going
to
keep
playing
with
it
and
make
it
more
feature
complete
and
that's
really.
It.
A
It's
it's
so
cool
to
see
additional
implementations
of
these
contracts
right
also
because
it's
implemented
here
for
a
service
and
as
a
go
capability
provider.
That's
really
awesome,
but
the
you
know
the
cool
part
is
like
Jordan
I,
don't
know
if
you
wrote
your
actor
before
or
after
you
started
this.
Did
you
start
with
this
with
like
the
postgres
provider
and
then
implement
lightstream?
Or
did
you
write
this
all
for
this
example?
No.
C
I
wrote
this
off
for
this
example.
That
being
said,
I
did
I
have
to
turn
on
the
post
Gres.
So
a
lot
of
my
encoding
troubles
right
was
understanding
how
seabor
was
doing
what
and
how
message
pack
is
doing.
What
so
I
did
turn
on
the
pet
clinic
and
subscribed
to
the
channel
for
the
postgres
actor
and
that's
how
I
was
actually
using
the
postgres
provider
to
study
the
check
my
sqlite
messages,
because
over
the
wire
they
also
see
the
same
so
I
did
use
it
as
I
say,
object.
A
A
Oh
well,
it's
going
to
come
to
me
in
a
second
yeah,
so
the
whole
the
encoding
mechanism,
with
message
pack
and
seaboor.
You
know
looking
at
what
we
should
use,
we
consider
just
using
Json.
We
consider
using
something
like
you
know:
Captain
Captain,
Proto,
or
something
like
that.
A
You
know
message
pack
was
a
great,
a
great
solution
for
something
that
was
really
efficient
at
encoding
over
time
as
we
were
using
it
and
Steve
is
actually
the
resident
expert
here
and
he
would
be
able
to
give
a
little
more
of
a
detailed
description.
But
C4
has
more
a
more
evolving
spec
message.
A
Pack
has
been
frozen
for
not
frozen
in
a
bad
way
but
Frozen
in
a
way
to
preserve
compatibility
for
four
years
now
and
C4
also
has
the
ability
to
do
it's
either
numbered
fields
or
or
custom
attributes
or
the
the
specific
term
is
escaping
me.
But
that's
that's
why
you
see
it
for
SQL
and
some
of
the
annotations
there
we
don't
actually
use
seabor
in
many
of
our
provider.
Implementations
but
SQL
is,
is
one
of
them
so
kudos
for
going
down
some
some
custom
encoding
things.
D
I
just
checked:
we
don't
have
an
ADR
on
an
architectural
decision
record
on
why
we
why
we
ended
up
going
the
way
we
did
but
I
know
there
was
extensive
discussion.
It
may
actually
be
logged
over
in.
D
It
was
like
the
was
impossible.
What
was
the
old
name
of
the
repo
Brooks
wasmar
PC.
A
Oh
yeah
like
in
it
might
be
in
the
world's
investor
PC
or
whatever.
That
previously
was
like
the
actual
crate,
repo
yeah
I.
D
Think
that
was
where
we
maybe
wrote
down
our
reasons
and
the
documentation.
I
I
think
I.
Don't
remember
that,
because
I
think
I
wrote
the
documentation
for
it,
which
but
I'll
look
for
Jordan
if
I
can
find
it.
I'll
drop
it
in
the
channel
because
it
was
an
interesting
read
through
I
interviewed
Steve
and
Phil
keedy,
and
a
few
of
the
guys
that,
were
you
know,
involved
in
that
decision
to
write
that
take
those
notes
if
I
find
it
I'll
put
up
a
slack
for
you.
Thank.
C
A
And
Jordan,
can
you
talk
about
any
other,
any
other
gotchas
that
you
found
when
you
know
I
know
that
you
were
kind
of
home,
rolling
your
own
provider
SDK
for
go
because
that's
not
something
that
we
had
support
for
you
talk
about
like
you
know
what
you've
been
able
to
separate
into
that
provider
SDK
and
like
how
much
work
it
took
to
write
the
provider
minus
the
encoding,
yeah.
C
So
really
the
pieces
that
I
moved
out
are
kind
of
like
the
life
cycle
pieces
all
right.
We
know
that
we
are.
You
know
at
some
point
we
have
to.
We
start
and
the
first
thing
we
have
to
listen
for.
Are
you
know
some
sort
of
Link
definitions
right
so
and
a
lot
of
this
I
did
it
with
go
channels
right,
the
the
providers
sending
things
to
the
you
know
the
SDK
over
you
know
created
channels
and
the
SDK
is
doing
all
the
NAT
stuff,
but
it
was
pretty
straightforward.
C
It's
about.
You
know
if
the
link
is
good,
you
know
start
the
actor
and
listen
very
specifically
for
the
certain
messages
defined
by
the
contract
and
and
then
once
I
understood
that
honestly,
the
provider
itself,
minus
minus
business
logic
is
only
like
47
82..
It's
only
like
it's
a
white
space,
it's
like
30
lines
of
code-
and
most
of
it
is
is,
is
goes
giant.
Error
block
error
checking
blocks,
so
it
actually
got
pretty.
C
You
know
became
pretty
straightforward
and
you
know,
without
the
only
thing
this
provider
doesn't
do
yet
is,
is
handle
multiple
connections
from
multiple
actors,
which
is
what
I'm
going
to
do
next,
trying
to
figure
that
piece
out
but
yeah.
No,
it
was
really
understanding
the
life
cycle
figure
out
the
encoding
and
everything
else
is
pretty
straightforward.
A
That's
that's
awesome,
I
mean
I.
I,
hope
that
you
know
you
well,
it's
open
source,
anyways,
so
I
hope
you'd
be
fine.
Putting
that
provider
SDK
out
there
and
and
letting
us
experiment
with
it
and
the
community
experiment
with
it.
And
would
you
be?
Would
you
be
interested
in
taking
this
light
stream
provider
once
you've
had
a
chance
to
work
on
it
and
polish
it
up
to
whatever
standard
you'd
like
to
and
getting
it
PR?
As
like
a
community
provider
into
our
capability
providers,
repo
I
mean.
C
This
would
be
a
sweet
alternative.
Where
do
you
see?
You
know
the
sqlite
capabilities,
the
funny
thing
about
live
stream?
Is
it
wasn't
released
as
a
go
like
as
a
module
right,
so
I'm
not
importing
light
stream?
What
I
had
to
do
is
pretty
much
go,
take
chunks
of
their
code
and
move
it
into
the
repo
with
it
and
I
call
it
directly,
but
if
we
can
get
them
to
just
expose
like
I,
think
I
use
two
functions
then
yeah
this.
C
This
can
be
like
a
full
light
stream
thing
that
we
can
do
that
can
have
a
life
cycle.
We
I
would
just
want
to
talk
to
the
light
stream
people
and
understand
why
they
didn't.
Why
they're
not
like
exposing
the
the
start.
Light
stream
function
essentially
publicly.
A
Yeah,
okay,
yeah.
That
makes
a
makes
a
lot
of
sense,
and
even
if
it's
even
if
it's
something
that
you
have
to
run
the
light
stream
database
or
like
run
a
mini
o
thing
anything
you
can
do
locally,
I
mean
that's
fine,
does
lightstream,
actually
is
it
its
own
service
or
can
you
do
like
an
in-memory
sqlite
kind
of
thing.
C
It's
it's
a
it's
a
command
line
tool
as
it
sits
right
now.
I
would
the
only
reason
it's
working
with
the
providers
because
more
or
less
they
have
the
the
replicate
function
is
a
long-running
service
that
you
can.
You
know.
Typically,
the
documentation
is
like
do
system
b
or
something
like
that.
I
just
took
that
function
and
stuck
it
in
the
provider
and
that's
what
we're
calling
so
the
provider
is,
is
the
is
the
parent
edit.
C
But
after
that,
so
I
am
interested
in
experimenting.
If
I
can
do
an
in-memory
database
like
go
to
S3
pull,
you
know,
replicate
it
down
into
memory
and
that
type
of
stuff,
because
I
know
sqlite
can
do
that.
But
I
don't
know
that
I
can
point
light
stream
at
an
in-memory
database.
Yet,
but
I
haven't
gotten
that
far.
A
Okay,
yeah
I
I'm,
just
really
interested
in
that
I
know
when
I
think
of
some
of
our
examples
that
we
have,
there
are
still
some
that
are
just
easier
to
run
using
Docker
or
Docker
compose,
which
is
fine
right,
like
with
the
pet
clinic
example,
we're
running
everything,
including
the
UI
with
wasm
Cloud.
But
you
can't
like
there's
no
in-memory
postgres
database,
so
we
launched
the
postgres
databases.
Container
I
would
love
to
start
identifying
some
of
those
places
and
and
removing
Docker
as
a
as
a
necessity.
A
C
Can
definitely
do
that
with
sqlite,
because
it
is
just
reading
a
file
off?
Yes,
that's
the
whole
database,
which
would
be
cool,
so
I
mean
we
can
make
this.
Even
if
light
stream
is
in
the
future
I
believe
they
get
a
full-fledged
sqlite
provider.
A
Well
then,
so
somebody
could
you
know
if
we
had
a
SQL
live
provider
and
a
light
stream
provider,
you
know,
there's
any
you
know
number
of
things,
but
you
could
develop
locally,
just
like
with
a
file
on
disk
as
your
database
and
then,
whenever
you
move
to
production
or
like
your
real
real
running
environment,
you
swap
it
out
with
the
light
stream
provider-
and
you
have
you
know,
I,
don't
know
if
that
provides
more
resiliency
or
a
better
experience,
but
your
actor
code
doesn't
need
to
change.
A
Cool
well,
thank
you,
Jordan.
That
was
really
cool
and
I
am
looking
forward
to
now
that
you
have
some
go
tools,
I
feel
like
you're,
just
like
running
Buck
Wild
like
now,
you
have
the
language
of
your
choice
and
you
can
do
what
you
want.
C
A
Yeah
awesome,
let
us
know
you
always
have
always
got
to
demo
slot.
A
Okay,
I
figure.
Now
at
the
end
of
demo,
section
it's
good
to
ask
Matt:
do
you
have
anything
else
that
you
wanted
to
show
for
wash
build,
I
know
that
you
put
in
a
PR
and
that's
kind
of
actively
in
review
and
I.
Can
I
can
talk
about
a
couple
things
with
like
about
the
merging
strategy
there,
but
is
there
anything
else
that
you
wanted
to
show.
B
A
A
Look
at
that
I
can
share
my
screen:
okay,
yeah
Matt,
that
if
you
wanted
to
narrate
this
go
ahead,
sure.
B
So
basically,
the
the
purpose
of
this
issue
kind
of
RFC
is
that
at
least
my
personal
goal
and
I
think
a
goal
of
a
number
of
other
people
on
the
project
is
to
remove
make
files
from
being
such
a
important
part
of
the
wash
process.
At
the
moment,
this
carries
a
number
of
benefits.
B
Some
of
them
are
listed
on
here,
so
we
want
to
delete
them,
make
files
entirely,
which
carries
number
benefits,
splitting
portability
just
pure
Simplicity
for
developers
and
more
kind
of
polished
feel
to
using
watch
and
using
the
was
and
Cloud
ecosystem
in
general
and
it
and
simultaneously,
with
some
of
these,
these
commands.
B
We
want
to
try
and
make
the
simple
case
of
developing
awesome,
Cloud
applications
to
be
easier
right
now,
there's,
as
many
of
you
guys
use
watch,
probably
know,
there's
a
lot
of
commands
that
require
multiple
different,
like
words
so
you'd
have
like
wash
reg
and
like
the
multiple
different
sub
commands
so
trying
to
reduce
the
number
of
times
you
need
to
use
those
just
to
make
it
easier
for
people
to
learn
these
commands
because
it's
easy
to
use
them.
B
So
that's
what
a
number
of
these
commands
are.
They
are
actually
coming
from
makefell
already
have
and
just
kind
of
translating
these
into
wash
commands
so
they're
kind
of
directly
in
there.
This
makes
it
easier
for
users
as
well
as
it
also
makes
it
easier
for
us
as
project
maintainers,
especially
when
we
add
more
languages,
because
having
a
make
file
that
is
split
between
multiple
languages
is
quick
is
quickly
going
to
get
kind
of
messy,
so
this
will
make
it
easier
to
implement
for
each
language
that
we
want
to
support.
B
If
you
take
a
look
at
these
some
some
of
these
commands,
I,
probably
won't
walk
through
all
of
them.
If
you
want
to
take
a
look
through
them
in
your
time,
go
ahead.
Some
of
the
important
is
a
kind
of
wash
test
so
automatically
running
the
tests
for
each
language
and
yeah.
There's
just
a
lot
of
simpler
commands
here
to
get
some
of
the
often
used
stuff
for
especially
for
active
Developers
yeah.
That's
that's
mostly
it
for
this.
B
If
people
want
to
take
a
look-
and
let
me
know
what
you
think,
if
you
have
any
comments,
I
know,
Brooks
has
a
bunch
of
comments
from
earlier
that
he
that
he
made
and
I
responded
to
those
so
we're
kind
of
in
a
dialogue
right
now,
so
people
want
to
take
a
look,
go
ahead
and
then
proxy
anything
else.
You
want
to
add.
A
Yeah
I,
just
overall
big
support
for
the
proposal.
In
addition
to
make
files
requiring
make
on
Mac,
we
actually
have
this
on
our
documentation.
A
A
Some
of
our
scripts
also
include
things
like
JQ
for
parsing
Json
output,
and
you
know,
if
we
kind
of
bring
this
all
into
wash,
then
we
get
to
parse
the
output,
so
it
removes
a
lot
of
I
think
we
actually
have
that
I
was
going
to
talk
about
something
else
for
our
install,
but
you
know
you
see
we
have
this
section
on
Unix
utilities
like
bash
git
JQ
make
removing
that
whole
section.
You
see,
there's
the
new
make
gnu
whatever
removing
this
whole
section.
A
You
know
takes
off
like
a
quarter
of
the
page,
it'd
be
real
nice
to
kind
of
build
it
all
into
into
wash
so
big
support.
The
only
comments
that
I
really
had
it
was
just
brainstorming
on
how
we
can
best
organize
the
sub
commands
here,
because
we
have
a
good
opportunity,
like
love,
wash
start
like
the
need
to
be
able
to
say
control
start
and
then
the
actor
or
provider
I'd
love
to
just
be
able
to
say
wash
wash
start
and
then
give
it
the
thing
to
to
start
so.
A
Big
fan
of
The
Proposal,
just
really
only
is
about
like
the
details
of
how
we
want
to
structure
the
sub
commands
and
washbuild
is
certainly
the
first
and
maybe
most
important
step
of
this
foreign.
B
Yeah
for
sure,
I'd
also
mentioned
that
a
lot
of
these
commands
will
be
will
be
leveraging
the
watsoncloud.tamil
file
that
we've
been
working
on.
B
So
all
those
kind
of
parameters
that
are
in
Brackets
there'll
be
stuff
that
is
either
gotten
dynamically
from
either
requesting
a
host
or
is
just
pulled
straight
out
of
a
config
file.
So
we
kind
of
infer
those
parameters
for
you
just
to
make
it
easier
to
use,
wash
and
learn
the
wasn't
Cloud
ecosystem
yeah.
If
people
have
other
commands
or
command
they
don't
like
these
commands.
You
know
feel
free
to
comment
and
let
us
know.
A
One
thing
that
we've
kind
of
talked
about:
that's,
maybe
hot
contention,
I,
actually
don't
really
have
a
preference
I,
don't
know
what
the
majority
of
things
are
is.
If
you
should
do
you
know
your
CLI
and
then
the
verb
and
then
the
noun
or
your
your
binary
and
then
the
noun
and
the
verb
like
if
you
inspect
claims
or
claims,
inspect
and
I
think
what
wash
has
now
is
and
I
think
Conor.
Maybe
you've
pointed
out
a
couple
inconsistencies,
but
largely
the
noun
and
then
the
and
then
the
verb
right.
A
So
we
have
like
wash
control
I
guess
we
have
start
after
it
goes
kind
of
back
and
forth
there,
but
it'd
be
interesting.
You
know
if
we
make
some
of
these
changes
if
we
do
want
to
reconsider
the
like
changing
the
order
of
of
our
commands
so
we'll
have
to.
Maybe
it
would
be
useful
if
we
lay
this
out
all
in
like
a
on
like
a
table
or
something
and
try
to
see
how
we
can
structure
the
the
commands.
A
I'd
probably
be
probably
be
useful,
and
you
know
we're
gonna
publish
a
breaking
change
for
all
of
this,
and,
but
you
know
whatever
is
most
ergonomic
and
what
most
tools
use
and
we
can
best
like
categorize
commands,
will
be
probably
be
the
better
thing
to
go
with
so
I,
like
the
the
more
I'm
talking
I
like
the
table,
idea
to
kind
of
lay
out
all
the
commands
together.
B
Yeah
I
agree
I
think.
Maybe
we
can
make
another
markdown
file
just
with
kind
of
the
existing
commands.
Then
we
can
keep
that
updated,
but
yeah
I
think
personally.
I
believe
that,
like
wash
verb
is
the
best
way.
If
we
then
need
to
do
wash
verb
noun
for
some
things
that
works
too,
but
I
just
I
I
think
that
washed
noun
verb
kind
of
makes
people
think
like
Yoda
style,
I,
love
it
backwards
and
confuses
people.
B
So
I
don't
know.
Maybe
we
could
put
this
through
a
vote.
I,
don't
know
what
how
people
feel,
but
I
do
think
that
having
a
verb
first
makes
sense
to
me.
B
A
Yeah
I
am
I,
but
that
that's
a
good
point,
Justin
I,
don't
know.
If
you're
in
a
place
to
talk,
you
can
say
it
or
I
can
I
can
say
it.
I
like
that
last
thing
that
you
said,
which
is
just
like,
like
speaking
the
command
out
loud,
can
be
useful
too.
A
To
think
of,
like
what
people's
brains
will
be
will
be
thinking
like
saying
wash
start
actor
makes
more
sense
to
me
than
saying
wash
after
start,
I
felt
like
the
Yoda
comparison,
but
that
would
be
maybe
we
can
take
a
step
back
and
think
of
the
things
that
we
want
wash
to
be
able
to
do,
and
then
we
can
organize
that
into
the
commands
that
we
would
want
watch
to
support.
A
We
could
do
everything
with
a
top
level
command.
The
only
can
I
did
drop.
This
concern
I
think
that
I
would
just
not
want
to
run
wash
like
dash
dash
help
and
then
get
like
a
whole
man,
page
of
commands
so
but
I'm
sure
we'll
be
able
to
organize
organize
that
it's.
A
Cool
well
I
dropped.
This
drop
this
issue
in
the
chat.
This
is
in
the
wash
repo
now
issue
299,
so
anybody
feel
free
to
drop
comments,
I'm
sure
after
today's
discussion,
probably
there's
going
to
be
a
good
bit
of
iterating
on
this
one
so
sounds
like
it'll
be
be
fun.
A
I
had
a
couple
things
that
I
wanted
to
share
just
general
wasmcloud
things,
and
but
that
should
be
should
be
pretty
quick.
This
is
actually
a
deploy.
All
you
can
see.
This
is
actually
a
preview
of
of
documentation,
I
I
put
in
a
draft
PR,
including
the
release,
change
log
for
the
new
0.57.
A
That
has
a
couple
like
little
bug
fixes.
This
is
really
a
high
level
view
of
the
fixes
and
in
the
git
repo
we
include
the
the
smaller
things
which,
thanks
to
robust,
and
maybe
it
was
the
last
one
yeah
taction.
These
folks
have
been
submitting
a
couple
PR's
for
the
Wasson
Cloud
dashboard,
which
has
been
awesome
like
just
little
little
fixes
for
errors
that
pop
up
in
the
console
someone
came
in
today
and
put
in
a
PR
to
let
you
close
modals
with
the
Escape
key,
which
is
super
nice.
A
So
so,
thank
you
all
for
that.
But
what
I
I
really
wanted
to
point
out
here
on?
The
call
is
that,
with
release
0-57,
we
changed
the
way
that
we
create
wasm
Cloud
releases
and
the
way
that
this
was
done
previously
is
with
a
library
called
Distillery
and
a
couple
of
years
ago.
This
was
the
de
facto
way
to
create
application
releases
with
Elixir
apps,
and
this
is
great.
A
It
had
a
ton
of
features
that
the
built-in
releases
didn't
have
at
the
time,
including
things
like
executable
thing,
executable
releases
and
being
able
to
start
applications
and
things
like
that.
But
it's
actually,
we
noticed
the
other
day
that
it
hasn't
really
been
updated
since
2020
February
17th
of
2020,
so
over
over
two
years
ago,
which
isn't
necessarily
a
problem,
but
we
did
see
when
building
our
project.
A
Let
me
see
if
I
can
find
I
guess
this
was
me
iterating,
but
the
build
started
to
fail
when,
when
creating
releases-
and
we
were
wondering
why
and
turns
out,
Distillery
I
can
actually
find.
The
issue
is
not
compatible
with
OTP
25,
which
is
the
latest
release
of
OTP,
and
you
know
I
I,
don't
have
we
didn't
have
a
huge
level
of
confidence.
The
Distillery
would
be
would
continue
to
update
with
new
new
versions
of
the
standard,
Library
and
actually
mix
releases,
which
is
the
way
mixes.
A
The
kind
of
cargo
equivalent
for
elixir
mixed
releases
are
a
way
they've
grown
a
lot
in
the
last
couple
of
years.
The
kind
of
eclipse
Distillery
so
I
knew
this
had
to
happen
at
some
point,
but
with
builds
and
the
CI
failing.
A
Ci
thing
you'll
never
know
that
this
change
was
actually
made,
but
I
did
want
to
point
a
couple
of
differences
out
which
are
documented
on
the
website
and
in
GitHub,
but
I'll
show
you
kind
of
live,
so
one
thing
to
point
out
is
that
switching
from
Distillery
releases
to
mix
releases
did
drop
our
binary
or
our
release
size,
a
significant
amount
which
is
pretty
sweet.
We'll
always
take
that.
A
So
what
you
all
are
are
likely
used
to
is
when
you
run
bin
was
on
cloud
host.
You
get
this
this
blob
of
help
text,
so
you
can
do
things
like
starting
in
the
background,
as
a
demon
start
in
the
foreground,
with
law
dogs
or
with
an
IEX
console
things
like
that.
Now
the
prompt
you're
going
to
get
is
a
little
different.
So
the
the
actual
structure
of
the
project
is
the
same:
the
bin,
erlang
runtime
system
and
everything,
but
the
commands
have
changed.
A
A
If
you're
looking
for
the
old
functionality
of
starting
the
host
in
the
background
kind
of
as
a
demon,
it's
aptly
named
demon
now
and
doing
that,
will
output
your
logs
to
Temp
log
or
Lang
log
one.
So
it's
it's
kind
of
the
same
as
before.
Just
temp
log
instead
of
VAR
log-
and
you
can
see
your
your
logs
there.
So
I
wanted
to
point
this
out.
A
If
you're
using
a
wasm
cloud
release
newer
than
newer
or
equal
to
then
whatever
version
57
you'll
you'll
have
to
use
start
to
start
in
the
foreground
and
Damon
to
start
it
in.
A
In
the
background,
this
probably
won't
affect
much
if
you're
using
the
docker
image,
we
take
care
of
the
start
command
for
you
anyways,
but
one
notable
decision
that
we
made
out
of
this
is
that
wash
lib
as
a
version.0.2
sure.2.0,
whatever
will
enforce
a
minimum
version
of
version
57.-
and
this
is
mostly
because
wash
lib
is
very
new
it.
It
implements
functionality
for
starting
laws
on
cloud
hosts
and
since
there's
no
plan
to
ever
go
back
on
Distillery
releases,
at
least
for
now.
A
There's
really
not
too
much
sense
in
maintaining
two
different
code.
Paths
for
starting
two
different
types
of
releases
additionally
wash
lib
or
or
the
wash
CLI
will
always
have
a
pinned,
wasn't
Cloud
version.
So
when
you
download
wash
it's
going
to
start
that
version
of
awesome
Cloud,
you
can
override
it,
but
the
default
use
which
is
going
to
be
start.
A
version
of
the
host.
That's
pre-compiled
in
I
would
argue
is
like
the
95
case.
Won't
ever
know
that
this
is
happening.
A
So
I
put
this
in
the
wasm
cloud
slack
just
to
see
if
anybody
was
violently
opposed.
Does
anybody
think
that
this
is
a
anybody
else
think
this
is
a
like
a
bad
idea
and
that
we
should
support
free
0.56?
A
Cool
then
yeah
with
with
slack
in
in
this
call
having
silences
acceptance.
I
think
that
this
is
fine.
I
think
that
there's
no
need
to
support
really
old
versions,
and
you
know
every
new
version
of
the
host
that
comes
out
is
such
a
massive
Improvement
at
this
point
it
it
might
as
well
rip
the
Band-Aid
off
now
and
we
can
just
continue
to
get
better
going
forward.
A
I
guess,
that's
all
that'll
be
extensively
documented
in
the
wash
lib
release
and
I
I.
Don't
think
I
have
the
right
version
of
wash
that
I'm
working
on
this
installed
locally
or
anything.
But
if
you
tried
to
start
a
version
of
the
host,
it's
like
0.55,
it
gives
you
a
nice.
A
A
Cool,
let's
see,
that's
all
I
have
for
now.
I.
Don't
think
that
the
schedule
for
cloud
native
wasn't
day
is
public.
Yet
as
soon
as
it
is,
we
will
be.
Is
it
public?
Oh
man,.
D
Yeah
look
over
in
chat;
it
just
went
live
in
the
last
couple.
Minutes
I
still
have
got
to
get
talk
in
two
talks
updated,
but
one
is
a
panel
that
we're
gonna
have
with
a
venture
capitalist,
a
media,
a
representative
Alex
Williams,
the
founder
of
the
new
stack
and
then
Renee
Shaw
from
amplify
and
then
the
other
one.
Is
the
Sig
registry
talk
that
Taylor
Thomas
from
cosmonic
is
going
to
be
giving
he
I
think
he
needs
to
update
a
few.
A
Okay
cool
well,
this
looks
like
a
this
looks:
pretty
sweet,
Luke
Wagner
from
fastly
Kate
golden
ring
from
Microsoft.
We
have
the
folks
from
farmion
who
gave
the
the
CMS
talk
at
the
last
wasn't
day
doing
looks
like
they're
doing
the
keynote
and
then
another
one
from
the
docker
folks.
That's
pretty
cool.
D
A
Docker
has
two
talks:
they're
launching
some
native
web
assembly
support,
but
I
think
this
talk
at
9.
50
is
probably
one
of
the
more
exciting
ones
of
the
day.
That's
jvm
support
for
Wazi,
which
is
kind
of
the
one
of
the
white
whales
from
a
language
perspective,
so
I've
been
following
that
work
in
TVM
for
a
little
while
and
that's
all
out
in
the
public.
D
The
pull
requests
about
that
one
is
super
excited,
that's
super
exciting
and
then
we've
got
this
python
talk,
so
we
not
only
have
Java.
We
also
have
python
with
an
update,
they're
moving
I,
don't
know
if
they've
announced
it
yet,
but
I
think
they're
planning
on
moving
to
making
Wazi
a
tier
two
language,
which
means
it's
more
officially
supported
in
Python
and
then
another
Docker
talk
and
then
we've
got.
D
We
have
this
guy
Brooks
Townsend
speaking
I,
don't
know
if
you've
ever
heard
of
that
guy
who's,
that
guy
another.net
talk
so
three
different
languages
and
we
got
Carl
back
from
single
store
and
the
Taylor
and
a
few
other
folks
who
are
doing
Sig
registries
and
then
c-sharp
and
wasm
interface
types
another.
Yet
another
language
talk
so
we've
got,
you
know
three
big
languages
coming
down,
so
there'd
be
a
lot
there.
A
Apache
wasm
components
from
oh
yeah
Taylor
has
two
talks
that
he's
doing
on
on
wazen
day:
very
cool
yeah,
oh
nice
and
then
Conor
from
suborbital
will
be
there.
I
know
he
was
a
this
of
orbital
folks
were
a
notable
missing
party
at
the
the
Europe
one
very
cool
yeah.
A
Well,
cool
this,
you
know
I'll
send
this
in
chat,
but
it's
pretty
easy
to
find
if,
if
y'all
are
looking
for,
if
y'all
are
looking
for
the
the
schedules,
like
Cloud
native
awesome
day,
n
a
for
for
2022,
that's
a
great
Justin
makes
a
good
point
in
the
chat
that
all
of
the
all
the
language
talks
represent
a
good
opportunity
to
to
network
and
to
get
developers
interested.
A
Just
very
true
whether
or
not
the
whole
Java
thing
is
is
using,
you
know
is
using
part.
You
know
TVM
or
you
know,
python
is
becoming
like
a
a
B2
or
whatever
any
advancement
in
in
languages.
Compiling
to
wasm
brings
on
a
whole
nother
group
of
people.
It's
really
exciting.
This
is
going
to
be.
This
is
gonna,
be
a
fun
one.
A
Sweet
I
was
yeah.
I
was
pretty
sad
that
that
wasn't
public
yet
and
then
here
we
are
very
cool.
Well,
I
didn't
have
anything
else
that
I
wanted
to
talk
about
today,
necessarily
I.
Think
it's
a
good
time
for
open
floor.
Does
anyone
else
have
anything
that
they
wanted
to
talk
about
from
the
Watson
Cloud
Community
stuff
happening
in
Watson
cloud
or
or
just
the
general
webassembly
community.
A
Thank
you
all
right.
Well,
pretty
sweet
community
meeting
thanks
everybody
for
coming
I'll
I'll,
stop
the
recording
and
we
can
just
hang
out.