►
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
B
We
pride
ourselves
on
people
being
able
to
onboard
quickly,
and
so
one
of
the
things
that
I've
gotten
to
learn
is
is
a
an
application
called
gitpod,
that's
available.
So
it's
getpod.io.
You
can
go
to
the
website
and
it's
a
very
similar
experience
to
what's
being
developed
in
the
space
in
terms
of
github
co-spaces
or
stack
blitz
or
any
of
the
others
that
are
out
there.
B
But
it's
a
little
more
heavyweight
you're,
actually
getting
a
container
with
some
with
some
decent
cores
on
it
and
some
decent
ram,
and
you
can
actually
do
quite
a
bit
of
development
and
vs
code
is
not
the
only
appliance
you
can
connect
to
it.
You
can
shell
directly
into
the
machines
you
can
join
in
through
intel,
intellij
and
jetbrains
products.
So
it's
it's.
It's
kind
of
broadly
gone
through
it.
So
I
come
from
a
background
of
we.
B
Do
we
do
a
little
bit
of
kubernetes
microservices
through
dapper,
and
so
looking
at
wassum
cloud
it
came
across
my
feet
is
and
it's
something
I'm
interested
in
as
a
rust
programmer,
because
I
like
rust
and
yeah,
it
seemed
like
a
nice
fit.
So
I
said
what
is
what
is
something
that's
smaller
than
dapper
and
certainly
a
little
bit
closer
to
kubernetes
agnostic,
just
pure
wasm
and
rust.
Look
like
for
a
back
end
application.
B
So
I
scoured,
through
the
docks,
learned
what
I
could
and
decided
boy.
This
would
be
really
good
to
have
in
a
template
if
I
ever
wanted
to
spin
up
a
back
end
with
getpod.
Really
quick,
I
think
that'd
be
some
really
good
synergy.
That
would
allow
people
to
see
the
application
and
work
on
it
and
experiment
right
and
it's.
C
B
I'm
gonna
take
this
link.
I'm
gonna
pop
this,
I
think
yeah
we've
got
it
we'll
make
sure
it's
there.
So
currently,
this
is
a
github
repo.
It's
a
public
template.
Anybody
is
free
to
access
and
go
clone
and
it's
just
got
some
kind
of
standard
files,
so
just
kind
of
real
quick,
I'm
just
going
to
walk
you
through
the
repo
from
the
github
web
editor
view
of
this
right,
and
then
we
can
walk
through
a
couple
files.
So
just
a
couple
things
that
getpod
kind
of
needs.
B
It
needs
some
rough
ideas
of
how
to
bring
up
your
dev
environment
right.
So
for
awesome
cloud
in
particular,
the
docs
indicated
a
great
way
to
do.
That
is
a
docker
compose
file.
So
what
I
did
was
I
made
a
little
awesome
cloud
base
folder
here
and
put
a
docker
compose
in
there
and
we
started
up
all
the
relevant
services
to
go
along
right.
So
that's
a
standard,
redis
image.
If
we
want
to
test
some
key
value,
work,
a
nat
server
for
the
controlled
message
plane
behind
it,
a
local
docker
registry.
B
If
we
have
any
signed
artifacts
that
we
want
to
go
ahead
and
push
locally
and
and
spin
up
scale
of
actors
right
beyond
one
actor
and
in
addition,
finally,
the
awesome
cloud
host
itself,
the
awesome
cloud
application
so
yeah.
So
that's
a
that's
an
important
step
in
the
thread
I
kind
of
walk
through
what
the
rest
of
these
are
some
things
that
it
does.
The
big
advantage,
really,
though,
is
the
pre-builds.
B
This
particular
init
step
will
actually
go
through
the
process
of
pulling
all
of
the
images
in
this
docker
compose.
That
can
be
a
pretty
expensive
task
locally.
It
can
also
be
a
pretty
time
consuming
task
remotely
too
so
that
init
task
allows
your
developer
environment
to
be
pre-built
on
get
pod
servers
before
it
ever.
You
ever
go
to
click
into
a
workspace,
and
whenever
you
want
to
load
a
new
workspace,
you
already
have
your
images
pulled,
you're
ready
to
go,
and
when
you
hit
run,
it
happens
as
quick
as
possible.
B
So,
starting
from
that,
I'm
going
to
go
back
here
and
we're
just
gonna
go
back
to
dot
com,
so
the
standard
experience
would
be
to
just
open
a
git
pod.
You
would
click
the
link,
I'm
on
a
dev
branch
right
now,
and
so
I'm
gonna
have
to
move
that
and
do
and
we're
just
gonna
make
this
a
little
bit
different,
it'll,
probably
yell
at
me,
and
it
says
I've
already
got
one
but
we're
gonna
pull
a
new
workspace
anyways.
B
So
one
of
the
great
parts
about
get
pod
is
you
can
just
pull
as
many
workspaces
right
up
to
fork
and
currently
is
up
in
the
professional
open
source
program.
If
you're
an
open
source
project
and
you're
an
open
source
organization,
they've
got
a
program
you
can
apply
to
as
well.
You
can
join
they'll,
basically
give
you
a
kind
of
an
elevated
license
beyond.
I
think
the
hundred
hours
of
runtime
anybody
off
randomly
off
the
web
would
get
so
yeah.
So
this
is
pulling
the
container
image.
That
is
the
base
for
the
developer
workspace.
B
B
D
B
B
You'll
have
a
rust
c
same
thing:
versioned,
it's
com,
it's
configured
with
rust
up.
So
if
you
rest
up
version,
you're
running
one,
two,
four
three:
these
are
all
configured
at
the
root
of
the
repository.
So
that's
a
rusttoolchain.toml
file
kind
of
at
the
at
the
root
there,
and
that
allows
you
to
just
configure
what
rust
do
I
want.
So
we
can
compile
rust.
B
That
does
not
want
to
download
very
fast
for
you
or
maybe
you
don't
want
to
pay
the
the
big
bucks
to
the
cox
or
the
verizon
of
the
world
to
give
you
all
the
bandwidth
so
yeah.
This
is
running
it.
You
know
cloud
data
center
speeds,
so
you're
not
going
to
get
better
network
latency
yeah,
we'll
let
this
build,
and
hopefully,
after
I'm
done
with
this,
I
will
get
the
target
path
of
our
new
actor.
B
That's
signed
and
built
with
this
make
file
once
that's
done,
I'll
plug
it
into
the
dashboard
here
and
say,
start
actor
and
that'll.
Let
us
configure
an
actor
with
a
specific
capability,
so
here's
our
actor
made.
It's
been
signed
with
keys
that
are
made
by
default
because
we
did
not
have
any
at
the
root
of
our
repository.
So
we're
welcome
to
add
those
in
a
repo.
B
B
B
B
If
we
go
back
to
the
shell-
and
we
kind
of
want
to
see,
what's
going
on,
we'll
see
that
indeed
the
host
that
this
dashboard
runs
picked
up
or
asked
to
run
that
capability
provider
and
told
us
its
name
its
instance,
the
various
nat
configurations
that
nas
configurations
that
it
needs
and
then
we
go
through
here,
pick
our
actor
pick.
Our
provider
pick
the
link,
awesome
cloud,
http
server
address
equal
8087.
B
and
so
an
interesting
thing
about
this.
This
is
a
remote
environment.
This
is
someone
else's
container
somewhere,
so
it
can
be
a
little
weird
to
think
about
what
ports
are
accessible.
One
of
the
ways
to
do
that
in
getpod.
Is
this
remote
explorer
window
so
believe
it
or
not?
Beyond
opening
this
just
to
this
application,
there's
options
here
to
actually
go
ahead
and
make
this
public.
B
B
B
B
This
is
this
is
remote
dev
environments-
and
this
is
this-
is
the
benefit
of
using
them
right
is
everybody
deserves
the
same
tools
or
the
tools
that
they
want
if
they
want
to
come
into
this
with
c
lion
and
intellij
they're
welcome
to
connect
to
it
over
over
a
remote
connection,
which
is
available
if
they're
die
hard,
vim
and
terminal
and
unix's
admin
guy,
they
can
sshn,
there's
a
getpod
local
companion
as
well
so
yeah.
This
is
how
you
template
a
project
and
yeah.
B
A
Hi
evan,
you
absolutely
crushed
this
dev
up.
How
cool
I
mean
you
didn't
do
a
good
job.
You
did
an
amazing
job
and
I
just
want
to
say
thank
you
so
much
for
not
only
the
twitter
thread
but
taking
the
time
to
come
in
and
walk
everybody
through
all
the
details
that
you
really
did
there.
Like
you
know
the
hot
reloading
and
taking
all
that
friction
out
of
the
development
process
is
incredible.
A
If
it's,
if
it's
okay,
what
I
would
love
for
you
to
do
is,
would
you
do
a
pull
request
over
to
wasn't
cloud
so
that
we
can
put
this
right
on
our
project
repo?
I
just
I
would
cut
and
paste
to
put
it
yours,
I
want
you
to
have
you
know
and
I'd
love
to
get
you
in
that
would
be
phenomenally.
It
would
be
phenomenally
awesome.
I'd
like
to
kevin
work
with
you
to
figure
out
what
would
the
right
like
start,
developing
and
wasm
cloud
now
button.
B
Yeah,
I'm
more
than
happy
to
transfer
ownership
of
the
template
I'll
I'll,
just
ship,
the
repo
your
way
and
kind
of,
let
you
run
with
it
from
now
on.
I
I've
I've
also
I'm
in
touch
with
the
get
pod
community
stuff.
So
in
my
volunteer
time,
I'm
part
of
the
get
pod
community
hero
program.
That's
a
program
that
I
joined
about
a
week
ago,
and
it
gives
me
some
elevated
permissions
and
a
little
bit
more
ability
to
talk
within
the
get
pod
organization
about
causes
that
are
worthwhile
and
and
worth
looking
into.
B
And
so
one
of
the
things
that
I
put
in
as
well
is
to
they
don't
have
an
elixir
language,
docker
image
or
anything
like
that.
So
probably
my
next
step
is
going
to
be
to
look
at
the
awesome
cloud
host,
github
repo
and
actually
ask
you
know
what
does
it
look
like
for,
for
you
guys
to
stand
up
remote
contributions
to
awesome
cloud
host
with
an
elixir
and
rust
stack
as
well,
so.
B
Again,
there
are
open
source
programs
available
yeah.
How
long
do
the
hosts
live?
There's
actually
some
great
questions.
I
would
love
to
answer
those.
There
are
run
time
defaults
depending
on
your
user
and
license
level
for
how
long
a
workspace
persists.
It's
a
really
great
question,
obviously
there's
things
that
persist
for
the
life
cycle
of
a
git
repository
and
then
there's
things
that
would
persist
like
I
have
local
changes
here
of
eight
files.
How
long
would
they
last
right
so
for
my
example
right?
B
B
I
think
that's
on
their
standard
tier
that
I
think
is
like
9.99
a
month
and
then
there's
obviously,
enterprise
pricing,
tiers
and
things
I
think
getpod.io
probably
slash
pricing
would
probably
have
any
information
you
might
need
on
capabilities,
and
what
comes
with
that
with
this
particular
platform,
but
there's
also
great
stuff
in
github
code
spaces
too,
and
with
yeah
with,
what's
in
even
the
beta
react
docs
that
have
been
done
right,
there's
a
tremendous
amount
of
work
done
with
stack
blitz
there.
That
makes
that
experience
pretty
seamless,
angular
as
well.
B
D
B
A
Options
evan.
I
also
wanted
to
welcome
you
to
our
community
program.
I
think
we
just
started
it
about
10
minutes
ago
and
you
know
you're
the
first
member
free
t-shirts
and
lots
of
stickers
and
things
like
that
yeah.
So
you
know
we
will
we'll
get
that
we'll
get
you
an
order
form,
so
you
can
get
that
some
swag
shipped
out
to
you
and
all
those
kind
of
things,
but
very
that
was
incredible.
A
B
Yeah
yeah,
please
do.
I
have
only
gone
through
very,
very
I've
like
the
hello
actor
and
I
think
maybe
I've
done
the
to-do
actor,
but
I
have
not
really
tested
kb
redis.
I
know,
there's
a
postgres
sql
db
capability
provider
that'd
be
a
great
thing
to
test
as
well
yeah
and
there's
and
there's
really
clever
ways.
B
B
And
you
will
actually
be
able
to
grab
the
echo
actor
without
having
to
clone
echo
right.
So,
in
my
case,
I
put
echo
in
hello,
but
that's
okay,
right
so
yeah,
just
a
little
quick
tip,
if
you
guys
are
walking
through
examples
that
might
be
a
quick
and
convenient
thing
you
can
do
that.
That
would
make
your
life
a
little
bit
easier,
subversion's
still
good
for
some
things.
It's
it's!
Okay,
everybody
has
their
own
version.
Control
system
gets
not
perfect
and
it's
always
good
to
have
tools.
C
D
C
B
Yeah,
so
I
didn't
show
that
workflow,
because
I'm
myself
and
I
own
the
template,
so
that's
probably
a
really
good
thing
to
talk
about.
If
I
don't
own
this
template
the
great
way
to
do
it
is
that
gitpod
understands
you're
working
within
someone
else's
template
repo,
when
you
would
sign
in
through
maybe
a
github
provider
to
get
pod,
and
it
would
say,
oh
you
want
to
commit
changes
like
hold
on.
B
B
Would
you
like
to
fork
it
and
it
will
fork
it
for
you
and
then
you
have
a
new
repository
that
you
can
go
and
rename
and
it
offers
those
options
for
you.
There's
a
workflow
around
that
I'd
love
to
demo
it.
But
I'm
not
myself.
I
guess
so.
I'd
have
to
find
a
separate
user
to
log
in
and
test
that
workflow,
but
it
has
worked
before
I've
used
it
with
other
templates
and
it
allows
you
to
to
start
using
a
template.
One
like
one
example
in
particular
simon
at
redis
university.
B
His
last
name
is
escaping
me
recently-
has
come
into
our
community
and
is
building
flask
redis
boilerplate
for
him
to
use
as
part
of
his
workshops
in
education
within
redis
university
and
he's
encountering
the
same
thing
as
well.
He's
got
templates.
He
can
send
them
to
someone,
someone
can
make
their
own
repos
with
flask
read
a
spoiler
plate
or
node
read
a
spoiler
plate
and,
and
they
can
be
ready
when
they
show
up
to
a
webinar,
to
be
able
to
code,
live
and
and
debug
problems
with
solutions.
B
A
Evan
I
was
able
to
when
you
did
your
twitter
thread
on
monday
or
tuesday.
I
was
actually
able
to
get
it
up
and
running.
So
I
think
that
if
somebody
in
the
background
wants
to,
would
you
mind
just
it?
Oh
brooks
you
dropped
a
link
to
the
repo
there
and
from
the
awesome
awesome
cloud
list.
If
somebody
wants
to
give
it
a
try
in
the
background,
maybe
someone
else
can
demonstrate
the
the
fork
button
in
just
a
second
here.
A
If
they
try
to
make
a
change,
if
there
are
no
more
questions
for
evan
or
if
evan,
if
you're,
if
you're
finished,
I
just
want
to
say
again
yeah
just
amaze
balls.
That
was
an
awesome
demo.
I
don't
know
if
you
just
did
that
off
the
cuff,
if
you
rehearsed
or
whatever
that
was
totally
off
the
cuff,
that
was
great
man
knocked
it
out
of
the
park
and.
B
A
B
A
Time
as
a
member
of
our
community
program,
of
course,
you're
welcome
anytime,
we'll
get
you
one
of
the
keys
for
kevin's
library
room
for
his
little
back
room.
I
think
brooks
you're
up
next
with
a
community.
Let
me
ask:
are
there
any
more
demos
or
demonstrations
today,
I'm
not
sure
matt
how
your
stuff's
coming
along.
F
I
actually
what
I
ended
up
doing
last
week
is
I
created
my
own
interface
for
the
dynamodb
kv-ish
provider.
I
guess
I
would
call
it
so
maybe
if
we
have
time
at
the
end,
we
can
talk
about
the
discussion
that
has
gone
on
with
respect
to
versioning
interfaces,
and
you
know
extending
interfaces
and
all
that,
but
maybe
yeah
like
save
it
for
the
end,
if
we
have
extra
time.
C
Yeah,
let's
do
it
look
at
that
sharing
permissions?
Okay,
so
this
is
our.
This
is
our
our
biggest
community
call
out
yet
which
I'm
excited
to
bring
you
today?
We
have
four
community
call
outs.
Two
of
them
are
in
the
host
and
two
of
them
in
wash,
but
we'll
get
into
them
really
quick.
So
these
are
all
these
are
all
great
places
to
get
get
started
for
contributing
this
first
one.
C
We
noticed
the
the
slack
user
vyk,
I
believe,
found
a
little
bit
of
a
bug
in
our
dashboard,
where,
if
you
are,
if
you
are
putting
in
base64
encoded
config,
which
is
something
that
we
commonly
do
for
link
definitions
between
actors
and
providers
when
the
the
values
kind
of
exceed
three
or
four
link
definition
values,
you
can
do
a
base64
encoded
config.
So
it's
just
one
easy
value.
One
string
and
the
logic
for
parsing
link
definition
values
is
a
little
more
naive.
It
was
designed
before
we
started
supporting
this
config
base64.
C
So
what
we
like
to
do?
You
know
this
is
a
bug
for
now
what
we'd
love
for
somebody
to
you
know,
a
contributor,
or
one
of
us
to
do-
is
take
a
look
at
this
and
handle
trailing
equal
signs
so
that
we
can
support
base64,
encoded
config.
So
I
pointed
out
where
this
is
happening
in
the
dashboard,
so
this
should
be
a
it's
a
great
place
to
to
get
started.
If
you
wanted
to
take
a
look
at,
you
know
just
kind
of
changing
our
algorithm
a
little
bit
here.
C
The
next
one
is
also
in
the
otp
host.
So
each
link
definition
that
you
make
between
an
actor
and
a
provider
essentially
involves
five
different
things:
it's
an
actor
id
a
provider,
id
a
contract,
id
a
link
name,
which
is
something
that
you
can
use
to
run
multiple
different
instances
of
the
same
provider
and
then
all
of
the
values
like
what
we
were
talking
about
in
the
last
and
the
last
thing.
C
But
what
we
found
is
that
each
actor
can
only
really
have
one
link
definition
for
a
contract
id
and
a
link
name
when
you're
writing
your
webassembly
actors
or
wasn't
cloud
actors?
You're,
not
you
don't
know
what
provider
is
going
to
be
fulfilling
a
capability
at
runtime.
You
only
have
a
concept
of
a
contract
id
and
a
link
name.
C
That
was
it
was
already
there,
but
the
capability
provider
will
essentially
ignore
it,
because
it's
seen
as
like
a
duplicate.
So
what
we'd
really
like
is
when
we
put
a
link
definition
and
again
here
I've
linked
to
the
point
in
the
code.
We
should
reject
that.
So
when
you
try
to
do
wash
control,
link,
put
or
you're
in
the
dashboard,
and
you
define
a
link
definition,
we
should
reject
this
essentially
duplicate,
link
definition
and
say:
hey.
You
know,
you
should
probably
be
deleting
this
link
definition
and
redefining
it.
C
Instead,
this
will
just
overall
lead
to
less
confusing
use
cases.
It
can
certainly
still
continue
to
work
with
this,
but
it
can
cause
confusion
depending
on
you
know
how
you're
scripting
or
how
your
development
environment
is
set
up.
Things
like
that
now
on
to
wash
I
just
kind
of
made
up
this
prefix
here,
it's
kind
of
a
request,
but
with
with
wash
context
what
we
do
is
by
default.
C
Whenever
you
start
a
wasm
cloud
host
on
your
machine
wash
uses
what
we
call
a
context
to
take
the
configuration
we
used
to
launch
the
host
and
automatically
connect
to
that
host's,
lattice,
and
so
the
net
result
of
this
is
that
anytime,
you
launch
a
host
and
then
immediately
go
to
use
wash
to
try
and
get
the
list
of
posts
or
start
an
actor
or
something
you
automatically
use
that
host
config
context
to
connect
to
that
host.
So
by
default
super
easy
to
interface
with
the
host
locally.
C
You
don't
need
to
provide
a
bunch
of
different
parameters,
especially
if
you
have
things
like
authenticated
maps,
but
what
the?
What
the
this
issue
here
specifically
is
talking
about.
Is
you
can
deserialize
that
host
config
json
incorrectly,
depending
on?
If
a
value
in
there
is
incorrect
or
or
what
happened?
I
think
it
was
that
the
host
wrote
out
a
null
value
and
or
a
non-default
value
and
wash
didn't
deserialize
it
properly.
Something
like
that
in
general.
We
take
these
errors
and
fail
silently
because
a
wash
context
isn't
required
for
interacting
with
a
host.
C
C
All
right
and
the
last
one,
which
is
something
that
actually
kind
of
bit,
says
cosmonic
devs
the
other
day,
and
so
of
course,
we
immediately
wanted
to
improve
this.
This
experience,
when
you
do
things
like
put
a
link
definition
or
you
start
a
capability
provider,
the
link
name
flag
is
optional,
and
this
is
because,
when
you
start
a
provider,
you
put
a
link
definition
for
a
specific
provider.
C
C
These
are
the
things
that
you
actually
need
to
have,
but
what
I'd
like
to
see
here
is
include
the
link
name
option
as
something
that
the
default,
like
failure,
output,
shows
you
and
the
reason
for
this
is
so
that
you
don't
forget
about
it.
If
you
do
watch
control,
link
put,
you
know,
help
or
something
like
that.
C
It'll
show
the
link
name,
but
it's
really
useful
to
remember
that
this
is
an
option
that
we
accept
for
this
command,
and
this
feels
like
something
that
clap
probably
has
and
an
option
to
show
a
an
option.
Man
too
many
options.
It
has
an
option
to
show
a
flag,
even
if
it
is
optional
and
and
that
would
be
the
desired
experience
there
all
right.
So
that
was
a
lot
of
talking,
but
that
is
our
community
call
out
four
issues:
two
on
the
host
two
on
wash.
C
C
Yeah
yeah
great
call
yeah.
We
we
tag
all
of
these
with
good
first
issues,
so
they're
easy
to
find,
and
I
will
also
I
will
also
make
sure-
to
post
all
these
in
slack
so
that
you
don't
have
to
search
or
remember
or
wait
for
the
youtube,
video
or
anything
and
we'll
also
get
them
tweeted
out,
so
that
they're
they're
easy
to
find
as
well.
A
A
A
We
hope
to
finish
reviewing
submissions
for
wasm
day
for
kubecon
valencia,
the
week
of
may
15th,
and
we
hope
to
finish
that
on
thursday
and
start
releasing
acceptances
at
the
end
of
this
week
and
we're
also
working
on
a
set
of
training
for
that
particular
conference,
so
expect
some
training
options
as
well
this
year,
if
you
don't
have
your
tickets,
yet
all
of
the
online
virtual
tickets
do
include
a
wasm
day
ticket
this
year.
So
last
year
we
were
the
top
rated.
A
Not
diversive,
inclusive
community
diverse,
is
what
I
meant
to
say
kind
of
stumbled
there.
So
at
this
point,
maybe
if
we
want
to
turn
matt
to
any
road
map
discussions
or
any
work
and
flight
discussions
steve,
I
think
that
folks
are
really
interested
in
the
wasn't
cloud
ml
stuff.
I
don't
know
if
you
feel
like
prepared
to
just
maybe
speak
to
it
today
or
not
just
open
floor
for
anything
folks
want
to
bring
up.
G
A
Okay,
those
meetings
do
end
up
on
youtube.
So
folks,
I
can't
make
the
thursday
at
three
o'clock
slot
those
do
show
up
there,
so
you
can
get
caught
up
and
up
to
speed
in
channel
there's.
Also
a
link
in
general.
If
you
want
to
check
for
the
repo
is
linked
right
there
I
shared
from
the
wasmcloud
ml
channel
out.
I
will
paste
the
link
for
tomorrow's
meeting
in
the
channel
into
our
slack
channel
now
into
the
end
into
the
comments
here
as
well.
A
F
Yeah
I
mean
so
I
was
just
working
on
you
know,
for
the
kv
dynamodb
provider.
What
I
ended
up
doing,
because
I
needed
to
anyway
was
I
had
been
doing
that
initially
as
a
branch
on
the
official
capability
providers
repo,
but
I
created
my
own
capability
provider
since
I
think
the
thought
is
that
that
would
be
like
a
third-party
provider
and
just
for
the
sake
of
getting
to
a
demo
faster.
I
created
my
own
contract
for
that
provider.
F
An
issue
came
up
on
the
interfaces,
repo
that
probably
surfaces,
two
discussions,
which
I
don't
know
if
it's
worth
having
here.
One
is
on
versions
of
interfaces
which
I
think
is
like
a
probably
really
large
discussion
that
perhaps
warrants
its
own
even
meeting
amongst
those
that
care
and
then
the
other
one
was
just
kind
of
going
back
and
forth
about
the
idea
of
any
interface
that
may
require
some
sort
of
paging
and
how
to
or
to
not
support
that.
F
A
E
Sorry,
apparently,
I
didn't
click
on
mute,
hard
enough
yeah.
I
think,
like
you
said,
the
the
discussion
about
versioning
is
probably
a
big
discussion
and
and
having
a
separate
meeting
about
that,
I
think,
probably
makes
sense
the
other
stuff
with
regard
to
paging
and
cursors,
and
things
like
that.
There's
it's
not
that
I'm
opposed
to
supporting
it.
It's
just
that
there
are
a
lot
of.
E
G
Yeah,
I
I
agree
with
that.
I
also
think,
on
the
one
hand,
all
of
figuring
all
of
this
out.
We
we
don't
want
to
make
it
a
bottleneck
to
that
prevents
you
from
getting
the
dynamo
db
working
aws
in
in
the
smithy
when
they
use
smithy
to
generate
interfaces.
They
have
developed
some
standards
for
pagination,
and
so
it's
a
similar
design
pattern.
They
use
for
hundreds
of
different
apis
you're
right.
C
F
E
A
like
a
cheat
code,
you
can
use
where
you
can
you
could
in
theory,
if
you
have
a
capability
provider
that
you
want
to
support
multiple
different
contracts,
you
could
in
theory
instantiate
it
or
start
it
multiple
different
times
and
each
time
you
started,
you
could
start
it
with
a
different
contract
id
I
understand,
but
that
doesn't
work
out
of
the
box
with
any
of
our
tooling
like
when
you
look
at
the
the
manifest
stuff
that
we
put
in
the
par
files.
Those
have
one
contract.
F
E
E
F
I
think
the
we
did
talk
about
in
slack
and
we
just
ended
on
the
convention
of
you
know,
throwing
the
rpc,
I
believe,
there's
an
rpc
error
not
implemented
a
new
variant.
I
don't
know
if
it
would
be
useful
to
have
a
specific
enum
variant
for,
like
anything
other
than
that,
but
perhaps
it's
just
something
that
you
handle
by
convention
and
then
ultimately,
actors
are
expected
to
see
that
if
they
get
a
particular
error
like
oh,
this
provider
doesn't
support
that
operation.
C
C
But
if
we
can
reach
a
point
where
we
do
integrate
something
similar
to
that
or
something
that,
on
the
dynamo
provider
side,
you
can
implement
the
skin
functionality
that
you
want
and
then
on
you
know,
on
the
redis
side,
we
could
do
something
different,
then
it
should
be
a
pretty
painless
switch
for
you
to
go
from
custom
interface.
That
has
it's
just
a
small
super
set
of
the
key
value
just
to
switch
to
the
awesome
cloud
official,
which
is
cool
yeah.
E
It
comes
down
to
the
there's
like
the
the
short-term
solution
for
how
we
take
care
of
the
scan
operation
and,
I
think,
like
brook
said,
there's
probably
a
pretty
provider
agnostic
way.
We
can
define
that
interface
so
that
you
know
both
redis
and
dynamo
can
implement
this
their
own
internal
ways
without
exposing
the
details
and
so
that'd
be
the
the
short
term.
But
then
there
are
larger,
more
philosophical
discussions
to
be
had
about.
F
F
A
Okay,
phenomenal
discussion
by
the
way-
and
you
know,
do
we
need
to
capture
any
tasks
on
you
know
our
github
to
sort
of
like
have
a
placeholder
to
have
those
discussions
or
you
know
kevin.
I
feel
like
some
of
those
are
going
to
ultimately
end
up
as
adrs
architectural
decision
records.
Is
there
any
way
you
want
to
track
that
stuff
on
there.
E
E
A
Yeah
that
I
feel
that
feels
like
a
good
approach
to
me
only
because
I
know
that
we've
got
a
lot
of
really
wicked
smart
people
here,
I'm
not
only
on
this
call,
but
folks
that
you
know
like
mats,
and
you
know
some
of
the
folks
that
are
really
regular
on
slack
but
are
in
a
different
time
zone.
So
I
think
it
would
be
really
important
that
we
include
their
perspective
and
give
them
a
chance
to
put
their
fingerprints
on
the
gun
before
we
pull
the
trigger
on
any
of
these
things.
A
A
Hey
man
just
try
to
bring
a
little
fun
to
architectural
decision
records,
I'm
open
for
suggestions
as,
as
usual,
never
make
anybody
bad
great.
Any
other
topics
roundtable
we
want
to.
You
know
we
want
to
touch
base
on
jeff.
I
would
ask
I
know
that
we've
touched
base
a
bit
on
the
importance
of
bindle
and
some
of
the
you
know
community
stuff
coming
down.
A
A
I
mean
I'm
not
nominally.
I
am
the
representative,
but
sorry,
sorry
liam
I
had
to
give
you.
Oh
you
are,
you
are
but
the
so
I
honestly
don't
know
exactly
I
mean
if
this
is.
This
is
human
politicking.
At
the
best
I
mean
it's,
it's
spec
writing
amongst
a
group,
so
we're
just
trying
to
decide
what
we
want
to
talk
about
right
now.
Really
so
it'll
it'll
take
a
little
bit,
but
this
is
normal.
That's
not
a
complaint.
That's
just
that's
how
these
things
work.
A
So
there's
there's
not
like
tons
to
report
there,
but
we
are.
We
are
having
conversations
those
registry
meetings.
Actually
I
have
to
go
to
one
here
in
just
a
second
they're.
Not
I
think,
there's
like
this
is
just
like
the
sig
member
meeting.
I
think
once
we
get
things
solidified
it'll
become
more
of
like
a
public
meeting.
Everyone
can
join
into
kind
of
thing.
I
I'm
still
getting
into
it.
So
I
don't
have
all
the
details.
A
Jeff,
I
know
the
privately
we've
touched
base
a
little
bit
and
you
know
we
taylor
has
been
doing
a
ton
of
thinking
inside
of
cosmonic
about
you
know
what
we
can
do
and
we
think
we
are
going
to
have
some
fun
open
source
stuff
later,
as
we
kind
of
get
into
this.
So
we'll
keep
you
in
the
loop
and
maybe
even
see
about
pulling
you
into
a
design
process
or
giving
us
some
early
feedback
on
stuff.
As
we
as
we
pull
our
plans
together.
A
I
want
to
make
sure
that
we
include
some
of
the
use
cases
that
you've
indicated
are
important
to
you
and
make
sure
that
you
know
this
is
something
you'd
look
at
and
at
least
consider
as
you
plan,
you
know
your
path
forward
over
the
next
few
years.
Yep
sounds
good
awesome.
Anyone
else,
I
think
we're
it's
been
a
phenomenal
meeting
today,
but
last
call
for
issues
today.