►
From YouTube: Rust and WebAssembly Working Group Meeting 2019-06-27
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
Hey
welcome
everyone
thanks
for
coming
out
for
another
Rustin
webassembly
working
group
meeting.
So
the
first
item
on
the
agenda
is
RFC
triage
and
there
are
not
any
new
RFC.
So,
let's
move
along
second
is
our
six-week,
sprint,
planning,
stuff,
I,
don't
think
we
have
too
much
to
say
here,
but
there
is
a
link
to
unclaimed
issues
that
we
have
kind
of
identified
as
a
priority
for
the
next.
However
many
weeks.
B
B
B
All
right
regarding
debugging,
there
was
a
session
at
the
recent
web
assembly
face-to-face
meeting
about
two
bucking
and
it
was
very
much
like
not
even
straw
proposal
kind
of
level
like
just
like
suggestions
for
directions,
and
so
my
hope
is
that
the
sub
charter
that
was
created
for
debugging
wasm
is
gonna,
start
picking
up
meetings
again
soon
and
I
don't
know.
Does
anyone
have
the
links
to
all
of
the
slides
and
stuff
that
were
talked
about
at
the
face-to-face
meeting
and
I?
B
B
C
B
We
have
web
IDL,
bindings
or
snowman
bindings
or
whatever
they
end
up
getting
called
and
he's
been
doing,
like
basically
rewritten
Wesson
by
Jen,
maybe
two
or
three
times
now
in
the
last
three
weeks
and
separately
from
that
I've
been
working
on
a
crate
just
to
kind
of
represent
Webb,
ideal
bindings
and
parse
them
and
omit
them
and
everything
both
the
text
and
binary
format
and
yeah.
The
the
eventual
goal
here
is
to
make
wasm
bind
Jen
and
ahead
of
time
polyfill
for
web
IDL
bindings.
B
B
Right,
sweet,
so
Aaron,
hello,
hello,
hey
so
I
got
invited
here
by
Ashley,
I
went
to
Jas
comp,
you
and
they
were,
like
you
usually
join
I
was
like
cool
so
and
I,
see
like
I,
know
like
half
the
people
in
the
chat,
which
is
pretty
crazy
but
yeah,
so
I
developed
by
some
boy.
I,
don't
know
if
that
means
anything,
but
so
I've
been
doing
a
lot
of
just
like
general
in
my
free
time,
like
hey
what
was
Emily's
cool.
So
that's.
B
Why
I'm
here
I'm
also
working
with
the
assembly
strip
team,
so
I
I,
don't
know
I'm
giving
so
much
context.
Anyways
one
thing
we're
trying
to
do
right
now,
or
at
least
we're
trying
to
do,
have
any
trying
to
pitching.
This
is
kind
of
like
making
assembly
script
like
a
major
or
not
major,
but
a
more
organized
open
source
project
like
a
I,
don't
know,
can
I
share
my
screen.
I,
don't
know!
If
that's
worth
sharing
my
spoon,
you
should
be
able
to
alright
cool
cool
I'll.
Just
do
application
window.
B
Do
application
window
now
I
just
do
my
entire
screen
I
always
I'd
like
the
hardest,
like
which
one
should
I
choose
boom.
Is
it
working
yep
all
right
cool?
Thank
you
all
so
nice
to
meet
you
Nick
I
watched
your
talk
before
yeah,
but
one
thing
I've
been
trying
to
do
and
like
talking
with
Daniel
the
author
of
assembly,
script
is
kind
of
start
getting
things
organized,
so
I
asked
them
to
create
like
a
meta
repo,
almost
in
some
sense
and
we're
trying
to
start
like
a
core
team,
and
things
like
that.
B
So
that
way
we
kind
of
have
a
place
to
be
like
hey.
This
is
like
what
we
do.
This
is
what
our
goals
are
and,
like
you
know,
we
have
a
team
here.
If
you
want
to
contact
us
things
like
that,
basically
probably
talked
to
you
about
it,
so
on
and
so
forth.
With
that
I'd
like
to
ask
my
first
question:
is
there
a
way
that
we
could
kind
of
work
together?
B
That
being
said,
I
feel
like
rust.
Right
now,
personally,
is
more
of
like
the
will
like
do
a
lot
of
the
glue
code,
for
you
would
do
like
a
lot
of
higher
level
abstractions
and
things
like
that
and
we've
been
trying
to
focus
more
on,
like
smaller,
like
I,
want
to
pass
buffers
and
pointers
back
and
forth,
and
that's
really
about
it.
You
more
like
I,
guess,
lower-level
type
stuff
where
just
because
our
team
is
smaller,
we're
starting
to
do
things
like
at
a
runtime
for
garbage
collection
and
things
like
that,
but
yeah.
D
B
If
I
may
share
my
screen,
one
more
time,
yeah,
so
boom
I
still
need
to
work
on
like
getting
it
more
known,
but
I
just
literally
launched
it
like
I
think
two
days
ago,
and
one
thing
is
just
a
lot
of
like
small
introduction
like
hey.
This
is
how
you
like
go
into
end,
build
a
small
hello
world
and
like
there's
some
links
to
like
there's
somebody
script
compiler
and
things
like
that
and
how
to
figure
that
out:
yeah
yeah
yeah
and
then
it
also
has
a
language
switcher
up
here.
B
So
you
can
choose
like
oh
I
want
to
do
rest
instead
and
then
I
have
like
a
whole
thing
here
about
like
how
to
do
this
in
rest,
so
yeah,
so
I'm,
gonna
topic
of
that
too.
If
someone
would
be
so
kind
with
some
leg,
look
through
this
and
make
sure
I
can
write
rest
cuz.
This
sex
is
the
first
time
ever
in
rush,
so
I'm
kind
of
like
I,
feel
bad
about
this,
but
Surma
told
me
it
was
okay,
so
I
was
like
okay.
B
Yeah,
does
anyone
want
to
to
look
through
that
sari
or
the
URL
in
the
agenda
so
that
we
can
definitely
repo
or
like
how
would
you
like
to
receive
yeah?
Definitely,
the
github
repo
probably
be
best.
If
you
know
I,
don't
want
to
like
push
on
like
hey
help
me
with
my
project,
but
yeah
I
mean
definitely
your
agenda
once
I
log
into
Dropbox,
and
then
yes,
but
yeah.
So
maybe
it's
something
I
can
come
back
in
the
next
meeting.
A
movie
could
be
like
yeah.
These
are
some
ideas
we
brainstormed.
A
The
sort
of
higher
level
bindings
would
be
a
really
great
fit
for
assembly
script.
Probably
it's
just
a
lack
of
resources
that
have
this
led
to
assembly,
scripts
being
so
low
level
right
cuz,
like
yeah,
you
know
you
can
pass
I
thirty
twos
back
and
forth
with
like
the
minimal
amount
of
effort,
but
something
like
assembly
scripts.
Like
you
know,
it
would
obviously
be
great
if
we
could
express
richer,
richer
types
and
passed
those
back
and
forth
seamlessly,
though,
maybe
that's
one
one
way
to
collaborate.
B
Definitely
yeah
yeah.
We
could
definitely
learn
from
y'all
on
like
terms
of
a
high
level
things
I
know.
A
lot
of
the
people
working
on
the
actual
compiler
are
just
reserved
about
implementing
a
lot
of
that
really
specs
when
they
know
they're,
gonna
change.
So
that's
kind
of
one
of
the
major
concerns
I
got
kind
of
pushed
on
like
hey.
B
Maybe
we
should
like
have
garbage
collection
and
they
totally
wrote
like
a
really
lightweight
runtime
for
it,
but
that
they're
still
very
strong,
like
we
cannot
wait
to
remove
garbage
collection
once
it's
actually
in,
like
the
browser
so
stuff
like
that
and
like
yeah,
so
yeah
I
guess
that's
kind
of
how
I
see
it
sitting.
If
that
makes
sense,
I
don't
know.
B
I
can
definitely
ask
and
I
would
be
surprised
if
they
said
no,
so
yeah
we're
just
yeah,
it's
open
source
project,
so
I
haven't
seen
them
say
no
to
anyone
before
I
mean
yeah,
so
yeah
cool
but
yeah
I
know
someone
else
us
to
talk
about
something
and
I
could
like
stop
talking.
But
yes,
I
just
want
to
say,
hi
and
be
like
this
stuff
I'm
doing.
Let
me
know
if
I
could
help
you
all
and
if
we
can
work
together,
yeah.
B
Before
we
move
on
to
the
next
topic,
what
is
the
tool
chain
for
assembly
script
looks
like
I
know
it.
It's
a
pretty
much
Daniel
like
decode
I/o.
They
wrote
compiler
and
C,
and
then
that
gives
you
binary
inand
then
binary
and
on
kawazu
modules.
Okay,
yeah
yeah
I
was
thinking
like
there's
a
story
for
linking
C
and
C++
with
rest
because
you
can
use
clang
and
as
long
as
you're
on
a
compatible
LVM,
everything
should
work.
It's
a
little
finicky
to
set
up
but
yeah
it's
with
with
like
completely
different
tool
chains.
B
I
think
linking
is
a
bunch
more
work,
but
the
other
alternative
is
kind
of
these
shared-nothing
modules
that
just
communicate
where
you
instantiate.
You
know
an
assembly
script
module
the
instantiate,
a
rust
module,
and
then
you
have
them
talk
to
each
other
via
imports
and
exports.
Or
something
like
that.
B
You
know
the
web
IDL
bindings
is
kind
of
the
intended
way
that
those
api's
will
be
expressed,
and
so
maybe
assembly
scripts
could
use
the
web
IDL
bindings
crate
that
we're
working
on
in
order
to
implement
that
and
then
we'll
also
own
and
then
we'll
be
able
to
you
know,
bridge
the
gap
that
way.
Yeah,
that's
actually
a
really
good
idea.
One
thing
Daniel
was
telling
me
like
yesterday
was
like
yeah
I
wish
like
we
could
find
some
type
of
like
downstream
dependencies.
That
could
do
some
of
this
work.
For
us.
B
A
It
also
talked
router
on
our
team
I
watered
yep,
yeah,
okay
wrote
a
like
C++
header
that
implements
in
a
streaming
interface
for
like
streaming
web
assembly
very
similar
to,
like
you
know,
binary
ins,
C
interface
for
streaming
binary,
annoying
art,
except
it
also
adds
in
all
of
the
extra
stuff
that
you
need
to
get
a
proper
wesam
object.
File
automatically.
D
A
A
So
right
now,
at
least
in
a
scripted.
What
we
do
is
when
you
create
a
multi-threaded
module,
because
it
doesn't
use
both
memory.
It
has
to
take
the
data
sections
and
move
them
into
a
separate
file
and
then
add
code
to
load
the
initialize
memory
by
reading
that
file-
and
this
is
to
avoid
a
situation
where
the
memory
gets
reinitialized
in
every
thread
that
you
spawn,
because
just
a
normal
passive
segments
get
initialized
on
instantiation
and
each
thread
is
doing
a
sum
of
instantiation
the
wise
module.
A
A
Breaks
Firefox
a
little
bit
with
threads
right
now
they
have
a
sort
of
a
flag.
You
can
toggle
in
about
config
to
turn
on
shared
memory
which
right
now
enables
threads,
but
it
doesn't
turn
on
book
memory,
so
that'll
be
insufficient
and
I've
already
dropped
an
issue
in
the
Firefox
bug
tracker.
Just
letting
them
know
about
that.
But
I
wanted
to
ask
this
group.
First
of
all,
how
does
the
rust
tool
chain
do
this
right
now
separate
the
active
data
segments
out
and
yeah?
A
B
Well,
so
it's
not
gonna
just
break
our
tool
chain
immediately,
because
we,
you
know
actively
update
LVN
right.
So
until
we
update
LVM
we'll
be
fine.
I
wish
Alex
were
here,
but
he's
not
available,
excuse
the
one.
You
actually
implemented
this,
but
we
essentially
polyfill
all
this
stuff
in
wesen,
fine,
Jen,
I
guess
it
sounds
similar
to
how
I'm
scripting
was
pulling
things
out
into
a
separate
file.
B
We
don't
put
them
in
a
separate
file,
but
we
make
everything
passive
or
not
passive,
so
that
it's
manually
initialized
and
then
somehow
in
the
start
function
we
determine
whether
or
not
we
are
the
main
thread.
I.
Don't
really
remember
what
the
mechanism
for
doing
that
is,
it
might
just
be
like
an
import
or
something,
and
then
we
branch
on
that
and
the
main
thread
is
the
one
that
initializes
all
of
the
data
segments
and
then
everything
else
doesn't
I.
B
Think,
okay,
but
yeah
like
definitely
we
don't
have
a
threading
story
without
the
non
passive
data
segments
because
there's
not
really,
as
you
said
like
you,
don't
want
to
initialize
them
multiple
times
and
overwrite
and
get
data
races
and
it
doesn't
make
sense
so
we'll
probably
need
to
modify
our
pass
and
Wes
mention
that
that
implements
this
stuff.
But
I,
don't
know
that
seems
fair
game.
Okay,.
A
Sounds
good
sounds
like
you'll
get
to
delete
a
bunch
of
code
because
the
linker
is
going
to
automatically
insert
a
function
that
initializes
all
of
the
memory
sort
of
at
the
right
time.
As
long
as
you're
calling
double
underscore
wise,
underscore
or
call
underscore
C
tours
at
some
point,
because
that's
that's
where
it's
gonna
get
inserted
and
since
you're,
just
using
LLVM
I
assume
you're
calling
that
somehow.
A
B
A
B
A
B
So
right
now
we
don't
well.
The
rest
doesn't
have
global
constructors
at
all.
So
I've
never
seen
that
symbol
in
any
wesen
binaries.
We
provide
as
nothing
would
reference
it
right.
I
see
right,
yeah,
okay,
yeah
we
as
long
as
the
stuff
is
like
documented
somewhere
or
something
yeah,
I.
Think
we'll
be
fine.
Okay,.
B
B
D
B
B
So
in
general,
like
from
just
a
development
experience
like
you're
gonna,
have
a
lot
better
time,
writing
test
cases
and
then
testing
them
in
the
suite
then
trying
to
run
all
the
tools
by
hand
because
generally
Wes
impact
is
the
thing
that
makes
like
wrangling
all
these
tools
easy
and
it
doesn't.
Support
like
using
local,
builds
right
now,
so
it'll
it'll
be
easier.
B
If
you
stay
within
the
test
suite
because
then
all
the
wrangling
is
done
for
you,
but
so
for
for
a
resin
pack
itself,
all
the
tests
use
this
kind
of
project
builder,
which
allows
you
to
kind
of
lay
out
different
files
and-
and
there
should
be
examples
in
the
test
directory
of
wasm
pack
and
I
suspect.
That's
where
a
bunch
of
this
implementation
is
gonna,
end
up
being
I
think
there
is
probably
some
implementation
needed
in
wesen
bind
gen,
and
for
that
I'm
less
clear
on
exactly
where
that
code
would
go.
D
But
yeah
I
guess
I'm
just
interested
in
like
like
how
was
wasn't,
bind
Jen
developed
like
without
like
it,
if
it's
just
like
I'm,
not
used
to
just
having
like
like
test
cases
for
development
purposes
like
I'm
used
to
being
it
like,
interact
with
tools
but
like
it's
obviously
a
very
like
there's
a
big
project.
Obviously
people
work
on
it
but
like
what's
the
yeah.
B
So
it's
mostly
integration
tests
that
are
doing,
like
you
know,
here's
an
interface
and
rust.
Here's
like
a
JavaScript
implementation
of
that
interface
and
make
sure
they
can
talk
to
each
other
back
and
forth,
and-
and
so
that's
that's
where
most
of
the
existing
tests
are,
is
kind
of
like
flexing
all
the
different
options
that
are
essentially
importing
JavaScript
functions
and
exporting
us
functions
and
right,
Ruxton
classes,
and
so
there
is
less
infrastructure
there
for.
B
B
Yeah
I
know
so
I
guess
like
that,
would
be
you
just
the
way
that
it
usually
works.
Is
you
add
some
new
binding
feature
and
write
a
test
that
you
know
uses
that
the
nice
needs
a
slightly
different
case
right,
where
we
actually
are
carrying
about
the
the
contents
of
these
artifacts
with
it
mark
exactly
yeah.
D
B
B
Actually,
if
you
don't
mind,
I
say
one
more
thing:
yeah
or
I
swim,
I
think
I
a
break
by
the
way
I
haven't
seen
you
in
a
while.
That's
crazy,
we're
both
here.
Okay,
so
think
I
want
to
ask,
is
in
terms
of
tooling.
What
have
you
found
very
useful
for
that
rest
community,
like
I,
know
wise
and
back
does
like
a
ton
of
the
stuff
for
you
in
terms
of
like.
Oh,
you
have
a
roast
module.
It
does
all
the
compilation
gives
you
a
web
pack,
there's
module
things
like
that.
B
Part
of
it
is
about
NPM
packages,
but
not
all
of
it.
Actually,
a
lot
of
it
is
just
like
having
a
single
tool
to
wrangle
all
the
other
tools,
so
I
suspect
assembly
script
might
suffer
from
this
a
little
bit
less,
because
it's
more
targeted,
specifically
at
wesam
but
like
compared
to
native
development,
with
rust.
B
Plasm
development
with
rust
introduces
a
lot
of
extra
tools
that
you
have
to
do
and
kind
of
like
just
extra
bits
that
you
don't
have
to
deal
with
on
native
and
those
are
largely
different
tools
like
the
wise
and
pungent
tool
has
mopped
tool,
wesam
snip,
wisdom
strip.
You
know
all
the
wabbit
stuff
right
and
like
having
one
tool
with
one
place
to
do
configuration
or
option
flags
to
manage.
All
of
those
things
is
like
a
huge
workflow
boon
like
even
just
typing.
B
Woz
impact
build
instead
of
cargo,
build
jeff
target
less
than
32
unknown
unknown
is
like
it's
great.
I
hate,
typing
Jeff
target,
wasn't
pretty
oh,
no,
no,
no
yeah!
So
I'd
say
that's!
That's
one
of
the
the
big
benefits
of
lesson
pack
and
then
the
other
is
wesen
by
gen
is
the
other
big
tool
that
we
have
and
that's
all
about,
the
integration
with
javascript
and
web
api
sand
and
stuff
like
that
and
okay,
yeah
nice,
so
I'm
a
topic
of
wise
impact.
Especially,
is
there
any
value
and
not
using
it
like?
A
B
The
CLI
the
II
use
the
last-mentioned
has
to
be
the
same
CLI
that
you
are
the
same
version
of
the
library
that
you
use
to
compile
Azzam,
or
else
they
won't
understand,
like
the
description
of
what
the
signatures
look
like
and
so
just
having
a
tool
that
makes
sure
that
those
are
always
the
same
version.
It's
really
nice,
okay,
cool
yeah.
Personally,
one
thing,
I
kind
of
felt
with
was
impact.
B
I
read
a
lot
of
JavaScript,
so
I
was
a
little
confused
while
I
was
generating
so
much
JavaScript
for
me,
so
I
was
trying
to
use
the
rusty,
compiler
and
then
I
realized.
Oh
wait
wasn't
packed,
does
cool
stuff
for
me
and
back
so
yeah.
That's
why
I
was
asking
if
X
I've
been
writing
a
bunch
of
rust
examples
and
I
want
to
make
sure
I'm,
not
writing
bad
examples.
So
you
know
I
asked
that
and
yeah
also
one
last
thing
about
assembly
script
we're
still
trying
to
working
as
a
project.
B
Don't
let
me
be
like
the
end-all
yeah
he's
the
assembly
I
mean
alright,
so
something
we're
trying
out
so
yeah
mm-hmm.
But
thank
you
I
appreciate
all
your
help.
Every
has
my
head
thinking
of
ideas
and
stuff
one
thing
that
might
be
useful
for
the
assembly
script
project
in
terms
of
like
organization
and
team
structure,
yeah
you're,
looking
at
like
how
RFC's
work
in
rust
and
in
particular,
I
think
having
like
a
road
map,
we
do
it
every
but
could
be
a
different
cadence
is
really
nice
for
like
providing
direction
for
the
community
and
contributions.
B
Deaf
I
am
yeah
yeah.
Thank
you.
Yeah
I
really
needed
that
I'm
sure
you
could
tell
and
yeah
yeah
someone
else
actually
mentioned.
We
should
maybe
have
our
C's
and
I
would
like
I.
Don't
know
those
are,
but
I
looked
into
it
a
little
bit,
but
now
I
look
into
it
more
because
yeah
I'll
throw
a
link
in
the
office
hours
section
to
the
Russ.
Wiser.