►
From YouTube: Lang Team Meeting 2019.05.09
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
A
A
This
is
the
it's.
What
happens
is
in
if
you
compile
a
rust
eyelid,
which
is
not
really
a
well
supported
use
case?
We
use
it
for
the
standard
library,
though
it
currently
embed,
done
some
whole
lot
of
symbols
in
there,
including
I,
think
this
symbol,
if
I
understand
from
all
your
see
dependencies
that
sometimes
overloads
Windows,
which
has
like
a
hard
limit
on
the
number
of
symbols
in
a
segment
or
something,
and
so
this
this
PR
kind
of
narrows
it
down.
There
is
some
crate
uncrating.
A
C
A
It
yes
right
right,
I,
think
one
thing
basically
I
think
we
should
do
is
document
that
it's
these
are
like
meant
for
internal
use,
and
not
really
you
can't
get.
You
can't
rely
on
much
from
them
and
at
some
point
we
probably
want
to
support
I.
Think
we
don't
yet,
and
at
that
point
we
would
probably
want
to
add
the
ability
for
people
to
control
what
sort
of
simple
thing
I've
exported
do.
A
A
A
Shall
we
write
was
a
consent?
Is
it
reasonable
to
write
a
statement
that
says
we
agreed?
These
are
not
documented
we'd
like
to
know
if
there
is
documentation,
maybe
where
we
could
document
it
and
we're
curious
if
you
could
actually
add
a
said
flag
of
some
kind
or
something
so
as
to
prevent
their
lives
from
being
used,
a
pardon
is
inter-library.
B
A
A
Okay,
so
say
APR
to
get.
It
was
a
bug.
Incoherence
from
2017
bug
has
been
fixed,
we've
didn't
like
slowly
letting
people
into
fixes.
There
is
still
about
20
crates
on
crates
that
I.
Oh,
that
hasn't
really.
Someone
did
a
lot
of
work
merging
all
like
opening
PRS
against
all
the
broken
crates
and
so
on,
all
of
which
have
been
merged
by
the
way.
I'm
gonna
try
to
move
me
because
I'm
getting
weird
feedback.
Okay,
that's
better!
Now,
I
think
it
was
Taylor
but
I'm,
not
sure
and.
A
A
Yeah,
we
could
do
that.
That's
actually
probably
a
good
idea.
I'll
write
up
a
report
and
yes,
here's
the
list
of
crates
Feynman's
curious,
but
what
I
mean
is
a
lot
of
not
fortress,
but
a
lot
of
0.1,
mostly
unsupported
sort
of
looking
things
but
I,
don't
know
it's
hard
to
say.
What's
in
here.
A
A
Scott.
Are
you
up
to
date
on
this
I
guess
we
can.
We
were
just
talking
about
it
because
third
time
but
I'll
give
the
real
brief
TL
DR.
If
you
have
a
variable
like
this,
you
do
this
and
you're
returning
something
and
let's
say
both
x
and
y-
have
a
destructor.
If
X,
if
the
destructor
for
X
panics,
then
the
destructor
for
Y
never
runs
the
reason
I.
Believe,
though
I
haven't
really
verified.
A
A
C
A
What
happens
with
other
variables
I
feel
like
we
should
if
we
should
go
and
write
a
whole
bunch
of
test
cases
for,
like
all
the
different
panicky
scenarios,
just
to
be
sure,
I
know
what
the
behavior
is
like.
What,
if
you
have
a,
let
say
here,
does
its
destructor
run
if
the
destructor
forex
panics
I,
don't
remember,
but
you
know.
A
Just
saying,
if
we're
gonna
talk
about
any
of
this,
we
should
like
right
there.
It's
not
bad
many
tests.
We
should
write
a
few
test
cases.
The
same
thing
with
struck
fields,
see
ya
figure
out
what
it
does
and
okay
well
who's.
Gonna
do
that
I,
don't
know,
but
I
guess.
Our
conclusion
from
this
meeting
is
we're
not
with
the
current
behavior.
We
consider
it
a
bug.
We
think
it
would
be
nice
to
to
sort
of
try
out
fully
document
the
space
and
then.
A
B
D
D
A
C
A
C
A
That
covers
the
things
we
had.
We
don't
really
have
a
good
action
plan
here,
but
that's
okay,
I
guess,
there's
we
we
we
didn't
decide
what
we're
gonna
do
today,
but
central
and
I
were
thinking.
We
could
do
some
like
general
triage
and
given
the
limited
attendants
that
sort
of
makes
sense.
One
thing
we
could
talk
about
is
it's
just
nominated
RFC.
B
B
A
Agreed
so
ampersand
raw
we
talked
about
this
is
ampersand
raw,
that
we
were
gonna,
try
to
schedule
that
this
I
wanted
to
go
to
this
one.
This
is
pointer,
metadata
and
V
tables,
and
the
last
time
we
talked
about
it,
I
basically
said
I.
A
A
A
B
Think,
there's
a
difference
between
like
saying
explicitly
that
this
is
something
wants
to
experiment
as
we're
as
having
a
bunch
of
stable
stuff
that
we
have
committed
to
that
art
adjustment.
In
both
we
mean
you
can
say
that,
okay,
we
don't
know
me
once
we
want
to
try
this
out.
Perhaps
it
makes
sense,
feel
free
to
implement
imagining
that
we
have
nuts
find
up.
D
B
D
Don't
know
I
have
this
feeling
like
we
are
bad
at
like
we
know.
We
want
something
here
and
we
definitely
need
a
big
write-up
for
it
eventually.
But
when
do
we
actually
do
the
big
write-up,
and
can
we
start
making
progress
on
little
pieces
of
it
somehow,
but
that
conflicts
with
we
want
the
holistic
thing
and
not
to
just
have
half-finished
things.
So
my
thoughts
are
all
over
the
place.
A
A
B
D
A
D
B
D
A
Think
I
would
call
this
the
it
would
be
a
custom
d'este
like
it's
part
of
a
costume,
DST
effort,
but
we're
not
ready
to
do
that.
But
we
are
ready
to
let
people
do
experimentation
under
the
like
expectation
that
we're
gonna,
wanna
holistic
view
before
we
stabilize
at
least
maybe
there's
like
tiny
pieces
comfortable
with.
A
A
D
A
A
B
B
Experimentation,
the
extreme
doing
experimental,
are
seen
as
more
of
a
general
from
these
from
my
side.
I
think
we
do
too
little
of
like.
Let's
try
something
either
might
eat
me
I've,
no
idea
this.
This
would
be
a
good
idea,
but
maybe
let's
find
out
I
mean
if
it
doesn't
take
a
good
amount
of
work
and
like
messes,
with
a
whole
infrastructure
of
either
like
if
it's
reasonably
feasible,
to
implement
an
experimental
pressure
that
doesn't
take.
You
generally
find
one.
B
A
A
A
C
A
C
B
B
A
B
A
A
B
A
B
B
A
A
A
A
A
B
B
A
D
I've
been
wondering
about
how
we
decide
if,
like
things,
Lintz
that
are
tied
to
specific
functions
actually
belong
in
the
compiler,
if
they're
part
of
the
language
or
if
they
become
part
of
cliffie
or
if
the
way
to
have
a
lint
in
the
compiler,
is
to
generalize
it
in
a
way
that
it
becomes
an
attribute
or
something
like
that.
As
we
have
like
you
know,
there's
a
lint
in
flippy
for
step.
I
0
which
just
panics
so
that
shouldn't
happen,
but
is
the
correct
answer
to
put
that
lynn
in
euler
or
just
to
say?
D
Use
as
an
attribute
like
that,
non-exhaustive,
it's
not
actual,
is
not
a
lint,
it's
sort
of
plenty,
but
no
yeah
you're
right,
it's
it's
semver!
So
it's
not
just
a
lot,
there's
a
bunch
of
like
if
we
can
find
patterns
and
things
and
turn
them
into
things
that
don't
have
to
become
limps
because
they're
just
attributes
that
would
feel
really
nice.
A
What
we
should
do
with
these
Lintz
I
think
I,
like
your
idea,
central,
except
that
I
think
we
should
try
to
get
a
smaller
group
of
people
who
feel
passionately
and
like
it
feels
like
something
where
it's
just
so
much
the
slog
like.
Do
it
RC
on
every
land
individually
trying
to
come
up
with
some
guidelines
and
like
working
through
example,
example,
but
coming
up
with
some
guidelines
and
then
saying
this
is
the
guidelines
we
did
here's
the
results.
We
did
I
think.
B
A
B
B
B
B
A
A
B
B
B
B
A
A
A
A
B
A
B
B
A
B
A
B
B
A
A
B
A
B
B
B
A
Yeah
I
think
we
should
wait
until
it's
harder,
but
it
seems
like
it
would
be
good
to
that,
and
we
should
try
to
get
some
estimate
of
when
that
might
happen.
By
starting
to
do
crater
runs
or
something
like
a
I.
Don't
know
if
it
that
estimate
that
my
forgave
was
very
informed
so.
B
B
A
Can't,
though
so
I
don't
know,
if
you
all
know,
but
we
are
now
doing
migration
mode
on
all
the
editions,
and
so
the
big
work
is
to
do
the
final
transition,
which
will
take
a
little
time,
but
we
have
to
kind
of
at
minimum.
Do
some
there's
some
bugs
we
have
to
fix,
and
then
we
have
to
do.
Some
crater
runs
to
see
what's
the
affected
crates
and
try
to
remove
them,
get
them
fixed
as
best
we
can,
but.