►
From YouTube: 2020-09-14 Lang Team Triage Meeting
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).
A
Okay,
so
we
have
one
pending
proposal
that
at
least
in
the
query,
which
is
this
question
of
allowing
different
preludes.
I
think
there's
really
no
change
here.
We're
still
wanting
to
do
this
still
lacking
sort
of
volunteers
is.
B
That
correct
there
is
actually
a
bit
of
an
update
on
this,
which
is
to
say
the
last
week
summary
that
probably
wasn't
a
complete
summary,
and
I
should
update
that.
But
we
last
week
because
of
our
discussion,
we
more
or
less
came
to
the
conclusion
that
this
really
needs
to
be
heavily
delegated
over
to
the
libs
team,
which
is
to
say
we
were
generally
in
favor
of
this
is
something
that
we're
prepared
to
do
some
work
on.
B
If
it's
something
the
libs
team
wants
and
has
a
proposal
to
use
and
that
the
actual
change
in
terms
of
what
should
be
in
the
new
prelude
should
be
up
to
the
libs
team
and
that
our
role
is
more.
Are
we
prepared
to
do
per
edition
preludes
and
just
change
the
part
of
the
compiler
that
says
what
prelude
do
you
import?
Based
on
what
edition
you
are?
B
A
C
Yeah,
I
was
just
going
to
agree,
I
feel
like
this
is.
If
libs
wants
to
do
this,
then
there's
some
work
from
the
compiler
team,
but,
like
I
guess
other
than
maybe
we
would
object
to
the
language
like.
It
doesn't
really
feel
like
something
that
we
really
are
involved
in
any
significant
way.
Maybe
there's
a
checkbox
somewhere
that
we
otherwise
were
like
kind
of
out
of.
B
It
yeah,
perhaps
we
could
just
fcp
this
mcp
and
like
decide
to
agree
what
that
fcp
means
is
we're
not
chartering
a
project
group.
We
need
to
decide
what
should
be
in
the
prelude
and
to
make
the
proposal
and
pitch
and
then
when
that
is
ready,
then
we're
prepared
to
review
and
sign
off
on
the
actual
mechanism
for
per
edition
preludes.
A
A
Yeah,
okay,
so
I
think
what
we're
saying
is
we
we
approve
of
this?
Someone
needs
to
do
the
implementation
work.
A
I
think
this
is
a
side
note
that
I
feel,
like
you
know,
drumming
up
and
directing
this
implementation.
Work
remains
an
interesting
challenge
for
us
as
an
organization
but,
and
it
kind
of
falls
on
us
and
kind
of
doesn't,
but
the.
A
A
I
guess
we
can
fcp,
I
tend
to
prefer
to
fcp
on
the
pr's,
and
I
think
I
would
advocate
for
that
also
here
versus
here,
because
it
this
is
very
high
level,
and
so
what
does
that
fcp
really
say
not
much
right
so
yeah?
I
would
be
happy
to
close
with
this
position
of.
We
are
approve
of
the
idea
and
we
encourage
implementation.
A
A
Okay,
great
we'll
circle
back
to
the
next
meeting,
but
that's
where
we're
at
that
sounds
fine.
Looking
at
the.
A
Looking
at
the
board,
I
don't
think
so.
We
have
some
pending
charters
and
things.
I
haven't
gotten
around
to
updating
this,
but
I
think
there's
no
new
proposals
there's
enter.
Are
there
any
comments
on
any
of
these
active
projects?
Const,
evaluation,
async
foundations,
safe
transmute,
there's
a
pending
rsc
for
safety.
Let
me
skim
through
and
look
but.
A
Like
starting
with
safe
transmute,
I
guess
I
wasn't
here
for
the
last
meeting,
but
well
I
guess
this
is
still
the
current
status
right.
People
should
look
at
the
rfc
and
post.
B
There
is
still
some
very
ongoing
discussion
about
the
mechanism
in
this
and
there
is
a
lot
of
trying
to
figure
out
exactly
what
that
should
look
like
and
avoiding
accidentally
exposing
the
ability
to
transmute
into
a
struct.
B
That's
happening
primarily
in
zulip
and
then
summaries
are
being
posted
to
the
rfc
thread,
but
yeah.
I
think
that
discussion
has
mostly
wrapped
up
and
is
getting
towards
like
what's
the
best
way
to
solve
the
problem,
but
there
seems
to
be
underlying
agreement
and,
like
here,
is
the
new
proposal
for
a
minimal.
B
Luca
thor,
you
made
a
comment
in
the
chat.
Would
you
describe
it
as
major?
I
haven't
had
a
chance
to
read
the
new
change,
yet
the.
D
Sort
of
the
minimum
viable
product
was
slimmed
down
from
the
previous
proposal.
As
I
understand
so,
it's
even
more
minimal
to
sort
of
create
basic
usefulness,
and
then
it
can
grow
into
the
previous
proposals
from
there.
D
Goals
the
minimum
amount
that
could
be
stabilized
and
deliver
something
useful,
while
not
accidentally
blocking
any
future
expansions
to
the
whole
overall
system.
A
D
I
think
that
the
updates
have
been
committed
into
the
rfc
so
reading.
It
now
would
be
a
useful
task.
I
see
okay.
A
All
right
sounds
good,
I
think,
on
the
foundations
I
to
be
honest,
I
haven't
fully
caught
up
from
the
last
week
or
two,
but
the
last
thing
I
know
of
is
that
we
were
going
to
post
the
stream
rfc
and
that's
still
true,
I'm
not
going
to
update
because
there's
nothing
to
change
here.
Does
anyone
else
have
thoughts
of
what
has
changed.
B
Not
something
that
has
changed
but
a
question.
Did
we
resolve
the
question
of
to
what
extent
this
is
a
lang
thing?
It
seemed
like
last
time
we
brought
up
the
notion
that
the
only
real
laying
interaction
is.
How
might
we
in
the
future
wish
to
handle
syntax
like
four
or
you
know,
generators
or
similar,
and
anything
else
would
very
much
be
libs.
I'm
wondering
if
that
has.
A
I
don't
know,
I
think
the
action
to
take
is
probably
I
guess,
who's
asked
to
we'll
tag
the
rsc
with
t-libs
and
make
a
comment
from
t-ling
about
these
language.
Intersections
that
sound
fine.
B
A
Yeah,
the
only
thing
I
would
also
point
out
is
there
has
been
some
work
on
a
towards
an
rfc
to
warning
about
values
that
are
live
across
yields
that
shouldn't
be
local,
though
you're
working
on
that,
or
am
I
mixing
that
up.
A
Async
things:
no
okay,
I'm
mixing
it
up.
I
forgot
who
who
it
is
somebody
else
with
an
l,
I
suppose.
Well,
okay,
there
has
been
some
work
towards
live
values.
Oh
yes,
I
remember
lucious,
but
the
idea
being
like
you
have
some
like
a
mutex
guard
that
is
live
across
a
yield.
A
A
Mutex
cards
are
not
sync,
and
so
they
can't
be
shared
so
futures
that
hold
them
across
yield.
Values
cannot
be
shared,
but
people
are
looking
into
a
more
targeted
and
then
similar
to
most
views.
I
haven't
caught
up
too
closely
with
that,
but
that's
probably
something
we'll
bring
up.
A
A
Josh,
I
see
your
comment
here
feel
free
to
edit
or
post
or
post
an
update
all
right.
Moving
on.
A
A
The
implementation
work,
as
far
as
I
know,
is
still
ongoing.
I
haven't
heard
from
the
person
yet
who
was
doing
that
at
least
not
super
recently,
but
we
don't
have
any
real
updates
just
that
batman.
Aod
was
interested
in
starting
to
to
push
on
that
a
little.
A
A
A
All
right
cool,
let's
jump
back
to
the
agenda,
so
project
board
review,
so
newly
created
rfcs
still
have
to
fix
the
thing.
This
makes
a
lot.
It
speaks
way
too
many
links,
but
we've
got
what
appeared
to
be
a
few
new
rfcs
to
me.
I'm
gonna
not
go
over
them
right
now
because
I
think
that's
probably
better
for
the
well,
I
was
gonna
say
better
for
the
backlog,
but
the
struggling
through
these
things-
oh
my
god,
there's
so
many
things
here.
A
Yeah,
that's
fine
skip
over
that
safe
transmute
was
nominated.
Did
we
already
cover
all
the
whatever
updates
are
needed.
A
A
So
this
first
one
75
502
use
implicit,
not
explicit
rules
for
promotability
yeah
we've
been
covering
this
for
a
while
looks
like
we're
ready
to
go.
Fcp
is
done.
I'm
going
to
remove
the
nomination,
then
nothing
more
to
say
here.
A
Ins
yeah.
E
So
we've
got
a
pending,
fcp
scott
nominated
this,
but
I'm
not
sure
why
maybe
just
to
tell
people
to
check
tick
boxes.
A
Yeah,
that
seems
fine,
so,
basically
right.
This
is
allowing.
A
Yeah,
I
remember
this
so
basically
we're
allowing
people
to
we're
slightly
extending
the
set
of
enums
to
include
those
with
no
variance,
so
the
rule
is
basically,
if
all
the
variants
do
not
have
data,
which
is
truly
true.
You
have
no
no
variance
and
that's
because
macros
sometimes
have
zero
variance
any.
I
better
check
my
box.
A
I
forget,
if
that's
the
syntax,
maybe
it
is.
Let's
move
on
lint
on
implicit,
abi,
okay,.
E
So
we
this
was
my
implementation
of
the
thing
we
discussed.
Probably
three
or
four
weeks
back,
the
crater
run
came
back
with
around
500
crates
broken
by
the
deny
lint
or
I
guess,
700
crates.
So
we
probably
don't
want
to
land
it
denied
by
default.
A
E
In
there
looked
too
out
of
the
ordinary,
just
people
not
having
the
api
on
various
functions.
A
A
C
C
E
Not
yet,
but
if
we
decide
to
land
it
as
something
that
is
worn
by
default
or
denied
by
default
regardless,
I
can
work
on
that.
A
E
I
think
my
personal
opinion
is
that
I
would
lean
towards
saying
that
we
don't
land
it
right
in
this
edition
and
we
make
it
in
addition,
idiom
lint
for
the
next
edition.
E
Deny
either
2024
or
2021
depending
on
sort
of
timelines?
And
whether?
A
C
A
Yeah,
so
so
I
think
the
plan
is
that
we'll
we'll
have
the
addition
can
make.
Some
things
go
from
warren
to
deny
like
can
affect
what
the
default
setting
is,
and
normally
we
would
go
to
deny
in
the
new
edition
and
warn
in
the
old
ones.
A
E
E
Patch
most
of
the
changes
are
from
test
failures
in
wrestling
rust
because,
like
half
of
our
ui
tests
have
extern
in
some
variant
or
another,
without
putting
the
ce
okay,
I
mean
I
don't
actually
care
if.
A
We
land
the
pr
as
it
is
or
not,
but
I'm
not
opposed
to
it.
Put
it
that
way.
A
E
Okay,
okay:
I
will
update
the
pr
and
put
a
comment
there
summarizing
our
discussion
and
then
find
someone
to
review
it
great.
A
C
What's
that
there
was
that
pr
that
there's
a
rfc
that
actually
okay,
I
can't
even
remember
what
it
was
so
again,
but
it
was
just
knox
made
up.
Pull
requested
got
a
lot
of
attention.
Last
week
around
dying
sized.
C
It's
just
2984
just
proposing
just
basically
to
get
proposing
downsized.
C
All
right,
it's
not
the
rc
itself
is
nothing
particularly
new,
it's
sort
of
just
bank
sized
and
a
way
to
get
the
meditative
of
any
given
type.
C
A
Now
I
guess
I
would
include
this
rc,
which
I
haven't
read,
but
I
do
feel
like
we
should
be
able
to
make
some
progress
here.
I
kind
of
feel
like
all
that's
missing.
Is
someone
to
survey
the
work
and
sort
of
talk
about
it
and
us
to
agree
that
we're
feeling
comfortable
like
a
lot
of
the
hard
work
has
been
done.
We
have
to
put
our
stamp
on
it
and
I
think
it
is
worth
reviewing
that
I
feel
like.
A
A
C
C
C
Cover
the
v
table
types
that
the
rc
has
so
it's
sort
of
a
so
I
think
the
motivation
for
this
rfc
was
that
the
author
wanted
to
restrict
something
to
types
that
are
a
single
word,
references
or
a
single
word,
so
that
would
be
include
extern
types.
So
that
would
be
like
you
know
something
that
it
is
in.
This
rc
would
be
question
mark,
dime,
size
plus
pointy
meta
equals
unit,
but
whereas
the
other
one
was
about,
you
know,
manipulating
the
table
type
and
stuff.
So
that's
like
two
different
motivations.
C
A
I
see
so
I
hadn't
yeah.
I
hadn't
wanted
to
open
that
bucket.
What
would
we
to
settle
the
question
definitively
in
favor,
like
in
favor
of.
C
B
So
I
still
generally
feel
the
same
that
I
did
then,
which
was
that
I
don't
think
we
should
introduce
this
distinction
in
the
type
system,
at
least
not
at
this
time,
and
that
I
don't
think
that
it's
a
blocker
to
doing
so-
or
I
don't
think
doing
so-
is
a
blocker
to
fixing
the
size
of
val
and
similar.
We
can
just
fix
those
to
panic,
so
I
don't
feel
like
that
is
inherently
tied
together,
but
I'm
also
fairly
open
to
being
convinced.
Otherwise,.
A
Okay,
that's
kind
of
where,
where
I'm
at
module
this
last
statement,
it's
hard
for
me
to
imagine
what
would
what
wouldn't
justify
such
a
it
was
sort
of
intrusive
change,
but
you
mentioned
in
the
chat,
I
think
josh,
that
this
desperately
wants
a
project
group.
What
were
you
referring
to
there.
B
Just
I
feel,
like
the
proposer
of
this
more
or
less
said.
Oh,
I'm
writing
one
rfc
because
it
feels
like
there
are
three
rfcs
in
this
area
and
I
want
to
make
progress
by
trying
to
find
common
ground
and
that
doesn't
sound
like
oh.
I
should
write
a
fourth
rfc
that
sounds
like.
Oh,
we
need
a
project
group
to
deal
with
a
coordinated
set
of.
How
do
we
solve
the
dst,
slash
metadata,
five
other
things
problem.
B
I
think
there's
a
lot
of
value
and
a
solution
to
the
problem.
Whether
we
should
do
it
right
now
is
a
different
question.
I
do
think
it's
a
problem
that
needs
solving.
I
also
suspect
that
it,
I
don't
want
to
say
overlaps
with,
but
rather
a
lot
of
the
people
who
are
currently
spending
a
lot
of
their
effort
on
safe
transmute
may
start
caring
about
that
once
safe
transmute
is
done
because
it's
a
reasonable
follow-on
to
say
hey.
I
want
to
do
interesting,
parsing
of
data
structures
that
may
be
variably
sized.
B
I
think
it
would
be
really
interesting
to
approach
the
safe,
transmute
folks
and
say
hey
after
you're
done
with
the
first
round
of
safe
transmute.
Would
you
be
interested
in
driving
forward
a
solution
to
the
dst
problem
and
trying
to
pull
some
of
this
together?
It's
a
group
we
have
experience
with,
and
it's
a
group
that
very
much
cares
about
this
class
of
problems
and
they
could
potentially
pull
in
another
person
or
two.
So
my
proposal
would
be
defer
until
safe
transmute
is
done,
but
start
talking
with
safe,
transmute
folks
to
say:
hey.
B
Yeah,
they
don't
overlap
in
function.
It's
that
the
same
people
care
about
both
they're
both
related
to
I
want
to
parse
and
work
with
interesting
binary
data
structures
from
memory
safe,
transmute
gets
you
the
initial
round
and
dsts
come
in
when
you
want
to
start
saying.
Oh,
I
have
these
27
different
data
structures
that
I'll
start
with
a
fixed
size,
header
and
then
it
might
be
a
bunch
of
different
sizes.
B
C
Yeah,
I
would
also
because
I
think
about
problem
space.
It
can
get
pretty
sprawling.
You
know
from
external
types
to
custom,
dsts
to
wanting
thin
trade
objects,
and
so
on
it'd
be
nice.
It
built
something
that
you
know
allows
a
lot
of
that
to
exist
in
the
crate
ecosystem
and
not
built
into
the
language.
A
A
I
see
local
thor,
also
posted
a
comment
saying
sleepless
plus
interrupt
people
seem
to
have
interest
like
I
think,
that's
what's
missing
for
me
and
would
be
useful,
is
both
elaborating
something
about
what
exactly
the
scope
of
the
group
would
be,
but
also
connecting
it
to
here
are
the
kinds
of
use
cases,
and
those
two
are
very
compelling
to
me:
safer,
unsafe
and
and
seatbelts
plus
interop
are
things
I
think
are
really
important,
but.
B
Yeah,
it's
worth
noting
that
you
can
get
pretty
far
here
with
purely
unsafe
code
and
no
additional
lang
support
in
terms
of
dealing
with
in
memory
data
structures
that
don't
really
have
compiled
time
understood
size.
The
point
of
teaching
the
language
more
about
this
should
be
to
be
able
to
do
it
with
more
safe
code.
C
A
A
Kind
of
goes
against
this
established
consensus,
though
it'd
be
good
to
skim
it,
but
so
somebody
could
leave
a
comment
to
that
effect,
but
I
would
like
to
follow
forward
with
rfc
2580
or
at
least.
A
A
For
the
extern
use
case,
okay,
and
about
the
should
does
the
design
meeting
talk
about
rfc
2580
then,
and
bring
some
of
this
back
in
scope
makes
sense.
The
only
reason
I
can
see
not
to
is,
if
we
wanted
to
say
you
know
what
let's
just
not
even
open
the
door,
because
we
might
want
to
do
a
more
thorough
group
around
this,
but.
B
B
A
B
E
C
A
A
What
does
this
introduce
that
might
cause
problems
with
some
of
the
ways
we
might
want
to
go,
but
but
I
do
think
this
more
general
question
is
interesting.