►
From YouTube: Lang Team Triage Meeting 2019.11.07
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
B
Ecstatic
voice
you've
got
a
lot
here.
Yeah,
there's
that
sorry
I'm
a
bit
verbose,
there's
not
a
lot
that
it's
a
urgent
concern.
Just
there
was
a
bit
more
discussion
about
what
we
talked
about
last
week
with
the
conclusion
being
that
it's
not
going
to
cause
any
new
issues
to
allow
something
like
Beck
new
in
an
array
initializer
that
aren't
already
cooked
caused
by
the
snippet
of
code
I
posted
here,
and
that
because
we
already
allow
this
there's
some
implications
down.
The
line
like
heap
allocation
and
constants
will
require
some
extra
hoop
jumping.
B
A
There
is
like,
like
I'm,
just
gues
I'm,
just
saying,
there's
a
difference
in
that.
That
is
a
context
where
side
effects
can
occur
today.
Long
as
the
resulting
value
from
the
evaluating
the
expression
is
copy,
you
can
have
whatever
you
want
in
that
arrange,
whereas
a
Const
initializer,
that's
not
the
case,
so
I
don't
know
how
relevant
that
is,
but
it
feels
like
it
might
make
a
difference.
C
B
C
B
E
E
A
I
mean
it
just
comes
down
to
I.
Think
this
case
actually
is
not
so
bad,
but
usually
when
you're
saying,
if
especially
on
the
coercion
czar,
something
it
implies,
that
you
can
decide
that,
and
it
might
be
that
you
need
type
information
later
on
to
make
that
decisions,
and
that
could
be
the
case
here
too,
but
like
it
could
be
that
the
full
type
of
that
expression
is
not
locally
inferred
but
results
from
later
statements.
We
haven't
seen
yet.
E
A
We
don't
know
if
it's
copy
or
not
that's
we
mean
we
have
to
defer
the
check,
whether
it's
constant,
which
we
probably
can
do
so
like
architectural.
A
this
might
be
a
case
where
it
works,
not
just
fine,
but
besides
their
concerns,
that's
Kaba.
Reading
I
would
raise
those
I
think
this
consultation
stuff
happens
actually
pretty
late
now,
I
think
that
it
like
a
mirror
yeah
yeah.
A
B
F
B
F
B
B
B
Point
in
my
summary,
the
Diagnostics
PR.
If
we
don't,
if
we
do
it
all
the
time
and
don't
wait
for
the
mere
contractor
to
fail,
it's
going
to
break
some
code
that
we
currently
allow
unstable.
Where
you're
have
a
loop
that
unconditionally
exits,
because
by
the
time
it
gets
lowered
to
mirror
it
doesn't
actually
have
any
like
things
that
are
typical
of
loops.
A
There's
no
real
updates
for
unwind
FFI,
or
we
just
didn't
have
time
to
do
much
for
the
didn't
rate
thing.
I
talked
about
last
time.
This
proposal,
that
was
something
of
a
breaking
change
to
notice.
It
is
sort
of
inevitable
I
did
a
crater
run
I
found
that
a
lot
of
crates
are
affected,
but
I
haven't
look
how
much
of
that
root,
breakage
versus
derived,
crates
and
I.
Could
I
haven't
analyzed
it
at
all
to
see
that
could
mean
a
lot.
A
A
B
G
G
No,
no
I'm
wrong,
I'm
wrong,
I,
misremembered,
I,
misremembered,
okay,
now
I'm.
What
this
is.
Thank
you,
okay.
This
was
something
this
is
something
we
talked
about
and
the
issue.
This
is
the
one
that
some
we
had.
A
discussion
and
boats
was
gonna,
leave
a
comment,
but
then
the
ylang
teaming
was
like
rapidly
approaching
and
I
decided
for
some
reason,
I
decided
to
take,
take
charge
and
leave
a
comment
stead
trying
to
representing
teams
perspective,
and
this
is
the
one
that
was
about
a
about
having
proc
macro,
attribute
style,
macros.
G
That
gonna
apply
the
modules,
and
the
thing
is
that
as
written,
it
only
is
this.
This
proposal
is
just
to
stabilize
it
solely
on
inline
modules
and
when
I
read
our
discussion
and
slash
were
on
my
memory
of
our
discussion.
I
had
thought
that
our
whole
problem
that
we
had
had
at
least
I
I
might
miss
remember
at
the
actual
situation
or
misunderstood
situation.
When
we
had
the
conversation,
I
thought
it
was
something
about
like
in
turn,
attributes
for
Zadar
attributes
in
terms
what
we
were
stabilized.
G
Some
reason
I
was
in
the
impression
that,
like
outer
attributes
on
out
of
line
modules
would
be
fine
and
would
work,
but
in
fact
that's
not
the
case
at
all.
This
is
truly
only
stabilizing
attributes
on
inline
modules,
not
any
kind
of
out
of
line
module,
and
so
at
first,
when
I
was
reading,
this
I
was
like
once
I
came
around
understand
situation.
G
I
was
thinking
to
myself.
Okay
is
the
situation,
one
where
we
should
stabilize
what's
being
proposed
and
then
I
switched
to
being
like
well
wait,
I,
don't
even
know
if
this
even
makes
any
sense
to
try
to
stay.
Look
it's
our
whole
problem
that
we
had,
or
rather
concern
that
was
raised,
was
do
we
have
to
stabilize
this
now.
G
So
it
seemed
to
me
like
petrol
tank
hubs,
response
it's
kind
of
like
well
it'd,
be
nice
and
stabilize
it
because
we're
stable
because
there's
other
features
around
macros
that
are
gonna
land
in
the
you
know,
1.40
cycle
I
personally
am
NOT
too
swayed
by
that
argument
myself,
but
maybe,
more
importantly,
if
somebody
else
posted
I,
you
skip
their
own
use
case
for
wanting
to
have
this,
even
if
is
only
for
inline
modules.
I
didn't
read
that
I'll
be
honest.
G
I
didn't
read
the
comment
they
posted
too
carefully,
but
at
least
there
are
other
consumers
out
there.
That
said,
they'd
have
a
use
for
this,
so
I
wanted
to
raise
it
through
the
team
with
the
meeting
with
the
meeting
again
in
part,
because
I
wanted
I
wasn't
even
sure
if
we
like
I,
said
I
may
have
missed
remembered.
G
The
situation
of
the
previous
meeting
might
be
everyone
understood
for
me
at
that
during
that
conversation,
but
I
want
to
be
clear
in
terms
of
what
is
supported
here
and
you
know
make
sure
you
all
understood
that
and
then
maybe
revisit
revisit
the
question
of
okay.
Well,
do
we
only
want
to
stabilize
this
one
piece
as
it's
being
suggested
here
or
do
we
want
to
push
back
and
say
no,
it's
better
to
wait
until
we
have
support
for
both
in
line
and
out
of
line.
G
G
If
you
can
call,
it
can
say,
wasn't
consensus
in
any
case
amongst
the
line
team
was
more
like
I.
Don't
think
I
was
your
person
the
concerns
ranking
properly
at
that
point.
So
I
wanted
to
be
racing
with
us
to
get
a
discussion
here
again.
F
G
H
H
Was
too
and
I
thought
that
was
sort
of
like
rather
complicated
rules
as
far
as
I
know,
we
also
just
want
the
semantics
that
we
don't
preload
the
modules
and
we
want
if
the
met.
If,
like
the
macro,
wants
to
read
the
inside
of
an
external
module,
there
should
be
some
sort
of
API,
so
I,
don't
know
why
I'm
just
just
side
those
semantics
and
stabilize
them
on
all.
F
F
H
I'm
not
sure
what
we're
factoring.
Do
you
mean
necessarily,
but
I
also
think
that
we
could
stabilize
it
without
having
the
API
to
let
you
you
know,
II
really
load
modules
useless.
It
could
be
that
if
you
put
the
attribute
on
the
external
mod
statement,
you
know
all
you
get
is
some
on
statement.
You
can't
really
do
anything
with
that.
H
G
The
only
questions
about
the
use
case
that
Robbie
pop
posted
in
the
read,
but
even
there,
it's
like
something
where
they
list
this
and
it's
like
they
said
this
is
it's
funny.
They
say
this
is
basically
one
of
the
two
remaining
lovely
features
we
are
using.
So
it's
not
that
this
is
the
last
one
remaining,
but
it's.
D
A
Seems
like
the
semantics
of
lazily
loading
modules,
I
tend
to
agree.
That's
probably
what
we
want
well
is
this
something
that
was
in
an
RFE
that
was
decided
like
how
are
these
just
where
and
how
is
this
discussion
happening
and
I
would
feel
pretty
good
out
saying
we
can
stabilize
this
without
API
to
make
it
useful
if
I
felt
good
that
we
had
decided
the
semantics
but
I
sort
of
feel
confused
about
that.
Well,.
H
A
G
F
Some
issue:
six,
four
one,
nine
seven
so
at
the
moment
a
face
condition
of
confirmation
works
by
they
like
when
you,
when
you
put
an
attribute
module
on
the
phone
margin,
then
that
will
directly
call
the
punishment
population
in
the
parcel,
which
is
a
quite
a
weird
thing
to
do.
But
nice.
But
that's
one
of
those
hacks
very
pleased
for
new
compiling
I
think
this
is
government.
H
I
think
you
know
somewhere
in
the
Restrepo,
we
use
it,
which
is
sort
of
one
of
those
weird
things
about
the
Restrepo
and
so
having
this
exception,
where
you
know,
there's
a
lot
of
inline
maaan
statements
unless
they
contain
non
inline.
Mod
statements,
like
that
exception,
is
like
never
something
that's
going
to
come
up
in
practice
for
people,
and
so
this
is
in
effect,
letting
them
apply
attributes
to
inline
mile
statements,
while.
A
H
F
A
A
F
A
It's
most
efficient
right.
Well,
the
point
is
we
could.
If
we
were
thoroughly
committed,
we
could
stabilize
it
such
that
you
have
no
API
to
do
lazy
loading,
but
you
get
module
tokens
and
all
you
can
do
is
reproduce
them
right
now
and
we
can
run
with
that,
but
yeah
I
don't
know.
It
seems
like
we're
spending
a
lot
of
time
on
this,
but
not
that
it's
not
important.
E
A
A
Don't
think
we
put
it
in
the
rest,
reference
just
yeah,
just
the
rules
that
you
can
apply
it
for
one
thing,
I
think
the
rule
of
to
me
personally
saying
you
can
apply
it
to
inlet.
You
cannot
apply
procedural
macros
to
out
of
line
modules
and
that
it's
like
transitive
if
they
is
also
true,
if
they
appear
inside
of
something
that
you
applied,
the
procedural
macro
to.
A
It
also
applies
to
like
our
line
modules
functions,
or
maybe
we
finally
got
rid
of
those
we
just
do
so
like
that
seems
fine
to
me.
I
would
sort
of
expect
and
that
sort
of
unified
to
all
the
rules.
H
Yeah
I
think
it's
fine
I
would
like
to
like
it'd
be
ideal.
If
we
had
I
could
feel
better
feeling,
like
a
group
that
was
organized
around
working
here
on
this
letter
check
out,
probably
like
being
in
the
lead
of
it
right.
That
was
I
know
what
the
band
are
like.
What
steps
in
the
future
we're
going
to
be
taking,
so
we
don't
feel
you
can
feel
confident
with
just
any
left
in
the
state
forever.
A
F
F
A
G
G
G
F
D
A
So
exciting
part
is
it's
three:
it's
30
minutes
in
and
we
have
a
little
time,
I
put
down
three
things
that
would
we
could
spend
some
of
this
time
on
for
turbo
fish.
I,
don't
want
to
say
a
lot
except
to
say
that
I
was
rethinking
after
our
discussion
last
time.
I
know
I,
agree
with
kind
of
everything
that
we
said,
except
that
I
said
something
like
I'd
be
game
to
take
on
that
thread,
and
then
I
was
thinking.
Oh
that's,
true.
A
I,
you
know
I,
think
I,
think
the
approach,
if
we
were
gonna
go
forward
is
I
would
want
to
go
in
and
get
a
good
understanding.
Everyone's
concerns
write
them
all
out
in
whatever,
but
I'm
just
it
seems
like
a
lot
of
energy
and
I
know
how
the
priority
lines
up
so
I,
don't
know
what
exactly
I
think
my
preference
here
is
I
think
we
can
wheel
have
to
region
the
media
discussion
decision
today.
A
F
F
A
H
A
I
think
that's
the
only
distinction
I
would
draw
is
that
in
that
instance,
cuz
I
kind
of
agreed
with
I
agree
with
your
assessment,
but
I
also
felt
I
would
add
the
one
other
bullet
point
of
like
that
was
sort
of
a
detail.
I
didn't
expect
many
end
users
to
encounter,
whereas
I
think
there
will
fish
is
something
that
the
only
reason
that
I
would
nudge
it
any
higher
is
that
ordinary
rust
users
do
encounter
it
under
alien
places,
but
it
still
feels
like
it's
not
I.
I
E
A
True
that
many
of
the
use
cases
fall
into
that
category.
Let's
talk
a
little
bit
about
roadmap.
Maybe
well
I
mean
do
we
want
to
reach
a
firm
conclusion
here?
Maybe
we
do
so.
G
I
G
A
F
Didn't
use
entering
so
last
time.
I
spoke
correct
me
if
I'm
wrong,
no
I
mean
he
said
something
right.
So
the
idea
was
that
we
would
do
some
experimentation,
almighty
figure
out
the
events,
unlike
text
editors
and
other
things,
and
maybe
after
hear
something
really
worked
like.
If
there
were
no
problems,
we
would
go
ahead
and
stabilize
them,
but
but
let
you
thought
that
you.
C
So
I'll
try
and
say
what
I
was
saying
again
in
case
I
can
do
a
better
job
if
compiler
wants
to
use
a
g,
lr
parser
or
something
like
that.
That's
totally
up
to
them
to
do
whatever
they
feel
is
best
I'm
happy
for
them
to
do
whatever
I'm
worried
that
this
is
one
of
those
cases
that
is
like
a
philosophical
feeling
more
than
something
that
would
be
really
just
a
mental
hey.
F
C
Think
I
have
it
at
least
somewhat
the
one
of
the
things
that
I
enjoyed
finding
about
rust
when
I
got
when
I
found
it.
The
first
time
was
not
having
like
Scott
Myers,
most
vexing
parse
thing
of
of
something
that
backtracks.
That
almost
always
does
what
you
want,
except
there's
this
one
particular
corner
over
here.
Where
sometimes
you
try
to
do
something
and
it
just
doesn't
do
the
right
thing.
If
you
know
what
I
mean.
A
Turbo
or
not
turbo
in
this
moment
but
I
mean
I
think
this
is
a
interesting
discussion,
but.
A
We're
really
trying
to
figure
out
what
to
do
with
this
thread.
Right
and
I.
Think
my
point
is:
if
I
were
gonna
go
forward
with
it,
I
would
want
to
I,
probably
close
and
reopen,
and
regardless
just
cuz
I
having
to
click
view
more
yeah,
nothing
else,
but
and
because
I
don't
want
people
to
talk.
G
D
E
Agree
with
that,
if
we
can
come
up
with
a
plan
in
the
next
like
week
or
two
that
amounts
to
here's,
why
we
feel
comfortable
reopening
it
and
resolving
it
then
I'm
all
for
it.
If
we
can't
then
I
would
be
all
for
just
you
know:
RFC
Bob
postponed
check
the
boxes
and
then
say
we'd
love
to
revisit
this
in
the
future.
When
we
have.
F
E
Like
I
actually
think
that
would
fit
well
with
insert
obvious
segue
here
with
the
roadmap
plans
that
Nico
and
I
are
looking
at
I.
Think
that
the
determining
factor
for
whether
we
want
to
postpone
this
or
whether
we
want
to
reopen
it
would
be.
Is
this
on
anyone's
roadmap
to
actually
work
on?
So
if
nobody
wants
to
spend
the
time
to
Shepherd
that
proposal,
then
we
should
just
postpone
it.
E
A
A
F
E
Like
to
propose
a
slight
topic
partition
here,
which
is
I,
feel
like
our
biggest
problem
as
the
Lange
team
is
not
well
specified,
features
that
have
been
accepted,
but
our
backlogged
on
implementation.
That's
a
problem
that
exists,
I,
agree,
but
that's
a
problem
where
that's
more
a
matter
of
developer
time
than
it
is
design
time,
and
there
are
a
lot
more
developers
than
there
are
designers.
So
my
bigger
concern
would
be
backlog
of
ideas
that
are
somewhere
between.
We
probably
care
about
this
and
actual
concrete
design,
RFC
and
acceptance.
E
E
F
F
A
A
F
We'll
review
the
post
this
number
of
parts
right,
so
it
says
I
can't
even
find
it
right
now
there
we
go
then
102
days,
so
I
think
I.
Think.
A
A
Okay,
we
have
ten
minutes.
I
had
another
bullet
point,
though
I'm
not
sure
if
we
were
done
with
this
discussion,
but
we
could
try
it
a
little
bit
more
to
the
like.
I
wanted
to
talk
a
little
bit
about
the
just
general
procedures
and
how
we're
managing
our
seeds
and
so
forth,
and
we
were
kind
of
getting
into
that
a
little
neighborhood
continue
with
that.
A
E
C
Might
phrase
that
slightly
differently,
as
we
have
a
lot
of
things
that
keep
coming
up,
that
don't
quite
turn
into
our
SES
and
aren't
necessarily
things
that
are
ready
to
get
approved
or
that
we
were
happy
to
approved
or
something
like
that.
And
then
we
have
a
whole
backlog
of
RFC's.
That
people
wrote
that
I
think
we
might
all
agree
are
not
things
that
we
want,
but
we
don't
go
through
and
do
anything
about
them,
because
we
have
an
awful
lot
of
RFC's
open,
for
example,
right.
A
A
Like
I
think
we
shouldn't
have
I
think
we
should
I
think
the
time
from
design
to
done
is
much
too
long
for
us
right
now
and
I
think
it
causes
a
lot
of
stress
in
different
grades
and
I
think
that
doesn't
necessarily
imply
that
we
should
go
for
multiple
things
that
were
RF
seed
and
not
implemented,
but
it
does
imply
that
we
should
try
to
not
create
new
ones
that
last
a
long
time,
and
probably
at
least
some
of
them.
You
should
be
closing
out
because
they
are
important.
Like
specialization
comes
to
mind.
A
E
A
H
Think
well,
there
are
two
issues
on
the
Nico
identified:
yeah
we
just
aren't
really
like.
There
are
several
features
that
we've
wanted,
for
you
know
three
or
four
years
now
and
haven't
don't
even
have
a
nightly
yet
and
then
I
think
just
that
the
RFC
process,
because
it
can't
handle
large
threads,
has
just
gotten
like
broken
in
a
lot
of
ways,
and
so
we
started
routing
around
the
RFC
process
instead
of
using
it,
and
so
that's
why
we
never
end
up
reviewing
RFC's
in
this
meeting.
G
C
A
H
And
I've
actually
I've
seen
several
features
were
like
there
are
several
proposals
and
I'm
necessarily
blamed
but
like
on
different
issues
where
I
thought
that
the
we
just
needed
to
have
an
artsy,
because
I
didn't
think
there
was
enough
evidence
of
consensus
right
and
the
problem
is
the
RFC
process
because
they
can't
handle
when
a
threat
gets
out
of
hand.
We've
all
sorted
like
avoiding
me
in
at
different
levels,
I
think.
C
Yeah,
we
sort
of
have
this
strange
thing,
I
feel
where
we
said
in
the
roadmap.
We're
gonna
finish
things
which
sort
of
implies
that
we're
not
gonna
have
our
FCS,
which
sort
of
results
and
people
phrasing
things
as
well.
I,
just
sort
of
cleaned
this
up
and
then
there's
an
issue
or
a
PR,
and
then
that's
there
and
things
just
sort
of
happen
in
random
places,
and
we
do
have
some
big
projects
that
are
doing
a
lot
more
coordination
like
people
we
have
here
and
that
currently
sort
of
thing.
C
A
So
we're
almost
out
of
time
for
the
meeting
I'm
really
enjoying
this
conversation,
I'm
wanting
to
have
something
like
this
for
a
while
I
want
to
propose.
We
should
probably
try
to
discuss
this
too.
I
have
some
very
specific
ideas
that
I
didn't
get
on
time
to
bring
them,
but
I'm,
tired
of
blocking
on
me
having
time
to
write
them
up,
which
never
happens
so
do
we
should
we
try
to
have
a
focused
discussion
on
this?
What
like
on,
how
to
run
ourselves
a
little
better
as
well
so
that'd
be
part
of
roadmap.
E
D
E
A
Okay,
I
think
I
agree
that
it's
worth
loggin
all
I
think
the
only
way
that's
not
is
that
it
might
give
us
vocabulary
with
which
we
use
to
form
the
road
like
talk
about
the
road
map
like
if
we're
talking
about
shepherded
things
for
people,
it's
just
a
different
way
of
looking
at
stuff,
but
it's
okay,
okay,
guys
start
somewhere.
But
let's
try
to
have
those
conversations.
I
guess.