►
Description
B
A
B
Right
welcome
everybody
to
the
node,
GGC
meeting
or
march.
Eight
can't
believe
it's
already
march
eight.
You
should
have
a
pretty
light
agenda
today.
We
don't
have
a
lot
of
folks
on
the
coal,
let's
go
ahead
and
do
a
stand-up.
First,
let's
leave
a
son
who
is
on
the
list
here.
I
think
Bradley.
You
are
first.
C
So
I'm
working
on
porting
and
fixing
the
loader
implementation
into
c,
plus
plus
so
people
can't
muck
with
it
and
working
with
some
people
on
exact
hooks
that
need
to
be
available
to
JavaScript.
That's
it.
Next
up
has
been.
E
A
G
Were
several
problems
in
the
a.m?
Module?
It's
the
sandbox
behavior,
that's
used
a
lot
and
just
enjoy
a
storm,
and
we
always
had
function
because
we
weren't
able
to
intercept
things
like
define
property
and
she
was
changing
or
still
is
changing
the
implementation
with
a
new
of
the
eight
API
so
that
we
can
intercept
all
of
the
definitions
and
setters
and
gathers
and
get
rid
of
this
copy
properties.
Hack,
okay,.
H
I
B
A
B
Other
sleep
right,
Vincent
right
now
is
anybody.
B
Right,
we
have
a
pretty
light:
the
gender.
What
last
meeting
we
talked
about
the
async
loader
for
module
stuff?
We
talked
about
Bradley's
a
sketchy
as
heck
loader
implementation
that
we
could
discuss
some
that
still
made
your
work
in
progress.
We
talked
about
updating
the
copyright
which
we
will
be
talking
about
again.
B
B
B
B
G
B
B
I
H
A
B
A
See
simple
properties
if
you're
like
using
inspector
an
actual
debugger
I'm
trying
to
think
like
when
you'd
want
to
see
these,
and
one
would
one
would
probably
be
ok,
not
necessarily
looking
for
things,
but
if
you're,
like
you
know
trying
to
debug,
why
maybe
I
know
a
future.
A
cheapy
API
is
working
a
certain
way
and
there's
a
bunch
of
things
hidden
by
singles
and
you're
wondering
ok,
what's
necessarily
the
state
in
this
certain
situation
like
most
of
us
console.log
for
debugging,
let's,
let's
be
real.
I
B
A
chrome
exposes
it,
I
think
the
Firefox
does
too,
when
you
do
console.log
an
object
with
the
symbol,
see
the
symbol.
Value
hey,
like
I,
said:
I'm,
not
going
it's
more
of
a
personal
preference
thing
for
me,
so
it's
nothing
that
I'm
going
to
and
in
a
way
and
I
think
you're
you're
right.
Most
people
will
be
the
console
log
you're
going
to
want
to
see
that
state
I
would
just
like
to
make
sure
that
we
still
have
a
way
of
turning
it
off
or
making.
B
B
B
B
F
B
B
B
B
A
Not
exactly
true,
oh
yeah,
p,
minus
1,
the
paver
is
intentional
just
for
the
sake
of
consistency,
and
that
would
be
fun
non.
Innumerable
string
properties
are,
we
sure
browser
show
the
month.
Cuz
I
would
be
weird
I
just
checked,
I
just
checked,
chrome,
definitely
chosen
and
I,
don't
know,
I
think
I
mean
probably
gonna
head.
A
C
So,
just
a
heads
up,
innumerable
properties
generally
are
considered
to
be
the
way
to
go
for
class
methods
and
things
things
that
are
not
associated
directly
with
an
object
itself
so
for
in
historically
we'll
skip
all
the
dom
methods
because
they're
non
innumerable,
but
they
will
show,
if
you
add
any
custom
properties
to
a
Dom
object.
So
that's
generally,
what
they're
considered
useful.
B
A
You
sit
I
mean
our
symbol
properties
innumerable
by
default,
because
then
you
could
probably
just
use
object,
ought
to
find
property
to
properly
hide
a
symbol
value
if
you
really
want
it
to,
and
that
would
be
consistent.
I
think
I
mean
that
means
not
printing
them
all.
If
they're,
not
not
numerable,
that
make
sense.
C
B
I'm
looking
to
the
right
now,
okay,
yeah
they're,
not
innumerable
so
so
edge
I'm
hides
them
by
default.
Firefox
hides
them
by
default,
though,.
B
B
B
B
Alright,
so
the
next
issue
is
17
for
updating
the
copyright.
This
one's
on
me
basically
same
status
where
we
were
last
week,
I
need
to
update
the
TR
and
then
get
it
landed.
I
was
going
to
do
that
last
week,
but
the
vm
summit
and
the
need
to
help
miles
out
with
exporting
stuff
to
forward
six
kinds
of
presidents,
so
same
status
on
that
and
no
update.
Yet
on
the
new
copyright,
we
still
have
not
heard
anything
from
the
foundation
they
lower.
B
B
C
So
Dominic
and
some
of
the
polymer
devs
are
concerned
about
path
searching,
so
that
means
file
extension
searching
and
searching
in
directories
for
index
or
package.json
when
you
give
import
a
relative
or
absolute
path.
So
when
we
were
discussing
esm,
loader
I
was
a
bit
careful
to
always
ensure
that
our
behavior
only
diverges
in
the
case
where
a
browser
would
produce
an
error
normally
so
the
what
WG
spec
is
compatible.
In
my
mind.
From
that
perspective,.
C
However,
they're
requesting
we
don't
do
past
searching
on
relative
or
absolute
pathnames,
we
only
do
it
when
you're
importing
between
packages,
it's
more
requests
so
that
they
can
move
all
the
errors
two
after
resolutions,
so
the
problem
for
them
is:
if
we
do
it
on
relative
or
absolute
pathnames,
that
means
we
actually
would
be
saying
for
the
browsers.
If
this
will
give
you
a
404,
then
you
need
to
do
path,
searching
which
they
don't
have
any
network
access
and
the
resolution
algorithm,
and
they
don't
really
want
to
add
hooks
for
that.
C
C
Sure
so,
if
I
do
require
dot,
slash
lib
right
now,
it'll
look
for
a
live
DJ
s,
it'll,
look
for
lib
/
index
jsf,
so
on
they're,
requesting
that
we
only
do
that
kind
of
searching
when
you
do
it
between
packages,
so
dot
slash
lib
would
only
look
at
dot,
slash
lib,
it
would
never
add
file
extensions,
it
would
never
look
for
indexes
or
a
package.
C
H
C
No,
we
don't
have
a
good
pan
for
packaging
core,
but
we
do
kind
of
imply
it
by
the
use
of
package.json
and
node
modules.
Look
up
paths
so,
when
I'm
talking
about
package
I'm
talking
about
a
directory
with
a
package.json
that
is
looked
up
using
a
bear
specifier
that
means
one
that
starts
generally
with
an
alphanumeric
or
some
character.
That
is
not
a.
B
A
It's
especially
into
other
packages,
so
if
you
want
to
require,
say
lodash,
/
defaults
or
something
then
in
low
dashes
directory
so
long
it
like
they
can
kind
of,
they
have
some
some
room
to
like
play
around
with
with
that
it
could
be
default
cas.
It
could
be
default
in
a
folder
with
some
other
stuff
to
help
them
make
it
more
modular
now,
I'm,
understanding
that
that
still
works,
though,
is
that
correct?
Yes,.
A
A
C
C
C
So,
on
the
other
side
of
things,
the
argument
here
is
somewhat
that
people
don't
really
control
other
packages.
So
if
I'm
importing
lodash
generally,
I
don't
control
the
level
of
abstraction
that
low
dashes
and
so
past
searching
seems
okay
there.
But
for
relative
or
absolute
things,
I
should
have
control
over
the
path,
and
so
refactoring
should
be
somewhat
expected.
F
C
A
C
We
could
not
use
URLs,
it
would
make
things
very
odd
between
browsers
and
node,
so,
but
there,
the
attempt
to
keep
everything
within
a
compatible
set
of
air
conditions
means
how
we
handle
URLs
matches
browsers
more
than
it
matches
require
unless
you're
doing
something
really
strange,
you
shouldn't
see
any
compatibility
problems,
but
there
are
cases
where
there
are
definitely
visible
changes
if
you
are
doing
things
with
filenames
that
had
special
URL
characters
in
them,
for
example,
it'll
be
visible.
C
B
C
B
We
want
to
have
any
more
update
on
module
implementation
or
read,
as
it
is
pretty
much
everyone
needs
to
go.
You
know
policy
case
you
everything
to
go.
Take
a
look
at
the
implementation.
C
B
So
it's
going
to
depend
on
the
timing,
I'm
going
to
be
cutting
the
eight
branch
on
monday
to
start
the
process
of
debate
is
and
the
release
candidate
I
think
the
goal
is
to
get
five,
seven
and
but
I
think
I've.
Still
a
work
in
progress
and
I
think
I
wish.
Targus
was
here
to
collect
what
were
that
that,
but
we
can
touch
bases
on
that
next
week,
not
first
rc
on
Monday
it'll
be
the
first
beta
I'm
and
I'm.
B
D
So,
as
for
the
schedule
from
the
v8
side
I,
my
expectation
is
that
it
5.7
should
be
stabled
student,
so
I,
take
it.
I
mean
57.7
to
definitely
be
on
track
to
be.
In
noted,
the
question
really
comes
around
OBS
5.8,
which
has
some
epi
changes
more
specifically
the
old
debug
date
on
protocol,
so
that
no
longer
than
500
turn.
D
So
probably,
if
schedule
point
of
view,
while
it's
possible
for
5.8
to
get
into
no
date,
I
think
it's
going
to
be
important
for
CDC
to
actually
make
a
decision
on
which
one
we're
going
to
pick,
because
the
lack
of
the
JSON
debugging
protocol
will
mean
that
we
will
have
to
make
start
doing
some
work
now.
If
we
want
to
be
able
to
shift
by
put
again
and
I,
don't
think
it's
necessarily
a
good
thing
for
us
to
ship
584,
because
it
will
necessarily
require
us
to
remove
the
old
debug
support
forum
loaded.
A
Yeah,
that
does
also
put
us
in
a
bit
of
a
pickle
it
later.
If
we
were
to
sew
in
I
think
both
six
and
seven
we
did
in
or
maybe
just
in
four
and
six
we
did
in
stream
v8
bumps,
but
it
sounds
like
that
would
not
be
possible
if
we
went
5-7,
or
at
least
the
difference
would
be
very
large.
I,
don't
know
if
it'd
be
possible.
The.
D
So
the
question
really
becomes
I'd
have
so
should
it
have
the
old
give
a
portable
or
not
right?
So
so,
if
you
want
to
do
with
increased
somewhere
major
or
a
bi
compatible
bunt
bump,
then
we
like
to
make
sure
that
we
can
put
that
back
in
when
you
bring
in
fight
for
it.
So
I
think
that's
really
fundamentally
the
question
whether
or
not
we
do
a
PDF,
a
droid
bumps
and
whether
the
feasibility
that
depends
on
what
note
he
tends
to
have
from
an
epi
dresser.
H
So
we
just
deprecated
the
old
debug
stuff
a
couple
weeks
ago:
we're
not
really
giving
I,
don't
think
that
much
time
I
know.
The
vs
team
here
has
actually
asked
me
about
that,
like
they
they
depend
on
the
old
one
and
I'm
sure
there's
other
stuff
out
there
a
note
inspector
as
an
example,
although
of
course
you
could
use
chrome
devtools
now,
so
it
kind
of
seems
like
that.
H
H
H
D
A
A
You
I
would
say:
go
57.
B
A
H
D
D
D
B
Right
so
my
only
challenge
there
is
light
and
not
too
comfortable
tying
the
release
too
I
think
it's
not
going
to
be
stable
until
the
end
of
April.
We
need
to
Democrats
that's
the
week.
We
need
to
actually
cut
the
80
release
that
that
would
mean
that
there
would
be
no
chance
of
getting
it
into
the
getting
the
stable
in
the
release.
Candidate
would
be
doing
release
candidates
with
a
beta.
D
D
B
D
B
H
D
B
B
B
B
A
B
B
B
B
The
current
state
of
napi
in
is
experimental,
not
behind
a
config
flag
but
behind
a
runtime
flight,
and
this
would
be
to
get
it
out
in
the
hands
of
native
module
developer,
so
they
can
actually
start
playing
with
it,
but
as
experimental
and
with
a
process
warning
emitted.
When
you
actually
go
to
use
it
to
say:
hey,
don't
do
this
in
production,
so
they've
got
that
two
estates,
the
napi
step
to
a
state
where
it's
it's
functional,
it's
nowhere
near
complete,
but
it
can
be
used.
B
D
B
You,
the
folks
that
are
working
on
the
napi
said
their
goal
is
to
epi
are
within
the
next
week.
Two
weeks
this
has
said,
everyone
take
a
look
at
that
getting
hit
the
comments
in
I
mean,
what's
the
understanding
that
it
would
be
experimental
and
changes
are
expected,
that's
why
they
want
to
get
it
in,
so
they
can
start
the
feedback
loop
from
people
actually
using
it.
B
The
really
good
thing
is
that
the
changes
there
are
sember
minor,
the
actual
changes
to
core
itself
are
really
quite
minimal
and
introduce
no
breaking
changes
to
existing
code.
So
everything
there
that
is
used
is
all
opt-in,
so
there
should
be
no
impact
to
existing
users
by
putting
it
in
as
experimental.
A
I
haven't
seen
any
I
did
ask
a
few
minutes
ago
now
for
a
questions
in
the
live
chat.
However,
the
stream
started
with
a
stream
chat
off
for
some
reason,
I'm,
not
sure
who
toggled
that
so,
if
anyone
asks
I'm
earlier,
not
they
wouldn't
have
been
able
to.
Fortunately,
it
doesn't
look
like
there's
any
questions.
So,
okay.