►
From YouTube: Node.js Tooling Group Meeting 2021-01-22
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
I'll
be
live
soon.
All
right,
I
see
the
little
thing
all
right,
hi
everyone
welcome
to
the
node.js
tooling
meeting
for
january
22nd.
A
B
You
know
on
that.
On
that
note
sorry
to
interrupt,
I
I
would
love
to
get
like
some.
You
know
some
tooling
in
and
I
know
that
what
is
it
pkg
js
has
been
working
on
some
like
hack,
md
integration,
sorry
about
the
noise
that
would
be.
A
B
Okay,
I'll
I'll
ping,
him
again,
I
think
I
pinged
him
before
the
end
of
the
year
about
it.
I
deal
with
this
at
the
cpc
and
try
to
help
with
the
community
committee,
and
I
think
I
even
had
a
pr
open
for
tooling
notes.
So
it's
like
drives
me
a
little
crazy.
A
Yeah
no,
I
agree
it
could
definitely
be
better.
I
also
it
also
kind
of
annoys
me
that
the
agenda
issue
gets
created
a
week
in
advance
of
the
meeting
or
like
five
days,
I'm
a
very
last
minute
kind
of
person.
So
usually
it
would
be
like
adding
things
even
today
so
yeah.
Maybe
we
can
speak.
So
it's
like
like
the
day
before
or
something.
B
I
I
was
thinking
about
that
the
other
day,
because
the
cpc
meetings
get
created
on
a
monday,
which
is
good
because
we
don't
meet
on
a
tuesday,
and
I
always
thought
it
was
just
like
the
day
before.
But
then
I
noticed
the
comcom
meeting
or
something
was
also
it
may
be
like
on
monday.
It
just
creates
all
the
meetings.
I
don't
know
right.
We
can
maybe
look
into
that
or
maybe
just
figure
out
this
other
tooling
solution
work
from
there,
yeah
anyways.
A
Okay,
I
guess
the
first
thing
on
the
agenda
is
the
recursive
copy
issue.
I
I
did
yeah
open
a
proper
issue
in
the
tooling
repo,
but
I
linked
up
the
rfc
that
I
put
together.
Did
anyone
have
a
chance
to
look
at
that?
A
C
A
I
guess
that
was
I
didn't
actually
have
like
like
here
is
the
proposed
solution
that
I
I
want.
I
do
think
we
should
vendor
something.
I
looked
at
a
couple
of
the
existing
packages
and
it's
like
much
like
rimroaf.
It's
like
way
more
complicated
than
you
would
think
it
is
so
probably
makes
sense
to
use
something.
That's
from
the
community.
That's
like
you
know,
been
used
heavily
already,
but
there
are
a
couple
options
for
that
too.
A
So
yeah
we
would
need
to
pick
one
and
it
might
involve
some
discussions
with
people
that
work
on
those
projects
as
well.
C
A
A
C
A
B
I
mean
I,
I
don't
think
we
we've
made
a
lot
of
progress
from
your
draft
rfc,
so
maybe
we
just
kind
of
invite
them
in
now
and
start
to
kind
of
talk
more
about
it
or
you
think
it's
still
too
premature.
A
B
C
C
A
Okay,
yeah:
I
can
try
and
reach
out
to
them.
A
A
Okay
yeah,
I
guess
that
seems
like
a
good
next
step
and
then,
after
that
maybe
open
the
issue
in
the
main
repo.
A
That
sounds
good.
Okay,
does
anyone
have
anything
else
they
want
to
mention
there.
A
Nope
all
right,
let's
see
we
have
foreign
function
interface
still
on
the
list.
I
know
basically
nothing
about
that.
I
was,
I
think,
brian
english.
I
think
we
had
a.
We
had
talked
last
time
about
trying
to
get
some
of
those
people
back
to
the
meetings.
C
C
C
A
Yeah,
because
there's
at
least
one
thing
on
here
that
he
was
sort
of
most
interested
in
okay,
so
we'll
skip
foreign
function,
interface,
chmod,
dash
r.
We
have
on
here
also.
D
No
update,
I
think
I
was
the
best
to
touch
this
and
I,
I
think
also
I
sort
of
punted
on
this
after,
like
some
of
the
recursive
stuff,
that
you've
been
doing
like
sort
of
kicked
back
off.
So
I
haven't
had
time
for
sure
to
pick
this
up.
So
I'm
not
sure
if
it
makes
sense
to
get
somebody
else
involved.
If
we
want
to
actually
move
forward
with
this.
A
We
could
potentially
just
like
back
burner
it
for
now
leave
the
issue
open,
would
take
it
off
the
agenda
and
yeah.
D
A
Okay,
next
on
the
list
is
esm
module
reloading.
I
saw
some
activity
around
this
in
one
of
the
various
github
repos
recently.
Does
anyone
know
what
I'm
talking.
A
I
there
there
has
been
some
activity
around
this
topic
in
one
of
the
maybe
like
in
the
not
the
modules
group.
Another
group
I've
been
getting
a
lot
of
github
notifications
related
to
it,
so
it
seems
like
it's.
You
know
a
topic
of
discussion
or
something
that's
on
people's
minds.
I
don't
remember
where
we
left
off
with
this
one
in
the
past
ben.
Do
you.
C
C
It
comes
down
to
that
thing
where
it's
like
a
problem.
People
have,
but
there
isn't
that
much
adoption
of
esm
yet
so
it
might
be
that
we're
slightly
ahead
of
the
curve,
and
it
would
be
good
for
us
to
recommend
something
or
maybe
there's
something
that's
already
getting
popular
in
the
community.
We
can
just
point
people
towards
if
the
question
comes
up,
so
I
mean
people
are
going
to
want
to
keep
doing
proxy
choir
and
stuff
in
esm.
B
A
It
wasn't
this
issue.
No,
although
this
one
is,
you
know
relevant
for
that
that
wasn't
the
one
that
I
was
getting
all
the
notifications
for.
I
wish
I
had
a
better.
I
wish
I
was
better
prepared,
yeah
and,
of
course
it's
not
in
my
notifications
today.
A
Anyways,
that's,
I
guess
not
not
the
most
important
part
yeah.
I
guess
I
just
don't
really
know.
C
B
You
know
I
I
saw
that
in
one
of
the
issue
comments
yeah.
A
Okay,
support
for
hooking
spawn
spawn
sync
that
was,
that
was
what
corey
was
interested
in.
C
Sorry
about
that,
yes,
I
think
some
of
this
stuff
was
specific
to
nyc
and
how
nyc
would
hook
into
like
be
able
to
inject
an
environment
variable
when
spawning
something
which
so
that
you
kind
of
have
these
sticky
environment
variables
that
bubble
down
to
the
sub
processes.
C
Let
me
find
the
actual
agenda
item
yeah
yeah,
that's
the
hooking
spawn
and
send
spawn
sync.
So
I
don't
know
I'd
say
that's
another
one
where
we'd
be
interested
to
see
where
cory
landed
on
this.
A
Yeah
we
can
just
skip
over
it
for
today
and
if
back
to
one
of
the
meetings
we
can
discuss
that.
C
C
A
All
right
and
then
yeah,
I
think
the
next
one
was
kind
of
related,
better
way
to
detect.
The
process
is
exiting.
C
C
I
believe
that
you
process.exit
will
get
no
event,
so
so
there's
not
actually
like
there's
not
a
great
way
to
detect
not
a
great
way
to
detect
exiting.
C
Yeah,
so
I
mean
we
use
the
signal
exit
was,
is
used
for
test
coverage
collection
so
that
you
can
kind
of
make
sure
to
dump
the
test
coverage
at
the
very
last
step,
but
it's
kind
of
silly
to
have
to
do
that.
It
would
be
nice,
it
would
be
nice
if
there
was
literally
just
like
a
on
really
going
to
exit
or
something
that
that,
like.
I
think
this
would
be
really
helpful
for
tooling
right.
I
will
go
so
far
as
saying
I
will
champion
that
agenda
item
all
right.
I
think
it'd
be
super
useful.
A
I
know
like
I
know
we
use
a
library
for
like
exit
hooks,
which
is
like,
I
think
again,
similar
kind
of
problem.
A
C
C
Kind
of
it's
definitely
an
insider
baseball
problem,
but
if
you're
like
creating
a
telemetry
library
like
coverage
or
diagnostic,
tooling
or
whatever,
and
you
want
to
make
sure
you
dump
something
at
the
last
minute,
it
would
be
nice
if
you
could
just
do
that
with
a
one
handler
and
not
have
to
install
a
library.
A
I
have
not
really
dug
into
how
it
works
in
detail,
because
you
can't
do
async
work
at
a
certain
point
in
the
shutdown
cycle,
so
I
don't
even
know
how
that
works.
A
C
A
C
Yeah,
I
don't
know
this-
is
it
feels
like
a
little
bit
of
a
wart
on
the
api
service
that
maybe
I
could
talk
to
joey
or
someone
who
knows
a
little
bit
more
about
the
system
code
than
me
and
see
if,
if
they
have
a
an
idea
of
a
better
design
because
snow
dumps
a
ton
of
diagnosis
stuff
at
the
last
minute,
like
we
don't
coverage,
we
don't
diagnostic
reports.
We
don't
performance
profiles,
so
so
you
know
there
obviously
is
the
ability
to
to
hook
into
the
last
moment.
It's
just.
C
We
haven't
exposed
it
to
users,
so
I
think
part
of
it
is
that
concern
that
you
could
get
node
in
a
state
where
it
never
exits
right
like
if,
if
you
were
to
do
something
like
an
async
exit,
there's
no
guarantee
that
that
async
behavior
terminates.
So
you
know,
there's
no
guarantee
that
the
process
actually
exits
which
people
are
concerned
about.
C
A
Okay,
next
on
the
agenda
is
source
map
v3
any
updates
there.
A
C
C
Yes,
I
I'm
gonna
get
rid
of
the
two
line.
Like
I
said
last
meeting,
I'm
gonna
get
rid
of
the
two
line:
error
reports
that
we're
doing
right
now
and
if
you
turn
source
maps
on
it's,
just
gonna,
look
identical
but
use
the
source
maps
and
that's
more
in
line
with
the
tc39
spec.
So
right.
C
A
All
right
sounds
good
yeah
up
next
is
argument.
Parsing.
I
saw
your
slack
message
joe
about
the
your
pr.
I've
not
had
a
chance
to
look
at
it
yet.
B
Yeah
I've
got
and
I
put
the
notes
in
the
doc
ready.
I've
got
a
pr
open
in
my
own
fork
of
node
against
my
own
master
to
just
kind
of
start
to
work
through
it.
Roy
had
volunteered
to
implement
the
bit
that
he
had
commented
on
in
the
original
pr.
B
D
B
No,
you
know,
for
my
own
knowledge,
some
of
the
stuff.
That's
in
the
pull
request
in
the
dock
wasn't
making
wasn't
totally
clear
to
me.
So
I
was
just
kind
of
researching
parsing
arguments
from
the
command
line
and
stuff,
and
so
all
I'm
really
trying
to
do
is
get
basically
a
new
pr
opened
with
the
same
work.
That
was
done
because
I
know
there's
a
lot
of
effort
put
into
it
beforehand
and
then
a
lot
of
work
that
chris
did
and
worked
through
with
the
group.
B
So
I
want
to
just
get
it
reopened
essentially
and
address
some
of
the
blocking
concerns
and
you
know
kind
of
hopefully
move
it
forward
from
there.
I
mean
it
seemed
like
it
was
really
close.
So
I'm
hoping
you
know
we
can
work
through
the
couple
of
bumps
in
the
road
and
then
and
then
get
it
wrapped.
C
I
I
think
chris
did
put
together
a
design
document
in
a
hackmd
pad
that
he
worked
through
with
me
and
roy
roy,
like
I
want
to
say
almost
a
year
ago,
so
we
can
probably
dig
that
up
and
darcy.
If
you
want
to
give
it
a
read,
I
feel
like
it
was
ended
up
being
a
pretty
good
compromise
between
a
bunch
of
research
like
it
was
like.
Chris,
looked
through
every
parser
that
was
around
and-
and
it
was
pretty
well
thought
out.
D
Yeah
yeah
I've
actually
read
it
quite
a
bit.
Sorry
joe
it's
just.
I
guess
my
thought
here
is
just
to
ensure
that
there's
no
more
like
there's
a
buy-in
from
the
people
that
had
blockers
first
on
the
approach
before
spending
time
coding
and
then
opening
up
a
pr
and
sort
of
having
the
back
and
forth.
It's
like
a
little
bit
easier
to
go
back
into
the
and
get
consensus
on
design.
B
You
know
I
I
can
talk
to
the
original
blockers
as
well
as
this
group
about
any
implementation
and
make
sure
that
it's
the
right
move,
yeah
yeah
I'll
I'll.
Definitely
do
it
diplomatically.
D
There's
also
a
raver
stream.
I
read
through
this
the
other
night
and
spent
a
lot
of
time
also
doing
research
in
here.
I
think
there
was
also
like
feedback
about
whether
or
not
return
value
could
be
a
string
or
array
and
then
also
the
casting
was
also
a
big
big
problem.
D
As
far
as
the
pr
I
remember
was
also
a
big
like
point
of
contention
in
terms
of
I've,
also
like
kind
of
stuck
in
here
a
bit
but
yeah,
I
can
probably
poke
you
async
joe,
with
some
thoughts
I
had
from
from
just
like
reading
back
through
that
history,
because
I
just
want
to
yeah.
B
Yeah
I'd
be
happy
to
you
know
if
you
want
to
meet
if
you
want
a
pair.
If
you
want
to
just
comment
in
the
doc
the
hack
md
doc,
I've
got
which
doesn't
have
a
lot
into
it.
C
My
experience
maintaining
yards
for
the
last
like
seven
years
now
or
six
years,
has
been
like
no
matter
what
decision
you
make.
It's
there's
gonna
be
an
opposite
opinion,
because
it's
such
an
opinionated
thing.
So
so
I
think
we
should
just
be
careful
that
we
don't
let
perfect
be
the
enemy.
The
the
good
okay,
because,
like
there's
going
to
be
a
dissenting
like,
I
think,
an
argument.
Parser
would
be
good
in
core.
I
think
it
would
help
tutorials.
I
think
it
would
help
people
who
are
learning
how
to
code.
B
Yeah
yeah,
I
I
I
I
will
you
know
likely
defer
to
this
group.
I
don't
really
have
like
a
horse
in
the
race,
I'm
just
you
know,
kind
of
trying
to
just
help,
get
it
over
the
finish
line.
So
I'll
definitely
lean
on
on
you
all
to
help
figure.
Some
of
these
things
out.
C
I
think
it
will
open
up
some
I'm
kind
of
excited
about
what
it
opens
up
about,
like
maybe
exposing
some
parts
of
the
node
code
base
as
command
line
helpers
too.
So
you
could
just
more
easily
rim
rafter
directory,
because
we
allow
ram
rafter
to
take
arguments
and
like
without
even
pulling
in
a
dependency.
You
just
have
it
as
a
npm
script
that
runs
node
with
like
a
rim
raft
flag
or
something
like
I
don't
know.
I
think
I
think
we'll
open
up
a
lot
of
cool
stuff.
I'm
just
I'm
excited
cool
yeah.
B
C
E
D
C
I
definitely
found
it
like
yeah.
I
agree
right
and-
and
I'm
thinking
like,
because
I'm
kind
of
we
actually
don't
have
that
many
node
experts
in
my
part,
at
google
and
like
when
people
edit
our
samples,
they
have
a
lot
of
trouble
understanding
what
yards
is
doing
for
them,
because
we
used
to
have
yards
as
a
dependency,
and
so
it
can
be
confusing
because
they're
like
well.
Why
can't
I
just
pass
in
an
array
of
things
and
like?
Well,
that's
not
how
process.v
works,
it's
a
bunch
of
strings
and
I
think
it's
definitely
been
a.
B
Well,
yeah
and
chris,
you
know,
put
a
lot
into
the
description,
so
there's
definitely
a
lot
there,
but
I
don't
see
a
link
to
the
doc
that
y'all
are
talking
about.
I'm
looking
through
the
the
argument.
Parsing
issue
number
19
in
tooling,
but.
A
Okay,
anyone
have
any
anything
to
add
on
that.
One.
D
I'll
I'll
follow
up
with
joe,
I
think
I
had
like.
I
literally
did
a
bunch
of
splunking
and
and
going
back
in
the
last
year
and
a
half
myself,
and
I
really
appreciate
that
joe's
picked
up
this
work
and-
and
I
want
to
make
sure
that
it
gets
through.
I
know-
ben
roy
and
chris
had
worked
quite
a
bit
in
that
first
phase
and
never
never
added
like
my
voice
into
the
the
pr
when
it
was
there
and
yeah.
A
B
I
I
I
perhaps
before
we
get
to
that
ben
responded
and
commented
in
the
issue.
Let
me
see
if
I
slide
in
my
clipboard
here,
is
it
this
one
dang
it?
Where
is
it.
B
Yep,
so
it's
short,
he
just
says
just
a
heads
up.
This
is
still
something
I
want
to
pursue,
but
don't
have
a
whole
lot
of
time.
Right
now,
number
two.
It
should
probably
be
removed
from
the
bi-weekly
agenda
unless
someone
else
wants
to
tackle
it.
Tangentially,
although
related
to
alternative
section
above
I've,
put
together
a
helper
library,
fast
native
fn,
to
help
folks
use
v8
fast
api
calls.h
when
available
and
still
compile
when
it
isn't.
A
C
Corey
just
responded
with
essentially
the
same,
he
said,
feel
free
to
just
mark
them
as
stale
or
stale
or
kind
of
not
being
worked
on.
A
Okay,
cool,
okay,
yeah,
any
new
stuff.
C
Well,
we
talked
last
week
about
we
talked
two
weeks
ago
about
trying
to
put
together,
like
a
wider
community
meeting,
to
try
to
get
feedback
on
whether
we're
working
on
the
right
stuff
for
the
year
trying
to
put
together
a
like
a
kanban
board
or
something
we
have
not
done
that.
Yet
I
think
we
have
some
great
stuff
here,
that's
shared
by
a
few
folks
right
now
like
I
feel
like
we
could
keep
busy
for
quite
a
few
months
on
this,
but
it
might
be
interesting.
C
A
Yeah,
I
agree,
I
think
we
should
be
kind
of
presenting
our
own
idea
or
roadmap
and
then
yeah
kind
of
opening
that
up
to
feedback
versus
just
asking
people
what
they
want
to
see.
C
Might
also
be
interesting
to
just
like
kind
of
codify
codify
how
we
decide
what
to
work
on
too,
like
you
know,
we
try
to
pick
things
that
we
know
have
some
major
stakeholders
who
would
love
to
have
them
and,
like
we
haven't
done
things
like
fs
in
a
holistic
way,
we've
chosen
things
that
are
really
impactful
that
people
have
requested.
So
I
don't
know,
maybe
just
as
we
think
about
what
we're
working
on
this
year,
maybe
think
about
how
to
describe
it
to
the
community.
A
C
So,
sir,
it's
hard
to
tell
without
my
video
on
that.
I'm
talking
I
apologize,
I
was
gonna
say
so
so
maybe
try
to
put
a
stake
in
the
ground
as
to
when
we
want
to
get
people
in
for
a
slightly
bigger
meeting.
A
Like
yeah,
I
think
when
we
talked
last
time
just
looking
at
my
calendar
here
two
weeks
ago,
we
said
we
were
going
to
do
it
in
a
month,
but
given
that
we
haven't
put
the
road
map
together,
I
think
we
should
like
defer
that
another
couple
weeks,
because
I
think
the
idea
was
we
wanted
to
put
the
the
road
map
out
there
like
open
up
an
issue,
make
sure
people
know
about
it
in
advance,
have
time
to
contribute
their
suggestions,
etc.
C
I
think
well,
I
won't
be
talking
much,
but
that
could
be
fine,
like
I'll
have
a
few
days
to
get
feeling
better.
A
I
think,
because
I
think
doing
it
like
two
weeks
from
now
is
probably
not
enough
notice.
A
I
agree.
Okay,
so
we
could
do
I
mean
we
could
do
the
19th
or
we
could
do
march
5th.
Then
these
two
fridays.
After
that,
I
guess
we
can.
We
can
figure
that
out
maybe
a
good
exercise
over
the
next
like
between
now
and
the
next
meeting
is
like
create
that
board.
You
can
share
the
link
on
slack
and
add
our
our
road
map
suggestions
to
it.
A
A
C
A
One
actually
this
was
your
suggestion
ben
we
should
work
on
removing
the
deprecated
rimder
recursive
functionality,
because
the
longer
we
wait,
the
harder
that's
gonna,
become,
I
think,
even
in
the
tsc
meeting,
where
that
was
discussed.
Like
I
think,
michael
even
said
it's
possible,
this
never
actually
gets
removed
because
it
is
kind
of
being
used
by
people
already.
So
I
think
that's
a
good
that
should
be
relatively
easy.
I
think.
A
I
actually
have
the
node
releases
page
bookmarked,
because
I
refer
to
it
like
at
least
once
a
week.
Yeah
note
16
looks
like
april-ish
of
this
year,
so
that
gives
us
a
couple
months.
I
think,
should
be
an
easy
code
change
right.
It's
just
removing
the
code.
That's
deprecated!
It's
already
deprecated
in
the
docs.
A
So
yeah
anyways,
that's
one
thing
that
I
think
we
should
do
and
I
I
mean
I
can
take
that
on
and
then
the
other
thing
I'm
afraid
to
bring
up
that.
I
wish
we
were
working
on
is
fetch.
Are
we
ever
going
to
get
fetch
or
something
like
that
in
node?
A
F
Ethan
arrowood
has
been
working
on
on
a
fetch
fetch
implementation
on
top
of
undici,
and
I
feel
like
that.
That's
probably
one
of
the
least
controversial
ways
to
do
it
since
undiichi
will,
I,
I
think,
theoretically
be
going
into
note
eventually
so
like
having
our
own
thing
built
on
top
of
indici
makes
sense.
F
I
think
I
I
could
be
wrong.
I
think
the
intent
is
to
eventually
put
it
into
node.
I
could
be
very
wrong
on
that,
but
implementing
it
in
in
something
in
node
makes
sense.
A
Yeah,
I
I
just
you
know
ben
you
mentioned
like
with
arg
parsing,
you
know
being
able
to
to
work
without
libraries
there,
like
I,
I
was
writing
a
lambda
the
other
day
I
needed
to
to
hit
a
url
and
didn't
really
want
to
set
up
a
whole
pipeline
and
dependencies
and
all
that,
and
it
was
a
big
pain.
C
A
C
A
F
I
was
scrambling
for
for
unmute.
I
would.
I
would
definitely
talk
to
ethan
airwood
into
mateo,
and
I
I
think,
there's
ronag,
I
think,
is
also
working
on
it.
Getting
the
landscape
there
and
what's
happening
and
figuring
out
moving
on
from
there.
I
think
it's
probably
the
best.
A
Bet:
okay,
I'm
just
adding
that
in
the
notes.
C
A
Sounds
good
adding
to
the
notes
here
too
deprecate.
E
Oh
last
thing:
I've
been
working
on
for
the
last
weekend
have
been
working
on
npm,
diff
kind
of
revived
that
idea
from
last
year.
So
I'm
wrapping
up
work
on
that
and
it
should
probably
land
next
week.
So
especially
folks
from
this
group
are
going
to
be
very
interested
on
it.
So
yeah,
let's
keep
an
eye.
C
E
D
There's
this
great,
like
workflow,
that
I've
seen
on
the
marketplace
that
was
written
by
this
eco
character,
but
yeah.
D
I
just
I
created
just
a
point
of
order.
I
guess
I
created
the
retro
board
if
you'd
like
to
add
tickets
for
the
existing
roadmap,
that
we
have.
We
can
circle
back
on
on
this
board
and-
and
you
know,
expose
it
to
other
folks
if
we
think
we
want
to
get
their
feedback
and
we
can
use
this
in
the
future.
Larger
group
meeting.
A
Yeah,
that's
awesome
thanks
for
doing
that,
yeah,
I
think
that's
an
action
for
all
of
us
like
over
the
next
like
a
week
or
something
let's
try
and
get
our
stuff
on
there
and
then
yeah.
Maybe
we
open
up
an
issue
for
when
we
want
to
have
the
meeting
and
link
to
the
board.
A
All
right
all
right!
Well,
thanks
for
coming,
everyone
have
a
good
weekend.