►
From YouTube: Node.js Tooling Group Meeting 2021-03-19
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,
welcome
to
the
node
tooling,
meeting
yeah
march
19th,
so
yeah
we
have
just
a
short
agenda.
It
looks
like
today.
First
on
the
agenda
is
recursive
copy,
which
I
don't
have
any
update
on.
I've
not
worked
on
that
last
couple
weeks,
also
just
looking
at
the
issue
right
now
we
were
someone
had
pinged
one
of
the
other
maintainers
of
fs
extra
and
I
don't
see
any
any
follow-up
there,
so
we
might
have
to
reach
out
but
yeah.
I
don't
really
have
any
significant
updates
on
that.
One.
B
C
Yeah
we
can
talk
about
that
and
we
could
also
maybe
talk
about.
I
think
at
one
point
we
had
the
idea-
and
I
I
think,
we've
lost
it
on
the
agenda
to
share
a
bit
of
the
road
map
that
we
have
with
externally
with
like
community.
I
feel
like
we
had
an
issue
we
had
put
together,
maybe
even
like
a
fun
retro
board.
Even
I
think
at
some
point
to
like
yeah.
A
C
A
We
were,
we
were
talking
about
doing
like
a
a
meeting
to
kind
of
get
ideas
from
other
people
and
and
sort
of
put
that
together
into
a
road
map,
but
then
I
think
we
kind
of
decided
like
we
actually
have
some
some
decent
projects
on
the
go
right
now,
like
argument
parsing
and
the
recursive
copy
and
a
couple
of
those
things,
so
I
think
we
we
kind
of
just
decided,
like
hey,
we
actually
kind
of
have
a
roadmap
already
we're
making
good
progress
on
these
things.
Let's
just
work
on
them.
B
Just
dropped
a
link
to
the
time
zone
or
the
time
stam.
What
am
I
trying
to
say
the
time
of
the
meeting
issue
of
the
schedule?
Also,
I
was
going
to
add
the
agenda
label,
but
I
don't
I
don't:
have
access
I'm
not
able
to
not
really
yeah.
B
C
A
No,
I
can
oh
and
it's
okay,
I
don't
know
why
that
is,
but
all
right.
A
Oh
yeah,
maybe
yeah,
maybe
you
guys
aren't
members
of
the
team
like
officially
in
github
I
mean
okay,
yeah.
B
Oh,
I
even
requested
to
join
this
team.
I
see
a
cancel
pending
request
button.
I
don't
know
who
manages
this
oh
looks
like
chris
hiller
is
a
maintainer,
so
maybe
we
should
create
an
issue
to
sort
that
out
sort
out.
The
members
list
start.
A
A
Yeah,
okay,
good
good
I'll.
Add
that
to
the
notes.
C
We're
good,
okay,
the
other
thing
I
was
gonna
ask
and
joe
might
have
insight
on
this.
So
like
that
one
community
piece
that
I
was
wondering
where
we
landed
on
and
it's
good
to
hear.
Okay,
we've
got
enough
on
our
plate.
We
have
a
good
product
road
map,
I
guess
was
the
collab
summit.
C
If
there's
a
collab
summit
happening
with
openjs
world,
I'm
not
sure
if
there
is
one
and
and
that's
a
question
I
had
and
and
just
thinking
in
the
future
us
getting
together
to
to
you,
know,
block
out
time
to
to
do
work
or
to
to
talk
about
things.
I
know
we
have
these
sort
of
bi-weekly
calls,
but
you
know
it
tends
to
be
around
those
that
time
of
year,
when
we
have
that
conference
that
you
know
we
we
try
to
make
some
substantial.
B
That
is
a
great
question
and
I'm
a
part
of
the
openjs
program
committee
and
I
don't
think
we've
talked
about.
Unfortunately,
I
don't
think
we've
talked
about
collab
summit
stuff
at
all
and
it's
probably
something
that
should
be.
D
D
B
C
B
C
Or
last
week
that
I
I
was
noting
with
michael,
it
was
there
as
well
asking
about
it.
Yeah.
B
C
Yeah,
like
my
guests,
there
also
was
like
I
know
last
year,
which
was
the
first
fully
virtual
collab
summit.
I
know
that
some
of
the
feedback
was
that
it
would
you
know,
especially
somebody
that
ran
a
session.
It
was
definitely
tough
to
delineate
how
that
was
different
than
sort
of
what
we
do
here
each
week
and
how
to
make
it
special
in
some
way
and
try
to
block
out
time.
C
I
know
personally,
I
wasn't
able
to
really
be
as
focused
since
really
I
didn't
block
out
time
and
I
think
a
lot
of
other
folks
were
the
same
where
they
weren't
able
to
sort
of
like
disconnect
from
work
and
were
constantly
sort
of,
like
you
know,
not
really
100
focused
in
sort
of
the
sessions.
It
was
kind
of
hard
to
like
jump
in
and
out
of
rooms
and
and
so
anyway.
C
B
Yeah,
so
let
me
actually
notes
from
most
recent
cpc
meeting.
A
Yeah
I
like
that
idea
too,
especially
because
hopefully
the
the
stuff
that
we're
working
on
now
like
around
argument,
parsing
and
recursive
copy
and
some
of
the
other
stuff,
hopefully
we're
we're
landing
that
sometime
around
mid-year,
so
that
doesn't
seem
like
a
good
idea.
That's
maybe
a
more
appropriate
time
for
that.
Like
road
map,
you
know,
what's
next
kind
of
a
session.
B
So
I
added
the
notes
to
that
issue
that
was
created
in
the
summit
repo,
and
I
think
that
you
know
the
consensus
is
like
what
you
were
saying.
Darcy
so
I'll
read
it
here.
Online
clab
at
summit
online
collab
summit
was
not
as
useful
doing
it,
along
with
the
conference,
was
also
not
as
useful
since
it
was
online.
If
we
have
to
do
it
online,
delaying
it
would
be
preferred,
didn't
feel
much
different
from
working
groups.
Consensus
is,
let's
not
push
to
do
an
event
alongside
openjs
world
event.
B
D
B
There's
a
repo-
and
I
think
that
you
know
people
volunteered
in
the
past.
I
think
a
lot
of
the
people,
some
of
the
people
that
volunteered
in
the
past,
I
don't
think,
are
as
active,
but
some
of
them
are,
and
I
think
that
the
foundation
you
know
would
would
certainly
support
the
effort.
B
C
Yeah,
I
know
it
was
tough
to
organize
last
year,
just
time
zones
and
stuff,
because
it's
everybody
not
everybody's,
getting
in
the
same
place
so
yeah
managing
that
also
was
super
tough.
I
remember
trying
to
get
all
the
schedules
and
stuff
organized.
C
Okay,
well,
it's
good
to
know
like
I'm,
I'm
still
interested
in.
Maybe
us
then
doesn't
have
to
even
be
around
that
time,
but
setting
up
maybe
a
a
session
like
to
kick
some
of
this
stuff
off,
like
you
said,
maybe
if
we
have
it
blocked
out
as
a
meeting
and
there's
other
folks
like
our
everybody
on
it,
it
sounded
like.
I
missed
one
of
those
types
of
sessions
earlier
like
last
week.
I
think
you
all
got
together
to
to
talk
about
the
or
work
through
actually
backboard.
D
D
Yeah
yeah,
that's
yeah.
I
was
going
to
mention
that
one
like,
I
think,
maybe
just
try
to
schedule
this
out
of
band
meetings
like
the
the
one
from
last
week,
like
it's
very
helpful,
like
I
know
yeah
I
know
like
this
was
really
deep
dive,
very
technical
like
and
very
specific
niche
kind
of
thing,
but
I
think
that's
very
helpful
to
get
together
like
so
we
managed
to
land
one
of
the
advanced
pr.
Now
we're
hitting
a
roadblock
with
ci
we're
having
some
problems
there.
We
have
to
debug.
D
A
Yep
yep,
I
find
those
useful
in
general
like
when
I
was
finishing
up
the
most
recent
like
rimrap
related
work.
I
think
ben
and
I
just
got
on
a
call
like
a
saturday
or
something
for
a
couple
hours
and
just
worked
through
some
of
the
final
stuff.
I'm
just
quicker
to
do
it
that
way,
then
going
back
and
forth
and
yeah.
I
found
that
really
helpful.
It's
a
good
way
to
like,
I
know,
we're
all
busy
and
stuff
too
right.
A
C
To
yeah,
if
I
agree
to
an
event
or
if
I
agree
to
like
a
meeting
then
like
there's
a
hot
like
that's
the
highest
percent
chance
I'll,
actually
block
out
the
time
and
actually
be
there
and
yeah
that's
the
best
way
to,
I
think
to
yeah
yeah
get
work
done.
Unfortunately,.
A
And
what
do
you
mean?
Okay,
I
think,
let
me
add
those
couple
things
to
the
to
the
notes
anything
else
we
want
to
talk
about
while
we're
waiting
for.
B
Yeah
yeah
I'll
I'll
I'll
bring
up
something-
I
don't
know
you
know
just
to
gauge
interest
and
curiosity.
B
A
couple
of
conversations
have
come
together
between
tyranny,
myself,
ethan
arrowood
and
ben
halverson
related
to
docs
and
related
to
like
moving
kind
of
migrating,
the
docks
to
a
slightly
different
format,
I
think,
or
at
least
like
kind
of
how
they're
annotated
and
such
and
using
electrons
like
doc,
parser
to
oh
there's
been
to
make
the
docs
better
and
and
maybe
change
some
things
about
the
docs.
So
you
know
I
know
it's
sort
of
tooling
related,
but
sort
of
not
really.
B
But
I
thought
I
would
mention
it's
something
that
we've
started
talking
is
that
type
scripts?
I
know
typescript
definitions,
yeah
yeah,
it's
related
yeah.
B
Like
I
think,
being
able
to
parse
the
docs
and
pull
out
types
and
also
pull
out
like
the
annotations
and
be
able
to
build
the
docs
into
like
a
doc's
website,
you
know
or
sub
website,
whatever
part
of
a
website,
I'm
still
kind
of
getting
my
head
around
it.
I
think
this
is
related.
B
Let
me
click
on
it
too
and
see
yeah.
This
is
the
block
parser.
Oh,
maybe
this
isn't
too
related.
I
mean
in
terms
of
informing
the
conversation
anyway,
I
can
I'll
I'll
get
more
context,
and
I
can
mention
it
next
time
with
more
clarity,
but
I'm
not
even
sure
if
anybody'd
be
interested,
but
I
thought
I'd
mention
it.
A
A
B
Yeah,
I
think
the
the
the
thinking
is
that
this
is
more
code
from
docs
rather
than
like
jsdoc,
which
is
docs
from
code
and,
like,
I
think,
pulling
things
out
into
some
sort
of
ast
and
I
and
I
think,
even
like
generating
json
related
to
the
docs,
regardless
I'll
I'll,
get
more
up
to
speed
and
and
kind
of
bring
it
back
to.
The
group.
F
I
I
could
put
my
hand
up
I'd
rather
work
on
recruits
of
copy.
First,
if
you
were
thinking
you'd
like
someone
to
pair
with
on
that-
and
I.
A
Would
I
would
like
that
actually,
just
before
you
got
here,
we
were
talking
about
how
the
meeting
last
week,
where
you
guys
were
working
on
back
porting
the
source
map
stuff
was
was
useful
and
it's
like
a
maybe
a
good
way.
I
mentioned
how,
when
we
were
finishing
up
the
rim
raft,
stuff
ben
you
and
I
like
paired
for
a
couple
hours
there,
it's
just
a
lot
quicker
than
going
back
and
forth
and
stuff
so
yeah.
A
We
were
kind
of
saying
we
we
all
kind
of
like
that
idea,
and
maybe
we
should
try
and
do
that
kind
of
thing
more
regularly,
so
yeah.
I
would
definitely
be
interested
in
in
pairing
up
on
that.
F
Cool,
so
so
I
think
my
note
for
my
note
for
a
better
way
to
to
exit
unless
someone
else
ends
up
taking
it
on.
I,
I
think
that
I
probably
won't
have
the
bandwidth
until
we
do
recursive
copy
then,
because
I
pretty
much
can
only
take
on
one
thing
at
once
right
now,
my
amount
of
spare
time.
A
Okay
and
then
I
guess
after
that
is
source
map
support,
so
I
have
a
short
update
on
that
right.
Yeah
we
managed
to
backport.
F
One
of
the
prs:
now
we
have
ci
cd
building
with
a
failing
with
a
memory
leak
issue
which
is
is
really
weird,
because
the
mainline
branch
doesn't
have
a
memory
issue
and
node
14
before
landing.
This
patch
doesn't
have
a
memory
issue,
so
it
does
seem
like
this
patch
is
somehow
triggering
a
memory
issue,
even
though
it
already
lives
on
the
mainline
branch,
and
it's
not
triggering
a
memory
issue.
F
So
my
hypothesis,
which
I
have
not
tested
yet,
is
that
maybe
my
moving
to
weak
maps
actually
addressed
something
that
was
triggering
the
memory
leak
detection
is.
Is
that
what
elsa
does
right?
I
think
it
basically
detects
if
you've
gone
outside,
of
a
known
like
dress
space
or
something
right
like
that.
You
shouldn't
have
gone
outside
of
kind
of
like
an
electric
fence
kind
of
deal.
F
Who's
here
does
c
programming.
Does
anyone
do
c
programming
here,
but
I'm
I'm
pretty
sure
it's
like
it's
identifying
like
a
bug
where
you've
you've,
basically
written
outside,
of
an
expected
memory
address
which
I'm
not
even
writing
c
and
the
code.
I'm
writing.
So
it's
a
little
confusing
like
it
seems
like
something
would
be
more
likely
like
a
pointer
overflow
or
something
I
think
would
trigger
it
anyways.
F
This
is
a
long
explanation
of
it's
actually
a
pain
in
the
neck
to
back
port,
and
my
plan
is
to
try
back
porting
the
weak
map,
migra
the
weak
map
patch
first,
to
see,
if
that
passes
the
memory
issue
and
then
try
to
land
the
one
that
I
think
maybe
would
work
with
weak
maps.
Second,
so
switch
the
order
of
landing
and
just
on
a
hunch,
because
that's
that's
the
only
thing
I
can
think
of
that
that
might
have
corrected
the
the
memory
issue
on
the
main
branch.
F
Well,
basically,
what
we
were
doing
before
it's
kind
of
interesting.
Historically,
basically,
we
were
attaching
we
were
using.
I
was
asked
when
I
implemented
the
source
map
support
to
not
use
an
object
for
tracking
the
source
maps,
because
you
can
delete
from
require
cache.
So
there
was
a
concern
that
in
some
types
of
programs,
like
maybe
someone's
repeatedly
deleting
from
a
require
cache
and
tests,
you
could
actually
have
the
source
map
cache
start
to
grow
unbounded.
F
Like
a
memory
leak,
so
I
was
asked
to
implement
it
using
a
weak
map,
but
the
problem
is
that
in
an
exceptional
case
of
your
program,
it
never
actually
adds
the
module
into
the
require
cache.
So
there
is
nothing
to
key
the
weak
map
on.
So
I
couldn't
so.
I
couldn't
do
source
maps.
F
I
couldn't
do
source
maps
if
an
exception
was
happening,
but
my
clever
work
around
was
to
key
the
source
maps
on
the
air
object
itself,
but
the
error
object
itself
might
garbage
collect
before
you
actually
get
to
the
stack
traces,
so
that
to
me
sounds
like
an
idea
I
had.
That
was
maybe
too
clever
that
maybe
is
triggering
a
memory
leak
issue
which
I
fix,
which
I
fix
when
which
I
fix
when
I
move
to
weak
map,
so
so
I'm
kind
of
like
fingers
crossed.
I
hope
I
hope
that
helps
that's
the
issue.
F
Last
question:
for
you
rory:
how
do
you
actually
land
the
back
port
pr?
Because
because
I've
done
everything
now,
except
for
landing
it
do
you
just
like
manually,
push
it
or
something
or.
D
Cherry
picking,
I
think
we
check
we
do
cherry
pick
everything
into
the
release
branch,
but
the
only
point
is
that
I
think
the
staging
branch
is
usually
just
the
release
team
folks
that
play
on
it
just
because
they
might
be
cutting
a
release
right
like
there,
it's
a
so
yeah.
I
think
I
think
that
I
might
be
wrong.
I
might
be
wrong,
but
it
might
be
fine
for
any
collaborator
to
land
things
there,
but
it's
basically
the
same
thing
as
learning
a
master.
So,
okay.
F
F
So
I
guess
the
only
concern
right
now
is
that
this
I
can't
figure
out
this
memory
leak
and
then
I'm
like,
I
guess
I
run
a
docker
container
and
see
if
I
can
reproduce
it
in
a
doctor
container.
Maybe
like
I
don't
know
it
doesn't
happen
locally,
so
I
get
different
memory
leaks.
Luckily,
so
I
just
don't
trust
this
test.
Assen
thing
to
be
perfectly
honest,.
F
F
F
D
The
way
we
were
discussing
last
week
on
the
release
group,
I
think
the
the
next
minor
release
for
note
14
is
coming
up
soon.
D
Yeah,
I
think
it
might
be
a
good
idea
to
try
and
speed
up,
but
I
mean
also
like
it's
it's
definitely
how
time
we
actually
have
to
work
on
this
but
like
if
we
can
prioritize.
I
think
it's
a
good
idea.
F
I've
got
a
pretty
mellow
weekend,
so
I
might
spend
some
open
source
time
on
the
weekend.
Just
try
to
get
a
docker
container
running
or
something
the
frustrating
thing
is
that
I
don't
think
github
actions
actually
publishes
their
docker
container.
They
use
for
action
so,
like
I
don't
even
often
find
a
container
that's
similar
to
it.
I
suppose,
like
like.
I
think
I
can
at
least
know
what
linux
it's
on.
E
F
A
F
I
was
thinking
of
starting
there
if
githubs
are
proprietary
or
like
maybe
roy
could
well.
This
feels
like
maybe
it's
not
high
enough
priority,
maybe
to
take
a
huge
amount
of
time
out
of
your
work
day.
Roy,
I
was
going
to
say
you
could,
like
you
were
darcy.
You
could
maybe
run
the
container
yourselves
like
get
out
of
work
environment
and
see
if
it
happens
to
you,
but
I
can
try
to
reproduce
it.
First.
A
A
F
D
F
F
This
is
actually
a
really
cool
thing
that
lives
in
the
node
code
base
too.
Now
it's
39,
35
915
is
the
pr,
and
basically
it's
this
new
thing
that
you
can
recently
do
in
javascript,
where
you
can
have
something:
that's
both
a
weak
map
and
iterable.
F
A
F
B
I
have
I
have
no
update.
I
was
gonna
work
on
it
today
and
then
I
had
a
bunch
of
thrown
on
my
plate,
so
I
didn't
get
any
work
done
on
it,
but
soon
I'll
get
something
and
then,
if
anybody
wants
to
hack
on
it
too,
we
can
I'll
get
it.
You
know
up
into
the
pkg
js
repo
and
could
start
building
on
it.
F
Yeah,
I
think
I
in
my
twitter
feed,
I
saw
someone
probably
recently
published
a
polyfill
for
another
new
node
feature
everyone's
super
excited
about
and
I'm
completely
blanking
on
what
it
was.
But
I
I
will
go
through
my
twitter
feed
and
try
to
remember
what
it
was,
because
maybe
we
can
copy
the
approach
they
used
yeah.
It
would
be
great.
F
D
A
A
Yeah
once
I
think
I
mean
even
just
getting
like
a
starting
point
on
that,
get
that
up
and
then
and
then
we
can
all
kind
of
like
pitch
in
working
on
it,
because
I
really
like
to
test
that
out.
I
have
I'm
trying
to
parse
command
line
arguments
in
a
slack
command
and
I
would
love.
I
think
this
would
be
a
great
use
case
for
this.
F
F
C
Is
there
a
way
that
we
could,
I
think,
before
ben
you
said
we
might
be
able
to
steal
a
bunch
of
the
like.
The
three
thousand,
some
odd
tests
that
you
have
written
for
like
yards?
C
F
It
might
be
worth
just
using
them
as
inspiration
for
like
edge
cases
like
like
there's
tons
of
edge
cases.
Like
I
don't
know,
unicode
characters,
I
don't
know
you
could
try
to
run
it
again.
So
I
think
there's
just
the
test.
Suite
will
probably
fail
because
we're
missing
like
a
pretty
like
it's
a
pretty
different
api
right.
F
You
could
try
to
build.
You
know
what
a
more
interesting
thing
to
do
might
be
is
try
to
use
because
you've
built
it
in
such
a
way
that
you
think
it
can
be
extended
on
to
build
another
command
line
argument
parser
on
top
of.
Maybe
I
could
try
to
do
like
a
york's
parser
flavor
on
top
of
it
using
by
extending
on
it,
and
we
could
try
running
that
against
the
test
suite
that
might
be.
F
That
might
be
interesting,
and
that
would,
and
that
would
see
how
it
could
show
us
how
extensible
the
your
idea
is
too
of
like
populating
the
multiple
keys
and
yeah.
I
guess
I'm
saying
yes,
but
I
I
think
we
need
to
build
like
a
little
shim
layer.
On
top
of
it
probably
makes
sense
yeah.
F
F
C
Say
I
could
help
out,
because
I
know
I
did
a
little
bit
of
work
kind
of
like
in
playing
around
with
like
what
it
would
mean
to
build.
Almost
like
kind
of
you
know,
aliases
optional,
like
fallbacks,
bring
your
own
validation
type
thing,
so
I
had
that
little
code
sample
as
well
like
to
kind
of
like
shim
it
and
make
it
more
like
a
more
verbose
like
implementation
with
like
defaults
and
options
and
and
descriptions
and
all
this
kind
of
stuff,
but
yeah.
F
I
was
kind
of
thinking
like
I
was
kind
of
thinking.
We
could
just
add
it
to
the
ship
to
the
polyfill
repo
you
write
like
we
could
have
it
as
like
a
little
example,
like
example,
here's
the
arc's
parser
being
done
with
this
polyfill,
and
then
we
could
just
pull
in
then.
The
test
suite
then.
F
Or
like
some
subset
of
the
part,
because
there's
stuff
that
those
yards
parser
does,
we
probably
can
just
drop
on
the
floor
like
object,
notation
loading,
config
files,
it
does
way
too
many
things
it's.
I
can't
wait
to
replace
it
with
node.
It's
gonna
be
great
cool.
I'm
excited
about
this
sounds
amazing.
B
Cool
yeah
we
had
yeah,
we
had
jumped
around
sorry
go
ahead.
Ian.
A
Oh
I
have
this
is
like
a
total
non-sequitur.
I
was
actually
going
to
mention
it
earlier
when
we
were
talking
about
polyfills
and
then
it
reminded
me
of
when
we
first
started
working
on
rimraff
and
we
adding
new
modules
to
to
node.
Like
top
level
modules
was
a
no-go,
so
I
don't
know
if
this
seems
vaguely
tooling
related.
I
don't
know
if
you
guys
saw
there's
a
new
like
node
prefix
that
you
can
use
for
require
now
so
like.
Instead
of
like
require
fs
yeah
there
you
go
darcy
whatever
they
call.
A
Instead
of
require
fs,
you
can
say,
require
node
colon
fs,
which
actually
kind
of
opens
the
door
to
being
able
to
define
some
more
top
level
modules
without
having
to
get
like
the
name
on
npm.
A
F
A
The
I
think
the
name
space
is
just
like
built
into
node
telling
it
that,
like
like
this,
is
an
internal
module
like
don't,
don't
import
it
from
node
modules.
It's
like
a
built-in
one.
F
Gotcha,
I
understand
yeah,
so
so
we're
just
less
likely
to
get
like
a
like
an
injection
attack
or
something
because
you
accidentally
install
if
you
actually
include
fs
vulnerable
instead
of
ffs
or
something
yeah.
A
Right
and
also
we
were
kind
of
stuck
before,
where,
like
we
couldn't
really
add
new
top
level
like
internal
packages,
because
if
the
name
existed
on
npm,
then
you
couldn't
really
use
it.
Oh.
A
That's
what
came
up
with
with
rimraf
right,
originally,
we
were
going
to
put
it
in
in
like
sh,
util
or
whatever,
and
then
someone
pointed
out
like
this,
that
name's
actually
already
used
on
npm
so
anyways.
I
saw
this
the
other
day
and
thought
it
was
interesting.
F
Eliminate
some
fud
around
the
the
gradual
growth
of
standard
there
being
more
standard
modules
because
it
doesn't
feel
like
they're
all
leaking
like
it.
Just
feels
slightly
more
contained
and
there's
like,
like
newer
modules,
are
in
that
name
space.
I
like
that
idea.
F
A
Oh
yeah,
we
were
talking
about
a
couple
things
before
you
got
here.
The
the
member
list
on
github
is
outdated,
so,
like
darcy
and
joe,
like
can't
add
labels
to
things,
for
example,
because
they're
not
technically
on
the
tooling
team.
A
B
He
recently
left
ibm
and,
I
think,
he's
still
working
on
what
is
next
just
fyi.
F
That
probably
means
that
he's
in
a
transitional
phase
where
he
might
the
last
thing
you
might
want
to
hear
about
is
tooling
potential
could
go.
D
A
Yeah
that
also
just
kind
of
came
up
while
we
were
chatting
while
we
were
waiting
for
you
yeah
talking
about
that
again,
oh
yeah,
no
problem
yeah,
we
were
just
wondering
if
there
was
it
sounds
like
there's
not
going
to
be
one
at
open.js
world,
but
we
were
talking
about.
Maybe
just
doing
we
could
even
just
do
something
of
our
own.
Around
that
time
we
we
had
kind
of.
We
were
going
to
do
that
roadmap
meeting
earlier
in
the
year
and
then
we
kind
of
decided
hey.
A
We
actually
have
like
some
good
projects,
we're
working
on
right
now,
but
hopefully,
sometime
around
that
time
frame
like
june
or
july,
we're
maybe
landing
like
argument
parsing
and
some
of
that
stuff.
So
it
might
be
a
good
time
to
to
do
that
meeting.
Then
sometime
around
the
the
open
jazz
world,
I.
F
Agree
that
makes
sense
to
me.
I
think
it's
been
exciting
for
me,
watching
new
people
learn
node
at
my
job
right
now,
because
I
see
them
using
some
of
the
stuff
we've
built
like
rim
raft,
recursive
and
vic,
derp
and
stuff,
and
I'm
like
we're.
Actually,
my
colleague
found
that
stuff
the
other
day
and
I'm
like
like
calling
you
actually
used
without
me,
prompting
him
used
rm
recursive
the
other
day,
the
rm,
and
then
I
had
to
remove
it
because
I'm
like
well
that
doesn't
exist
until
node
14.,
so
you've
got
to
use.
F
You
got
to
use
rimder
until
then,
but.
F
Of
that
one
and
I
saw
pulling
in
shoot,
no,
I
saw
him
pulling
in
fs
extra
for
just
for
the
copy
dash
r
behavior.
F
A
Speaking
of
that
kind
of
thing,
I
I
think
we've
maybe
talked
about
this
before,
but
I'll
bring
it
up
again.
I
was
thinking
it
would
be
interesting
to
like
we
should
add
something
to
the
the
node
like
eslint
package,
so
that
like
if
you're
running,
node
14
and
you
import
rimraf,
it
should
tell
you
like
hey.
This
is
actually
built
in
now.
You
don't
need
this
anymore.
F
That's
a
good
idea
and
I
think,
to
a
certain
degree,
this
exists.
I've
noticed
if
I
change
the
engines
field
in
my
it
was
telling
me
I
couldn't
use,
I
think,
render
recursive
for
typescript
until
I
switched
to
my
engine.
Oh
sorry,
this
is
typescript.
Not
yes,
then
so
typescript
will
catch
that
you're,
trying
to
use
a
version
of
node
that
doesn't
yet
have
those
features
in
it
and
someone
must
be
updating
those
typescript
bindings
for
node
on
a
pretty
regular
basis.
C
Now
I
wonder
if
we
should
be
doing
something
in
npm,
though,
to
also
be
letting
you
know
that
you're
pulling
in
potentially
a
module,
especially
with
like
the
confusion
here
between
the
fact
that,
like
I
don't
know,
crypt
crypto
or,
like
maybe
other
like
libraries,
may
not
be
like
you're
pulling
down
something
that
is
now
like
native.
I
don't
think
we've
fully
blocked
out.
I
could
be
wrong,
but
if
you
go
to
the
registry,
I'm
not
sure
if
we
fully
deprecated
any
package
that
potentially
is
now
a
native
module
in
the
registry
itself.
C
So
it
might
be
good
that
we
or
we
like
that,
we're
doing
something
there
like
if
you're,
installing
a
package
that
we
know
is
now
conflicting
with
like
a
native.
Alright.
D
D
A
Well,
it's
it's
a
little
bit
tricky
though
too.
Right
because,
like
rimraf
is
largely
implemented
in
node
now,
but
if
you
want
the
bin
link,
for
example,
you
have
to
still
use
the
package
right,
so
the
package
isn't
totally
deprecated.
D
A
Yeah,
that
was
part
of
the
reason
why
I
like
the
eslint
idea
too,
is
because,
like
you
know,
if
you,
if
you
just
had
a
dependency
on
rimrap
like
that's
fine
you're,
not
gonna,
you're,
not
gonna,
see
maybe
you're
using
the
bin
link
and
you
just
have
a
script
that
uses
it.
But
if
you
actually
like
try
to
require
from
rimrap
in
your
code,
then
it
could
tell
you
like
hey.
You
actually
don't
need
this.
A
That's
smart,
yeah,
there's,
probably
a
few
things
we
could
do
there
like
ben.
You
were
mentioning
the
typescript
bindings.
You
know
making
sure
those
are
are
current
and
then
adding
something
like
eslint,
maybe
or
the
eslint
node
library.
F
F
Think
when
we
reached
out
to
james
halliday
about
mcdurp
originally
he
was
not.
I
mean
he
wasn't.
He
wasn't
the
most
supportive
he
likes
kind
of
that.
F
Well,
oh
yeah
yeah,
oh
you're,
talking
about!
If
you
like.
C
Yeah,
so
that's
what
that's
one
where
it's
like
there
wouldn't
be
anything
confusing
there
right
like
there's
you're
not
going
to
require
make
their
for
as
the
node
right.
So,
whereas,
like
what
I
think
we
were
alluding
to
was,
we
might
be
able
to
embrace
this
for,
like
parsers
or
other
future
packages
like
the
that
node
prefix
kind
of
like
helps
eliminate
some
confusion,
but
for
anybody,
that's
still
like
pulling
in
like
these
things,
potentially
npm
or
other
tooling.
C
F
I
think
for
the
polyfills
could
be
really
interesting
like
if,
if
we
like,
if
we
had
blessed
polyfills,
that
we've
done
in
the
package
org,
where
like
for
the
first
like
like
james
snell's,
one
or
the
one,
that's
being
done
for
parsers
like
mpm,
could
even
just
have
a
map
of
those
ones,
and
we
would
be
like
hey,
looks
like
you
were
experimenting
with
our
polyfill
before
it
actually
became
part
of
node.
Why
don't
you
just
use
the
one
that's
built
into
node?
C
A
Anymore,
cool
yeah-
I
guess
we're
pretty
much
out
of
time
any
any
last
minute
business
or
anyone
got
any
announcements
or
anything.
F
I've
been
working
for
like
four
months
to
make
eric's
async,
which
has
been
a
pain
in
the
neck,
but
I
feel
like
it's
almost
there,
so
it's
I'm
pretty
excited
about
yarg's
v17,
interesting,
it's
gonna!
It's
been,
I
feel,
like
it's
been
a
really
good
exercise
in
just
like
code
golf
to
try
to
make
it
work
properly.
F
Complete
completely
synchronous
until
you
do
the
first
thing
asynchronous
and
then
it's
asynchronous
so
it
like
it's
hasn't,
broken
the
contract
too
much
with
how
yards
used
to
work.
But
if
you
actively
try
to
use
async
stuff,
you
now
have
a
promise
instead
of
a
instead
of
a
parsed
arguments,
but
then
it
applies
the
middleware
and
the
handlers
and
the
checks
all
in
appropriate
order,
and
they
can
all
be
async
too.
So
you
can
have
like
you
can
have
like
async
validation
of
your
command
line.
F
A
Well,
I
actually
was,
I
had
not
an
issue,
but
was
I
had
like
an
async
handler
and
then
in
my
code
that
was
calling
yards
though
I
wasn't
awaiting
yards,
but
it
it
still
worked.
But
then
I
kind
of
went
down
that
rabbit
hole
too.
I
was
like
wait.
Does
this
need
to
be
what
I
call
a
yard's
parse
or
whatever
do?
I
need
to
await
that
and
then
so
it's
I
like
the
idea
of
that
being
more
explicit.
Maybe
that's
the
idea.
Try
yard.
F
Set
next,
and
it
should
be
correct,
now
like
if
you
await
yards,
it
will
not
resolve
until
your
handler
is
resolved.
A
F
A
Grateful
the
only
reason
I
was
using
it
with
webpack
is
because
I
was
bundling
a
cli
with
ncc
and
that
uses
webpack
to
do
that
and
it
worked
fine
with
yards
15.
But
then,
when
I
upgraded
yards,
my
build
broke.