►
From YouTube: Node.js Tooling Group 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
All
right
welcome
everyone
to
the
node.js,
tooling
meeting
it's
january
7th
2020.
A
I
guess
let's
get
started
here.
What's
up
first
on
the
agenda.
A
Oh
actually,
the
first
thing
in
our
last
meeting
I
was
trying
to
remember
there
was
an
issue
that
had
been
raised
on,
like
the
main
node.js
repo
that
I
thought
was
applicable
to
us,
and
I
could
not
remember
what
it
was
so
when
I
found
it
after
the
call
and
it's
this
recursive
reader
issue.
A
That
I
mean
maybe
who
knows-
yeah
anyways,
I
just
so.
I
just
linked
this
into
our
repo,
so
it
appeared
on
the
agenda.
I'm
not
sure
doesn't
like
there's
been
any
real
updates
on
it
for
the
last
month
or
so,
probably
not
that
surprising,
but
yeah
I
mean
this
was
open.
A
Yeah
it
looks
that
way
and
then,
like
more
recently
like
when
it
came
to
my
attention,
I
think
that
ben
you
had
commented
on
this
and
then
ethan
said
he
was
maybe
gonna
gonna
work
on
this
sometime
soon.
A
So
I'm
not
sure
that
there's
anything
for
us
to
do
here
unless
you
know,
maybe
I
think
he
just
like
changed
jobs
as
well.
So
you
know
maybe
he's
a
little
busy
at
the
moment.
We
could
always
try
to
help
out
with
this,
but.
A
And
actually
I
ran
into
this
and
something
I
was
working
on
recently
too.
A
A
I
have
not
taken
a
look
into
how
complex
that
package
is.
It
has
one
dependency.
A
I
think
the
some
of
the
issues
at
the
time
were
like
which
sort
of
like
format
would
it
follow?
Because
I
don't
know
if
there's
like
a
posix
standard
for
that.
A
A
C
Both
ethan
and
tierney
obviously
are
in
the
open,
js
slack.
If
we
wanted
to
tag
them
in
a
tooling
channel
and
just
you
know,
our
conversations
should
probably
be
in
the
github
issue,
but
if
we
wanted
to
kind
of
prod
them
a
little
bit
in
slack,
we
could.
A
Okay,
yeah:
let's
start
with
that,
okay,
anyone
have
anything
else
on
that
or
should
we
move
on.
E
A
Up
next
is
the
yeah
doing
something
to
like
call
out
new
features
in
node
that
don't
require
libraries.
I've
been
more
focused
on
things
around
argument
parsing,
so
I
haven't
done
anything
with
that.
One
lately.
A
So
yeah
no
updates
there.
Next
up
is
recursive
copy
anything
new
there
ben.
D
What's
new
one
second,
here,
just
looking
things
up,
there's
a
note
here.
D
So
james
said
he
would
not
james
said
he
would
not.
He
would
be
supportive
of
us
moving
to
stable.
He
was
one
of
the
folks
who
james
snow
was
one
of
the
folks
who
said
he'd
like
to
see
us
support
buffers
before
we
call
it
stable,
but
he
said
he'd
he'd.
Actually
you
know
for
not
getting
many
bugs
open.
He
said
it's
enough
of
an
edge
case
that
he
would
probably
not
he'd,
probably
be
supportive
of
us
moving
to
stable
without
that.
D
So
one
of
the
I
linked
one
of
the
things
that
maybe
we
would
want
to
experiment
with
before
we
called
it
stable,
though,
which
is
there's
apparently
open
dir,
which
is
a,
I
think,
like
a
async
generator,
instead
of
works
as
an
async
generator
instead
of
iterating
instead
of
loading,
everything
in
memory
and
they
were
pointing
out,
this
would
reduce
the
memory
footprint
of
copy
recursive,
so
that
might
be
worth
looking
at.
I
think
it's
like
a
drop-in
replacement.
A
Yeah
that
that
seems
worth
well
he's
trying.
E
D
D
D
I
guess
the
maybe
you
end
up
with
50
file
handles
open.
That
way,
though,
I'd
have
to
reread
the
algorithm
okay,
if
it's,
if
it's
recursive,
you
might
end
up
with
yeah
you're,
not
loading
all
the
directory
into
memory,
but
you
have
73
file
handles
that
are
open
and
iterating
on
a
directory.
So
I
I
would
need
to
look
at
the
algorithm
again.
I
think
see
if
the
recommendation
actually
works.
D
A
Totally:
okay,
anything
else
on
that.
A
A
Okay,
should
we
move
on
to
everyone's
favorite
topic
argument,
parsing.
A
All
right,
let's
do
it,
I
have
a
very
minor
update,
so
I've
been
trying
to
get.
I
extracted
that
node,
yes,
lin,
plug-in
node
core
from
node
core
into
my
own
library,
which
we're
using
in
the
in
the
project
and
started
a
conversation
with
michael
dawson
about
getting
that
moved
over
to
the
pk
gjs
org.
A
You
joe
you-
and
I
talked
about
this
in
slack-
I
don't
think
either
of
us
could
remember
if
we'd
come
up
with
a
better
name
and
then
just
a
couple
days
ago
someone
opened
an
issue
on
my
repo
and
mentioned
this
being
used
in
node
core.
So
it
seems
like
the
confusion.
Issue
is
real,
so
I
probably
should
rename
it.
A
I
don't
know
to
what,
though,
what
is
the
currently
so
it's
called
eslint
plug-in
node
core,
but
there's
like
a
placeholder
package
on
npm
with
that
same
name,
and
then
this
this
exists
directly
in
the
node.js
code
base.
So
it
is
technically
a
yes
link,
plug-in
it's
not
published
to
npm,
though
node
just
uses
it
like
internally.
In
its
own
source
code,
so
yeah
it
sounds
confusing,
so
it's
definitely
confusing
ideally
like.
A
It
would
be
great
if,
if
someone
just
published
it
to
that
package
on
npm
right,
but
we
thought
that
would
maybe
be
more
work.
So
we
were
just
going
to
make
our
own
version
in
the
pkg
js
org,
which
is
basically
done
now,
but
I
think
we
probably
should
change
the
name.
D
A
Be
the
one
that's
a
placeholder
yeah
I
mean
that
is
also
an
option,
I'm
not
sure
if
that
would
be
easier
or
harder
or
more
or
less
confusing
would
be.
I
mean
it'd
be
interesting
like
ideally,
that
package
would
be
published
to
that
existing
name
and
node
would
actually
use
that
package
rather
than
containing
everything
in
its
source
code
right.
E
I
are
you
talking
about
the
thing
that
people
are
already
doing
right
now,
or
is
this
a
separate
thing
that
you're
proposing,
because
I
I
swear
to
I
swear
that
was
just
proposed
and,
like
it's
being
done,
are
you
talking
about
that?
Or
are
you
talking
about
doing
that?
I
don't
know
I
I
saw
that
being
discussed
somewhere.
E
E
E
A
A
Yeah,
okay,
I'll!
If
there
is
an
existing
discussion
about
this
I'll,
I
can
reach
out
to
rich
and
see.
If
you
can
point
me
in
the
right
direction,.
D
A
F
Yeah
it'd
be
nice
if
you
could
do
like
a
recap
of
what
happened,
maybe
for
the
last
six
months
on
the
argument
who
wants.
C
Yeah,
I
don't,
I
don't
know
if
you
know
how
much
has
really
been
done
actually
in
the
last
six
months,
because
that
is
last,
six
months
has
gone
by
pretty
quickly.
There's.
D
Been
I've
been
kind
of
noodling
on
it
a
bit
because
there's
john
gee
who's,
john
gee
who's,
one
of
the
people
who
works
on
a
commander
has
came
in
and
opened
a
ton
of
issues
about,
came
in
and
opened
a
bunch
of
issues
about
just
outstanding
bugs
and
suggestions
about
usability
and
they've
all
been
super
reasonable
and
he's
been
he's
been
contributing
a
bunch
of
patches.
D
D
D
We
the
reason
we're
playing
a
lintering
is
we're
trying
to
get
this
repo
to
be
as
much
a
drop-in
thing
that
we
that
it's
the
right
style
is
same
style
as
node.js.
We
can
just
drag
and
drop
it
over
and
make
it
a
library
inside
of
node.js,
so
we
put
linting
in
we
had
a
couple
backs
and
forths
on
behavior.
D
I
think
we've
hammered
out
a
lot
of
the
edge
cases
in
terms
of
behavior.
The
like
there
was
a
question
about,
maybe
the
biggest
one
was
about.
Do
we
accept
named
options
with
a
single
dash,
and
I
think
that
for
leading
now
towards
not
accepting
named
options
with
a
single
dash,
because
it
eliminates
a
lot
of
ambiguity
and
then
we
can
do
grouped
booleans
like
rm-rf,
which
you
know
is,
opens
up
a
whole
set
of
clis
that
we
couldn't
do.
D
If
we
had
this
ambiguity
so
we've
had
those
discussions,
we
haven't
actually
implemented.
Boolean
option
sets
yet,
which
I
think
is
one
of
the
last
features
for
implementation.
D
We've
done
a
few
minor
things
like
we
decided
to
call.
I
think
it
was
called
options
and
flags.
We've
decided
to
call
that
values
and
flags.
I
think
because
option,
because
both
things
are
options
like
they're
boolean
options
and
options
with
values.
D
So
so
we
made
that,
like
one
little
semantic
change,
what
else
is
there
worth
saying?
I
think,
oh
sorry
go
ahead.
If
you
have
more,
the
the
only
other
real
one
I
had
was.
A
The
I
think,
the
one
other
thing
unless
it's
in
this
pull
request
that
I
haven't
reviewed
is
we
we
don't
have
any
implementation
of
the
strict
argument.
So
the
idea
is
like
in
the
non-strict
mode
you
get,
you
would
get
an
error
object
back
for
any
invalid
flags.
I
guess,
whereas
in
the
strict
mode
it
would
just
throw
so
I
don't
think
that's
implemented,
and
I
was
I
was
kind
of
interested
in
maybe
giving
that
a
try.
B
Put
a
question
in
there
because
I
I
know
I
haven't
been
here
in
a
while,
then
I
was
trying
to
play
catch
up
with
what
you
just
said:
there
ben
sorry,
the
the
the
shorts
or
aliases
are
going
to
be
able
to
be
defined
together
or
we're
going
to
parse
those
out
in
some
way
yeah.
I
think
what
oh
thank
you.
A
B
You
could
pass
that.
Oh,
it's
dash
a
b
so,
like
the
thing
is
with
aliases
and
shorts,
they,
they
aren't
necessarily
one
a
single
character
unless
we're
saying
that
we
have
some
sort
of
like
validation.
I
think
the
idea
originally
with
parsecs
was
like
bring
your
own
validation
and
so
doing.
This
kind
of
stuff
starts
to
like
really
muddy
the
waters
right.
D
Yeah
right,
yes,
yeah
because
I
mean
like,
like
I
think
both
both
myself
and
the
person
from
commander
agreed
that
it's
kind
of
a
weird
affordance
to
allow
for
the
single
dash
like
like
there's,
no
reason
to
not
force
someone
to
say
dash
dash
foo,
most
command,
like
a
lot
of
command
lines,
require
that
and
it's
and
you
get
more
utility
out
of
being
able
to
do.
Rm-Rf
like
like
cause.
That's
like
get.
Does
that
rm?
D
If
we're
going
to
eliminate
a
source
of
ambiguity,
I'd
rather
us
force
you
to
have
two
dashes
for
an
option
than
to
lose
the
ability
to
build
like
rimraf,
with
really
really
easily.
Where
have
we
lost
you
darcy.
B
Just
the
exception
of
shorts
being
only
a
single
or
like
aliases,
only
being
a
single
character,
because
npm
is
not-
and
this
is
before
my
time
of
managing
the
cli.
But
log
level,
for
instance,
is
like
dash
d
dash
dd
dash
dd
for
like
verbose
map,
silly
log
levels,
which
is
obviously
not
the
standard.
Let's
say
like
the
the
normal
practice,
but
yeah.
A
B
We
have
others
that
are
not
this
way
as
well.
Is
that
that's
just
one
example,
ws
is
another.
One.
Workspaces
can
be
passed,
this
ws
that
what
that
was
a
bad
decision
on
our
end,
but
we
have
there's
a
whole
bunch
of
other
flags
like
that.
I
know.
Other
people
have
aliases
that
are
multi
characters,
so
so.
C
D
I
think
we're
relitigating
like
because
we've
argued
about
this
quite
a
bit
like
like.
I
think
you
lose
a
lot
of
benefit
if
someone
can't
hammer
together,
something
like
git
or
rimroaf
like
like
that
option.
Groups
are
really
common
in
command
lines
and
when
I
look
at
something
like
ddd,
like
that's,
actually
represented
in
command
line,
parsers
sometimes
differently
like
what
you'll
sometimes
do
is,
is
have
a
count.
Account
argument
where
you
actually
tell
how
many
times
it's
been
provided
so
so
like
in
yards.
D
C
C
Just
kind
of
I
you
know
I
I
don't
want
to
relitigate
either,
but
you
know:
could
we
assume
parsing
short
code
single
characters
but
also
have
another?
You
know
object
or
array
that
we
return
that
passes
it
along.
If
you
don't
want
to
use
the
shortcodes,
I
don't
know.
E
Thank
you.
I
just.
I
think
that
there's
also
like
I,
I
think
ddd
is
one
example,
but,
like
you
know,
I
think
that
there's
yeah
I
I
guess,
given
that
we
I
mean
you're,
correct,
that
that's
like
a
power
user
thing.
We
also
have
a
power
user
here,
who's
kind
of
asking
for
that,
and
I
think
that,
like
I
don't
know,
I
feel
like
we
should
try
to
figure
out
ways
that
we
can
yeah.
A
I'm
I'm
in
favor
of
supporting
like
the
multiple
of
the
same
single-letter
flag,
but
maybe
not
in
favor,
of
supporting
multi-letter
or
character.
Aliases.
D
I
feel
the
same
way
like,
like
I
think,
like
ddd,
would
be
easy
to
differentiate
from
dash,
say
rm,
because,
because
like
we
could
we
could
like
special
case,
we
could
special
case
increments.
It
doesn't
work
well
for
the
other
example,
which
was
what
like
dash,
can
you
what
was
that
other
example
of
star
citizen?
Oh
yes,
it's
ws
yeah,
so.
B
Yeah
I'm
wondering
like:
do
we
have
exactly.
D
A
F
A
F
D
E
D
C
A
So
just
to
clarify,
are
we
saying
like
by
default?
We
would
parse
that
as
single
letter
or
single
character
flags,
and
then
we
would
also
pass
the
whole
string
somewhere
so
that
if
you
want
to
do
something
else
with
it,
you're
free
to.
F
That's
what
I
was
trying
to
suggest.
I
think
what
ben
was
suggesting
before
was
the
other
way
around
right.
We
don't
do
anything
because
it
might
also
be
a
way
we
could
do
it.
We
don't
do
anything
by
default,
but
then
we
show
the
you
know
what
provide
examples
provide
like
okay.
This
is
an
easy
way.
You
can
just
turn
a
group
of
letters
into
as
if
you
have
sent
many
options.
B
At
once,
like
this
becomes
a
positional
right
like
so
this
becomes
a
positional
and
what
you
can
do
is
you
starts
with
to
find
the
dash
and
then
basically
split
and
do
exactly
what
you
want
to
do,
which
is
to
like
map
these
shorts
to
whatever
you
want
right.
That's
what
I
imagine
so
we
don't
do
anything
special
for
when
we
sing
a
single
dash.
We're
only
do
parsing
when
we're
seeing
multiple
dashes
right.
Does
that
make
sense,
and
then
the
skits.
C
Well,
I
was
gonna,
try
to
make
sure
I
understand
you,
so
it
would
be
in
the
positional
array
or,
and
it
would
just
be,
it
would
start
with
a
dash
as
a
positional.
So
you
would
know
this
is
what
you
would
you
know
you
would
do
your
own
parsing
on
that.
B
Yeah
and
then
that
would
limit
because
again
I
know
I
know
there
was
a
an
ask,
and
I
think
you
answered
it
joe
at
one
point
or
maybe
michael
dawson,
had
the
ask.
I
forget
around
what
happens
when
we
see
three
dashes.
What
are
we
doing
that
we
see
ten?
You
know
ten
dashes
like
you
know
like
what
are
we
eventually
gonna
do
right
like.
D
I
think
if
he
did
like
dash
dash
dash
foo
the
recommendation
of.
B
D
That
I
was
going
to
say,
like
these
discussions,
have
been
happening
on
the
repos.
I
think
we
have
to
have
the
discussions
there
because
we're
not
we're
not
looping
in
we're,
not
looping
in
sorry.
I
always
forget
his
name,
john
we're
not
living
in
john
gee
who's
been
making
good
arguments
from
the
commander's
perspective.
You
could.
A
Make
full
requests
for
the
read
me
with
these
changes
and
then
and
then
use
that
to
kick
off
a
discussion.
C
D
C
I
mean
an
easy
path
forward
is
just
you
know
if
it
starts
with
two
dashes
and
then
no
other
dashes,
we
handle
it,
and
if
it's
anything
one
or
three
or
more,
then
we
pass
it
as
a
positional.
The
whole
dang
thing
it'd
be
like
an
easy
way
to
just
two
dashes
are
a
thing.
Any
other
dashes
is
a
positional.
You
know,
as
it
is.
B
E
I
I
think
I
also
think
out
of
the
box,
and
you
can
turn
it
off
is
fine,
like
I
think
that's
a
fine
thing
to
do.
If,
like
someone
wants
to
turn
that
off
and
like
write
their
own
custom
handling
of
it,
I
think
we
should
probably
allow
that.
D
D
D
F
B
E
D
B
D
F
D
E
I
think
ethan
ethan
and
I
both
joined
because
joe
pinged
us
in
a
slack
channel.
Yes,.
D
A
Yeah,
let's
go
back
the
recursive
reader
agenda,
so
there
was
a.
There
was
an
issue
in
the
node.js
repo
that
the
two
of
you
were
involved
in
and
then
so
we
are
big
fans
of
adding
recursive
things
to
node
and
so
yeah.
We
were
basically
just
wondering
you
know.
Ethan
you
had
said
you
were
interested
in
working
on
that,
so
we
were
wondering
if
that
was
still
the
case
or
if
you
wanted
us
to
pick
that
up
happy
with
either.
G
I'd
definitely
love
to
contribute
it.
I
think
I'd
love
to
hear
from
tierney
and
what
they've
done
in
the
like
recursive
reader
kind
of
project.
G
If
we
wanted
to
vendor
something,
if
folks
just
wanted
me
to
add
a
recursive
option
to
the
reader
like
function
like
whatever
we
all
think
is
the
best
way
to
do
it.
I
definitely
have
some
time
this
month
that
I
can
get
this
started
and
working
and
we
can
try
and
land
it
and
discuss
over
pr,
usually
works
best
in
the
node
repos.
E
I
just
context
on
my
quote:
unquote:
work
in
recursive
reader.
I
was
trying
to
build
this
and
then
I
found
there
was
a
module
and
it
hadn't
been
published
and
it
had
a
problem.
So
I
just
asked
if
I
could
be
added
and
I
got
added
and
then
that
was
it,
so
I've
not
done
work
on
recursive
reader,
but
I
have
published
access.
So
yes
that
that
is
the
extent
of
my
work
because
I
asked
nicely
if
I
could
be
a
maintainer
and
then
I
never
published
anything
so.
G
I
think
I
can
in
that
case
I
can
probably
just
take
what
exit,
because
recursive
readers
downloaded
like
four
million
times
or
something
so
a
lot.
I
can
just
take
this
and
see
what
I
can
sort
of
vendor
it:
sort
of
copy
paste
copy
the
license
and
all
into
node,
and
then
we
can
take
it
from
there.
I
know
devsneck
wrote
a
pretty
cool
like
async
generator
version
and
I've
added
some
additional
pieces
to
that
in
that
original
issue.
So
we
can
take
it
either
way.
E
This
is
I
I
wish
the
npm
website
was
slightly
better.
Is
there
a
way
to
see
like
the
most
downloaded
packages
that
rely
on
a
package
like
I
don't
I
don't
know
how
to
find
that
the
answer
might
be.
There
isn't
one
and
I
have
to
like
loop
through
all
1900
dependencies,
but.
E
D
Oh
sorry,
if
you
look
at
the
dependents
rc
can
carry
everybody.
Can
you
can
correct
me,
but
it
should
be
sorted
in
in
order
of
dependence
like
in
order
of
number
of
people
who
depend
on
it.
If
you
click
on
the
dependents
tab,
so
so,
like
the
top
one's
going
to
be
the
most
downloaded
one.
E
Interesting,
I
never
knew
that
it
seems
no.
That
is
not
the
case,
because
the
first
one
has
two
downloads
weekly,
really
yeah,
there's
other
ones
that
have
much
more
like
aws
s3
deploy
probably
has
nope.
It
has
six
well,
okay,
it
has
more,
but
not.
It
is
not
ordered
in
that
order.
So
yeah,
just
figuring
out
that
and
like
reaching
out
to
people
who
like
screen,
relies
on
it.
I
don't
know
if
that's
still
being
developed
or
not,
but
you
know
it
has
twenty
thousand
dollars
a
week.
D
E
E
Like
react,
dev
utils
relies
on
it.
Oh
hey!
This
is
all
of
the
downloads.
Yes
you're
you're
responsible
for
all
the
downloads,
so
yeah
we,
we
can
probably
wait
that
was
published
recently
by
yeah.
Okay,.
D
E
The
sort
order
years
if
ian
changes
react
wtills
to
this
implementation,
four
million
downloads
will
go
away
from
a
recursive
reader,
so
that
works
out.
G
Great
all
right,
another
aspect
is,
we
probably
haven't.
We
could
probably
also
do
some
benchmarking
and
see
what
this
you
know.
What's
the
best
way
to
do
it
fastest
way
to
do
it
and
incorporate
that
into
the
into
the
vendoring
addition
to
node
core.
E
D
Was
just
gonna
inject
if
you
take
a
look
at
the
copy,
recursive
copy
code
that
went
in
recently
ethan,
but
like
that's
taken,
that
was
just
dropped
directly
over
from
fs
extra,
so
it
was
vendored.
But
you
can
look
at
how
we'd
like
what
directory
we
chose
to
look
at
put
it
in
and
the
cleanups
that
we
ended
up
having,
because
we
did
a
bunch
of
cleanup
just
to
make
it
look
more
like
node.js
style
so,
but
but
that
was
actually
vendoring
like
it
was
just
the
file
was
brought
over.
I
think.
A
G
Yeah,
that's
what
I'm
saying
copy
and
pasting:
that's
what
we
did
in
udichi.
We
took
my
indici
fetch
and
we
copied
and
pasted
it
and
then
it
as
well
as
the
license,
and
it's
like
this
is
ethan's
original
work,
but
it's
then
been
you
know,
added
to
removed
from
et
cetera.
So
that's
what
I
was
thinking.
I'm.
D
Saying
the
same,
I'm
saying
the
same
thing
I
just
priced
interviews
the
word
vendor
I
copy
and
paste
it
I
copy
and
pasted
copy
recursive
over
from
fs
extra,
but
did
a
bunch
of
notisms
to
it.
My.
E
Tiny
tiny
problem
with
it
is
the
license
for
recursive
return
doesn't
actually
list
the
copyright
holders.
It
just
has
copyright
holders
in
brackets
because
they
didn't
fill
it
out.
E
A
We'll
figure
it
out.
Okay,
maybe
we
can
figure
that
out.
I
was.
I
was
just
going
to
make
two
other
comments
about
the
implementation
details.
One
before
you
joined
the
call
ethan
we
were
saying
the
current
recursive
reader
has
a
dependency
on
mini
match,
so
we
were
kind
of
proposing
that
maybe
we
would
just
exclude
the
globbing
functionality
for
the
time
being,
which
is
exactly
what
we
did
with
rimraf
as
well.
A
So
that
was
the
first
one
and
the
second
thing
I
was
going
to
say
in
terms
of
like
the
api,
I'm
thinking,
maybe
just
the
adding
a
recursive
option
to
reader,
makes
sense.
What
is
I'm
curious?
What
everyone
else
thinks.
A
D
I
might
be
repeating
something
you
said
already
so
sorry
if
I
did,
and
but
we
were
talking
earlier
too,
about
glob
not
being
a
thing
in
node.js,
but
what
copyrecursive
did
to
give
you
something
kind
of
like
globbing
was
you
can
provided
a
closure
and
then,
if
the
closure,
if
you
provide
a
closure,
you
can
return
true
or
false
as
to
whether
you
want
the
file
in
the
result
set
so
like
you
could
maybe
think
about
that
as
an
option
instead
of
glob
and
it
and
it
matches
how
we've
dealt
with
it
elsewhere.
G
Yeah,
well,
I
think
that
right
now,
it's
you
know.
If
you
call
fs.reader
you
just
pass
it
a
a
string
or
a
path
that
is
a
directory
and
it
will
error
if
it's
not
a
directory
and
then
it
will
just
return
the
the
top
level.
I
think
the
addition
of
the
recursive
option
is
simply
to
just
continue
down
the
path
of
each
of
those.
You
know,
if
it's
a
you
know,
if
a
path
is
a
directory,
just
continue,
adding
it
to
the
resulting
list.
G
Does
reader
already
have
an
option
such
as
like
filter
or
or
whatever,
where
like?
If
you
were
to
you
know,
just
call
it
as
it
is
now
not
recursively.
G
Could
you
filter
out
some
part
of
the
results
and
if
it
doesn't
have
that,
then
maybe
I'll
say
we
shouldn't
be
making
any
change,
I
think
at
first
we
should
only
just
add
deep
recursion
reader,
and
if
we
want
that
kind
of
filtering
feature,
that
could
be
like
a
second
thing
to
do
after
the
first
bit
lands.
It
doesn't
have
that
option
right
now,
I'm
just
reading
the.
D
A
Cool,
I
agree
with
what
you're
saying
there.
You
can
always
add
it
later
and
also
there's
a
possibility
at
some
point.
We
end
up
with
a
glob
implementation
in
in
node,
and
then
we
just
use
that
so.
G
Nice,
so
I
will
get
started
on
this
now,
I'm.
Luckily.
I
have
a
lot
of
free
time,
so
I'll
try
and
get
something
going
and
then
I'll
make
sure
at
the
right
people
and
we
can
get
some
reviews
going
through
thanks.
Ethan
awesome.
A
All
right,
that's
kind
of
everything
for
the
agenda.
This
was
a
good
meeting.
It's
the
most
people.
We've
had
had
a
meeting
in
a
while.
E
I
realized
it
is
not
on
my
calendar,
like
I
thought
it
was
so
I
haven't
been
showing
up.
I
will
try
to
show
up
more
in
the
future.
A
D
D
D
Are
folks?
Okay,
with
the
moving
up
to
the
word
values
from
options?
The
argument
was
very
sound.
I
thought,
which
was
that
they're.
Both
options,
that
was
the
only
like
I'd,
say:
significant
change.
That's
landed.