►
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
B
Okay,
hi
everyone:
this
is
the
no
jazz
tooling
group
meeting
for
december
10th,
2021
yeah.
So
should
we
should
we
start
with
the
stuff,
that's
on
the
agenda
and
then
get
to
the
new
business
at
the
end
or
that's
good
to
me.
B
C
Brief
introduction
for
daniel
too,
to
explain
what
we're
we're
up
to
I'm
happy
to
do
that
sure.
So
we've
been
having
this.
I
think
I
explained
this
to
you
a
little
bit
and
when
we
were
talking
about
it,
daniel
that
basically,
we've
been
having
this
meeting
for
a
few
years
now
and
it
was
originally
started
to
try
to
improve
the
api
surface
in
node.js.
For
folks
writing
things
like
yargs
or
ian
works
on,
create
react
app
and
some
folks
from
mocha
were
in
the
started.
It
too.
C
C
Most
of
the
work
we've
done
has
been
to
add
improvements
to
the
file
system,
we've
which
we've
kind
of
championed
over
the
years
and
then
there's
code
coverage,
and
then
we've
been
talking
a
lot
about
argument
parsing
lately,
which
I
think
is
in
the
agenda
today.
D
B
The
the
first
thing
on
the
agenda-
I
guess,
is
me
the
this
thing
we
were
talking
about
some
kind
of
like
lint
rules
or
whatever,
to
tell
people
about
things
that
are
built
into
node,
but
I
don't
know
it's
been
kind
of
like
on
pause
for
a
while,
I'm
not
really
sure
what
the
best
way
to
go.
There
is
so
no
update
for
me
on
that.
C
Yes,
one
second,
though
I
was
going
to
find
one
link,
I
want
to
link
if
I
can
find
it.
I
might
have
shared
it
last
time
here
it
is.
I
think
I
probably
shared
this
last
time,
but
I
keep
seeing
these
threads
on
twitter,
where
people
are
basically
asking
for
exactly
this
linting
tool
or
like
something
that
would
point
them
in
the
right
direction.
So
I
thought
I
would
share
that
with
that.
Tweet,
oh
and
I
don't
know
if
I
shared
this
before,
but
the
person's
like.
C
So
it's
like,
I
don't
know
it's
kind
of
neat,
because
it's
someone
saying
that
they
like
what
we've
been
doing,
but
also
that
maybe
kind
of
shows.
We
need
to
find
a
way
to
make
sure
people
right.
All
these
new
features
exist
right
so,
but
maybe
we
can
get
some.
Maybe
we
need
to
get
more
people
excited
at
an
in-person
meeting
or
something
in
the
new
year.
C
B
C
C
So
I
was
asking
if
we
could
just
treat
this
as
a
missing
feature
and
be
open
to
addressing
it
as
soon
as
we
have
some
users
who
are
blocked
on
using
copy
because
they're
in
some
keyboard,
some
unicode
form
that
doesn't
work
properly
with
the
string
and
james
said,
he's
fine
with
that.
So
I
think
we
could
probably
move
forward
with
calling
it
stable
pretty
soon.
C
There
aren't
too
many
open
issues.
Maybe
we
wait
for
maybe
we'll
get
more
adoption
with
note.
16
did
note.
16
recently
go
stable,
or
is
it
going
stable
soon,
yeah,
maybe
give
it
a
couple
more
months
before
we
call
it
stable
see
if
we
have
any
issues
open,
but
I
think
recursive
coffee
is
in
node
16
right,
yes
from
the
gecko
yeah.
B
C
Yeah,
I
think
you
I
mean
as
long
as
we're
cognizant
and
keep
an
eye
on
if
there's,
if
there
actually
are
issues
that
are
biting
customers
right
or
if
I
say
customers,
I
mean
users
because
they're
not
really,
I
think
yeah.
My
work
brain
is
running
right
now.
Yeah
I
mean:
when
did
it
go
out?
Recursive
copy
went
out.
C
C
B
C
A
A
C
To
update
this
edit,
I
have
so
that
it's
flags
and
values
instead
of
flags
and
arcs,
but
I
basically
based
on
the
feedback
from
let
me
actually
link
it.
So
if
you
know
what
I'm
talking
about,
I
have
a
straw
person
proposal
to
rename
ops
and
what
was
it
originally.
It
was
args
and
values
originally
args
and
values.
C
Shadow
spawn
is
their
name,
what's
their
actual
name,
non
github
name
john
gee,
who
is
one
of
the
main
maintainers
of
commander.
So
this
is
great.
We
have
yards
people
and
commander
people
working
together
to
try
to
add
nursing
into
node.
So
I
think
this.
I
think
this
will
really
help.
I
think
when
we
actually
try
to
merge
it
into
node.
This
will
make
a
really
strong
case
that
we
have
two
of
the
main
argument.
Parsing
implementers
can't
wait
to
stop
rating
argument.
C
Parsing
implementary
implementations,
so
that's
not
kidding,
but
I
mean
it
looks
like
you
signed
off
ian,
but
are
you
okay
with
me
switching
it
to
values
instead
of
args.
B
Yeah,
I
don't
actually
remember
seeing
these
comments
when
I
reviewed
it,
I
must
have
just
missed
them.
So,
yes,
I
am
fine
with
that.
I
think
I
think
that's
they
make
a
good
point.
So
let's
do
that.
C
C
C
C
B
C
B
C
I
think
that's
related,
I
think
number
11
and
number
six
are
pretty
related.
They
both
relate
to
like
how
strict
is
the
parser.
Do
we
throw
errors?
Do
we
just
return
an
error
object,
so
I
think
if
we
could
solve
how
we
handle
errors,
then
that
would
close
two
of
the
blockers
for
proposing
it
landing
a
node.
B
C
C
I
I
don't
know
I
don't
feel
super
strongly
about
whether
we
return,
whether
it's
possible
to
have
an
error
or
not,
and
whether
we
throw
trying
to
think
of
how
in
jargon's
approach
I
definitely
want
to
do
which
is
it
suppresses
errors
and
prints
out
like
a
pretty
error
as
its
default
behavior,
which
I
think
would
be
better
if
someone
either
got
an
error,
object
or
caught
and
then
did
that
themselves.
So
that's
what
they
wanted,
because
I
think
what
we've
been
talking
about
is.
B
C
If
you're
writing
a
cli
that
takes
two
arguments
or
just
something
simple,
you
know
like
you
might
probably
just
want
to
put
in
a
try
catch
block,
and
then
you
decide
whether
or
not
you
want
to
print
the
error
or
not
right.
Yeah
maybe
like.
If
you
want
to
raise
an
error,
then
you
just
don't
put
a
try
catch
and
if
you
want
to
print
an
error,
you
catch
it
and
print.
Just
the
message
right
right.
C
B
C
There
have
been
a
few
folks
suggesting
that
parse
never
fail
on
user
input
and
perhaps
air
information
could
be
a
property
of
in
the
results.
I
see
the
attraction
and
interested
in
and
try.
I
see
the
attraction
and
I'm
interested
in
trying
that
approach.
A
C
C
B
I
guess
I
had
one
there.
I
have
one
issue
on
here
that
I'm
working
on
as
well,
which
is
just
getting
the
that
islam
plug-in,
that
I
made
moved
over
to
pkgjs
and
I
actually
have
access
to
like
the
npm
account
now,
so
I
can
publish
it.
C
E
Just
talk
over
here:
oh
no,
it's
okay!
I
was
actually
on
another
question
on
the
arc:
parse
strict
option.
So
what
happens?
If
you.
E
So
it
says
it
throws
an
error
on
unknown
args,
but
what
happens
if,
if
it's
expecting
a
value,
and
you
pass
in
an
arc
like
a
like
it's
a
flag
like
something
that
expects
something
at
the
end,
but
you
don't
like
give
it
like
an
actual
value,
should,
should
it
throw
if
strict
is
false
or
how
do
we
want
to
handle
that
situation?
I.
B
C
I
think
so
and
then
like,
if
you
have
anything
that
if
you
have
any
things
that
don't
take
proper,
don't
always
take
arguments,
then
those
would
be
probably
flags,
I
think
at
which
point
it
probably
wouldn't
be
an
area
it
would
just
be.
If
you
don't
give
it
an
argument,
it
becomes
true.
I
think
right.
So
that's
probably
the
behavior
we
landed
on.
C
B
Yeah
I
was
thinking
it
would
be
like
yeah.
This
field
produce
this
error.
Maybe
I'll
try
building
this
strict
mode.
This
weekend
sounds.
C
C
All
right,
I'm
going
to
add
the
new
business
stuff
in.
So
let's
talk
about
daniel's
idea
of
find
up,
which
I
thought
was
I'd,
never
thought
of
and
actually
let
me
make
sure
I
think
I
directed
daniel
towards
find
out,
but
he
described
a
different
problem
to
me.
So
I
want
to
see
if
I'm
going
to
call
it
what
ben
is
describing
is.
E
Yeah,
I
think
something
like
I
think
it
would
be
nice
if
packages
had
the
ability
to
click
to
quickly
and
conveniently
get
its
its
nearest
package
json.
E
So
that
way,
you're
able
to
get
let's
say
if
you
want
to
print
out
the
version,
if
you're
a
cli
tool
or
if
you
just
want
general
metadata
from
the
package
json
and
currently
there
isn't
a
clean
way
to
do
that,
except
for,
if
you
do
like
a
a
relative
import,
but
the
problem
with
that
is,
if
you
use
something
like
typescript,
where
it
might
move
your
your
your
build,
and
it
might
break
that.
E
So
I
think
in
a
language
it
would
be
nice
if
there
was
a
convenient
way
to
get
the
factory
slump.
C
Looks
like
bernoulli
has
a
question
I
may
be
related
to
this.
Do
you
want
to.
B
Notice
here
let
me.
C
D
Yeah,
so
I
don't
think,
there's
really
a
problem
with
it,
but
there
are
some
like
caveats
that
probably
should
be
discussed.
This
has
been
brought
up
a
bunch
of
times
in
the
issue.
Tracker
one
is
your
package.json
might
not
be
the
one
at
the
root
of
your
package
if
that
makes
sense,
so
some
tools
do
have
like
nested
package
jsons.
D
The
other
is
like
knowing,
when
you're
exiting
your
package,
like
you,
probably
want
to
have
some
kind
of
barrier
to
not
cross
node
modules
directories,
like
probably
don't
want
that
also
preserve
sim
links,
makes
this
like
an
absolute
nightmare.
D
D
Because
it
would
be
like
asking
what's
the
package
json
for
any
file
that
could
point
to
me
and
there
could
be
multiple
like
reserve.
Simulinks
is
a
very
bad
state
of
the
world
so
like
if
preserved
sim
links
is
on.
This
has
to
throw
like
it's,
it
would
work
sometimes
but
other
times
it
would
just
give
you
the
wrong
answer.
B
I
have
a
question:
is
there
I
mean
there's
no
like
opposite
of
this
functionality
in
node
either,
though
right,
there's
no
like
find
there's
like
fs
find.
Is
there
they're.
C
D
D
There
is
an
alternative
that
has
been
discussed
a
few
times
that
one's
a
little
bit
harder
to
do,
because
when
you
do
a
resolve
operation
like
you're,
resolving
babble,
slash
parser,
there
are
actually
multiple
package
jsons
involved.
There
is
the
resolution
package.json
and
then
there's
the
metadata
package
json.
This.
I
think
the
feature
request
is
just
to
find
the
closest
one
towards
the
root,
not
there's
only
ever
one,
because
there
is
an
inverse
operation,
but
it
it
has
multiple
potential
things.
People
want
right.
C
Like
I
think,
daniel's
original
question
to
me
was
basically,
how
do
I
know
the
package
json
I
care
about
as
as
a
see
as
an
example
of
cli,
so
that
I
can
show
the
version
number
of
the
specific
package
json.
I'm
like
an
example
in
yards
would
be.
C
I
have
yards
as
a
dependency,
but
I
am
myself
a
package.
I
am
myself
a
cli
tool,
but
I
have
yards
as
a
dependency.
C
I
want
to
print
out
the
version
number
of
the
thing
wrapping
yards,
not
not
yargs
itself
right.
So
you
end
up
in
these
positions.
D
D
C
E
I
yeah,
I
think,
it's
yeah
and
even
in
the
arcs
case,
if,
if
there's
a
way
that
the
the
local,
the
the
thing
that's
depending
on
your
arts,
to
get
its
version
and
then
just
inject
it
into
yard,
because
an
argument,
I
think
that
could
work
as
well.
D
Yeah,
so
meow
was
a
popular
package
that
did
something
similar
and
they
stopped
doing
that
functionality
and
now
do
require
a
parameter
be
passed.
It
does
that
make
sense.
C
What
do
I
find
interesting
like
so
in
the
typescript
example,
say
you're
going
into
a
build
directory
daniel?
You
basically
just
want
to
find
the
package.json
that
lives
outside
that
build
directory,
because
that's
going
to
be
the
version
number
you
carry
that's
going
to
be
the
metadata
you
care
about
right,
so
correct.
E
Exactly
yeah
exactly
and
if
and
if
one
of
the
parent
directories
is
like
note
modules,
then
we
just
throw
an
arrow
raise
an
arrow
if
it's,
but
if
you
recurse
up
and
you
find
a
package
json
grab
it
get
this
metadata.
C
C
D
D
B
C
I
thank
you,
for,
I
think
you've
actually
asked
for
literally
the
opposite
thing
ian
which
isn't?
I
don't
think
it's
a
bad
idea
either,
but
I
think
it's
that's
why
we
need
to
get
the
naming
right,
because
this
is
like
going
like
brad
says
you're,
where
you
are
going
up
up
up
up
up
until
you
hit
a
package
json,
where
I
think
you
were
saying
like
find
me
a
file
from
where
I
am
in
the
directory.
C
E
C
C
B
Is
this
that
thing
that
ethan
harwood
was
talking
about.
B
I
think
you
were
talking
with
him
about
it
on
twitter.
It
was
some
file
system-related
thing
as
well.
So
I
thought
maybe
that's
what
you
were
referring
to,
but.
C
C
Fs
hooks
would
be
like
just
being
able
to
react
to
to
as
files
change
on
disk
or
like.
D
No
being
able
to
enter
intercept
the
fs
modules
behavior.
D
Yeah
so
yarn
and
electron
and
web
package
well
now
called
bundled
exchanges.
D
I
think
all
are
three
forms
of
run
times
which
are
able
to
have
archives
load
things
directly,
so
they're,
not
real
files
on
disk
they're
out
of
a
zip
or
they're
out
of
an
asar
file,
an
electron's
case,
but
they
they
are
currently
modifying
the
fs
module
in
node
core
and
that's
not
working
so
great
with
es
modules,
because
the
s
modules,
don't
let
you
patch
in
the
same
way.
C
That
makes
sense
I
understand
so
now.
Instead
of
you
like
overrating
the
global,
you
would
have
like
some
blessed
injection
you'd
have
some
blessed,
like
extension,
points
for
different.
D
C
That's
not
currently
on
our
agenda.
The
thing
I
wanted
to
bring
up
was
a
breaking
change.
C
And
I
don't
want
to
leave
them
hanging
on
their
issue,
like,
I
guess,
the
the
underlying,
and
this
came
up
when
we
when
we
built
mcder
like
when
we
built
the
recursive
version
of
victory
years
ago.
C
That
there's
like
this
question
of
whether
it
always
returns,
is
what
its
return
value
should
be,
the
the
deepest
directory
that
had
to
be
created,
or
just
the
deepest
directory,
I
think,
was,
and
I
think
the
behavior
is
the
deepest
directory
that
it
actually
had
to
create,
which
is
kind
of,
I
think,
useful,
because
it
tells
you
what
happened
like
you
know.
It
provides
some
information
about
what
actually
happened
so
and
I
think
we
actually
implemented
it
the
opposite
way
initially,
and
then
someone
who
used
mick
victor
asked
for
us
to
implement
it.
C
C
C
Tyranny
seems
to
have
a
pretty
strong
opinion,
but
I'm
unfortunately
it's
it's
a
feature,
that's
been
out
for
years
and
it
seems
like
a
risky
breaking
change.
So
I
don't
know
what
a
good
compromise
is
like.
I
don't
like
the
idea
of
switching
the
vet,
the
value
like
just
because
it's
a
breaking
change
for
the
community.
E
The
what
the
separate
fs
call
but
the
return
value
of
the
of
the
larger
thing
you
can
just
get
with
fs.resolve
or
I
think
it's
resolved.
C
I
mean
in
their
sample
code,
they
are
calling
resolve
themselves
so
that
they
have
the
fully
resolved
I'm
just
reading
through
their
example
code.
C
I
don't
need
to
solve
it
here,
maybe
I'll,
like.
Maybe
the
action
item
is
reach
out
to
tierney
and
just
see
if
there's
a
if
there's
any
compromise
at
all,
I'm
a
little
worried
about
I'd
be
more
worried
that
there's
some
implementations
in
the
while,
because
it's
so
it
would
be
hard
to
notice
this
breaking
change,
because
just
suddenly
your
app,
maybe
you've
written
an
app
that
relies
on
this
providing
information
about
which
directory
was
actually
created,
and
now
it's
just
a
slightly
different
string.
C
No,
no
sorry,
I
was
just
re-reading.
Sorry,
I
was
very
slowly
re-writing.
I
was
re-reading
the
thread
yeah.
I
think,
just
when
I
reach
I'll
reach
out
to
to
hear
any
explain,
explain
our
point
of
view
because
I
think
it's
a
risky.
C
I
think
it's
a
risky
breaking
change
to
be
honest,
so
different
story,
if
we
hadn't
landed
it
like
three
or
four
years
ago,
like
I
don't
care.
D
D
D
D
D
C
Yeah,
that's
kind
of
what
I
was
figuring
like.
I
thought
there
was
probably
some
nice
kind
of
elegant
elegance
that
that
was
annoying
that
that
tyranny
wasn't
getting.
So
that's
why
I
was
trying
to
be
at
least
open
to
the
recommendation
if
we
can
figure
out
a
way
to
do
it
without
like
breaking
the
world.
C
So,
okay,
I
mean
I
don't
know
it
sounds
like
we
talked
this
one
through
pretty
well
reach
out
to
your
knees,
make
sure
see
if
the
option
make
sure
the
option
isn't
you
know
a
dead
end
like
see
if
there's
a
compromise.
C
B
I
have
been
trying
and
failing
to
find
the
thing
that
I
was
thinking
about
so
I'll
maybe
keep
looking
and
if
I
do
I'll
post
it
on
slack
or
whatever
yeah.
C
C
I
kind
of
I'm
excited
about
doing
something
like
I
don't
know
if
this
is
a
good
idea
or
not,
but
I
continue
to
think
I
continue
to
think
it's
a
good
idea
to
do
something
like
mpx
for
node,
where
npx
is
a
bad
example,
but
something
like
python's
thing,
where
you
can
just
run
some
of
the
standard
modules
with
a
flag.
So
so,
like
you
could,
if
you
wanted
to
do
rimraf,
you
can
just
run
our
fs
module
dot.
C
Rimder
recursive
like
be
able
to
pass
some
flags
into
some
common
fs
operations,
because
then
your
npm
script
could
just
be
your
npm
script
could
literally
get
rim
raft
by
just
running
the
fs,
remove
recursive
or
it
could
get
copy
by
just
running
the
fs
recursive
copy
or
like
make
a
directory
with
with
just
the
fs.
Mr.
C
Exactly
because
people
do
that
today
and
they're
we've
talked
about
this
before
so
I'm
like
bringing
it
up
again
partially
to
nerd
snipe
daniel,
I
think,
but
but
but
basically
like
everyone
in
their
package,
jsons
will
frequently
pull
in
rimraf
or
say
the
cpr
module
or
like
mcderp,
so
that
they
can
just
have
it
as
part
of
their
mpm
scripts,
which
do
some
sort
of
bring
up
and
tear
down
between
tests
or
whatever.
We
can
totally
do
that
with
node
itself.
B
Can
do
that
right
if
you
do
like
node
e
fs,
dot,
rm.
D
Oh
yeah,
but
that's
really
frictiony
against
shell
command
line
argument.
Styles
right.
C
B
C
B
All
right,
let's
wrap
it
up,
then
thanks
for
coming.
C
Everyone,
we
should
probably
say:
oh,
do
we
want
when's
the
next
one
because
we'll
be
basically
christmas
right,
so
we
should
probably
mention
we're,
probably
not
going
to
have
the
next
meeting.
Oh.
C
B
Yeah,
okay,
well,
that'll
get
that'll,
get
the
issue
and
everything
I'll
get
scheduled
and
whatever
like
next
week,
I
think
automatically.
So
we
can
just
reply
on
there
and
say
like
we're
skipping
this
one.