►
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
A
We
are
live
on
youtube,
hey
everyone.
This
is
the
node
tooling
meeting
for
may
13th
2022,
I
guess,
let's
get
to
the
agenda
here-
sounds
good!
Sorry
go
for
it
all
right,
yeah.
So
the
first
thing
on
the
agenda
actually
is
about
regular
failures
in
test
fs
rmd
recursive
on
windows.
I
could
have
sworn
we
fixed
that
good
question.
I'm
gonna
open
up
the
issue
here,
because
this
is
an
old
one
or
maybe
pour
a
new
one.
Who
knows.
A
Oh,
I
see
we
did
make
some
fixes
and
I
don't
think
they
fixed
the
issue.
A
Okay,
I
didn't
even
realize
that
was
still
happening
so
good
to
know.
I
guess
yeah,
like
the
last
comment
here
on
this
issue
is
from
ben
from
february
21st.
A
So
I
guess
it
does.
He
had
a
suggestion
there
of
something
we
could
do
serializing
those
tests,
even
if
that
just
makes
it
easier
to
figure
out
where
the
failure
is
happening.
A
I,
like
I
said
I
didn't
realize
that
issue
was
still
open,
so
I
might
I
might
say
we
can
take
a
look
at
that
other
than
that.
Next,
on
the
list
we've
got
recursive
reader,
which
is
very
exciting.
I'm
also
not
super
up
on.
What's
going
on
with
this,
has
anyone
else
been
following
this
recently?
A
Yeah
I
mean
it
seems,
like
seems
like
this
is
close,
but
one
kind
of
sticking
point.
A
A
Yeah
yeah,
a
lot
of
this
stuff
is
related
to
him.
Unfortunately,
like
the
next
thing
on
the
agenda
is
recursive
copy.
I
don't
know
what
the
update
is
on
that
one
either
it
is
in
node.
I
think
it's
still
listed
as
experimental.
I
think
that
was
probably
why
it's
still
in
the
agenda.
D
D
Well-
and
let
me
let
me
say:
okay,
I
see
something
here,
you
know
if
we
wanted
to
waste
a
little
bit
of
time
until
ben
joins
us.
You
know.
One
thing
I
was
thinking
is
collab
summit.
You
know
if
we
want
to
take
a
few
minutes
to
talk
about
what
we
might
want
to
do
there
or
there's
issue
140,
which
has
a
link
to
a
discussion.
I
guess
maybe
I
should
just
share
the
link
to
the.
A
Yeah,
there's
really
anything
there.
I
agree,
though,
if
we
all
are
gonna
be
in
austin
together,
we
should
definitely
do
the
tooling
group
session.
D
And
then
you
know,
I
know,
we've
we've
talked
in
the
past
about
you
know
road
map
thoughts.
So
that's
certainly
something
I
wonder
if
we
should
have
like
you
know,
maybe
a
a
brief
sort
of
enablement
session,
for
maybe
any
anybody
come
new
coming
to
the
space.
If
they're
joining
us,
we
could
spend
you
know
a
half
hour
or
so
talking
about
what
we
do
and
if
folks
want
to
get
involved.
D
We
can
spend
some
time
talking
about
argument
parsing,
maybe
any
other
things
that
are
sort
of
lingering
and
then
talk
about
future
future
work.
Yeah.
A
I
think
so
yeah
yeah
yeah,
you
got
two
good
ideas
there.
I
would
say
one
like
anything.
We
can
do
to
get
any
new
people
interested
would
be
great.
I
think,
and
then
two
yeah
just
talking
about
like
what
do
we
wanna
do
next,
because
really
a
lot
of
our
road
map
was
like
a
bunch
of
the
recursive
file
system,
stuff
argument,
parsing
a
few
things
like
that
and
if
we're
for
hopefully
close
to
to
wrapping
up
a
lot
of
that
stuff,
what
do
we
do
next.
A
Okay,
those
are
good
ideas
should
we,
I
guess
I
started
this
issue
in
the
tooling
group.
I
can
maybe
make
some
notes
here
like
I've
got
something
started
so
I've.
A
E
D
Yeah
yeah,
if
we
were
I'm
looking
at
the
list,
I
got
here
get
new
people
interested
what
we've
done
in
the
past
year,
current
state
of
affairs
and
then
what
do
we
want
to
do?
Next
include
discussion
of
include
discussion
of
the
scope
of
the
work
of
this
group.
D
Okay,
yeah
I
mean
we
could
we
could
maybe
split
that
into
two
or
three.
You
know.
There's
really
aren't
constraints
on
how
we
want
to
approach
this,
so
whatever
we
think
makes
sense,
I
could
see
like
getting
new
people
interested.
What
have
we
done
in
the
past
year,
the
current
state
of
affairs
that
could
be
kind
of
one
session.
You
know
maybe
an
hour
or
two
whatever
be
adequate
and
then
maybe
like
two
hours
for
what
we
want
to
do
next.
D
A
Could
work
we
could
just
cover
that
in
current
state
of
affairs
or
have
an
informal
session
at
the
bar,
where
we
all
just
yell
at
each
other,
about
our
arguments
for
a
couple
hours
that
will
happen
regardless
yeah.
A
A
A
Also,
how
does
this
stuff
get
scheduled?
Did
we
just
decide
when
we're
gonna
have
the
meetings.
D
It's
a
good
question.
I
will
have
to
work
with
walid
to
see
what
you
know
what's
what's
taking
shape
but
as
far
as
I
know
there
there
are
some
sessions
proposed,
but
I
don't
think
that
we're
maxed
out
on
what
they
were
expecting
to
have
for
node.
So.
A
Okay,
I
just
I
I
think
tentatively
for
me,
it's
looking
like
I'll
be
around
all
day
thursday
for
the
collaborators
summit,
and
only
for
a
little
bit
in
the
morning
on
friday,
okay,.
D
I'll
say
would
prefer
thursday,
okay,
okay,
I'm
gonna,
say
disgust
in
today's
meeting.
Okay,
so
I'm
gonna
comment
there.
Y'all
can
take
a
look
at
this.
A
Issue,
can
you
comment
that,
on
our
issue
or
on
the
the
tooling
issue.
D
Yeah
you
shared
the
link
again:
okay,
great
yeah
and
then
I'll.
I
can
take
the
action
item
to
take
this
to
walid
or
propose
it
in
the
summit
repo
and
go
from
there.
D
A
All
right,
yeah
that
didn't
that
did
not
kill
nearly
enough
time.
We're.
D
D
I
don't,
but
one
thing
that
comes
to
mind
for
me
is:
I
am
a
mentor
in
an
internal
like
new
contributor
sort
of
sessions
that
we
do
kind
of
ongoing,
and
so
I
have
a
bunch
of
people
that
are
interested
in
getting
involved
in
node.
I
always
recommend
folks
to
find
like
a
group
like
this,
so
that
they
can
have
some
support
and
finding
some
issues
and
getting
some
guidance
in
the
work.
D
So
I
I
may
be
able
to
bring
a
couple
of
folks
along
to
the
tooling
work
in
the
coming
weeks.
We'll
see.
D
E
So
some
of
these
wrap
essentially
existing
fs
functionality.
Primarily,
I
think
it
has
to
deal
with
like
owner,
like
ownership
and
a
few
nuances
of
what
our
own
needs
are
with
fs,
but
might
be
something
in
here
that
we
can
pull
back
in
to
core
in
the
future,
so
just
want
to
bring
out
this
potential.
Like
next
work
item.
A
E
A
D
E
Yeah
we've
got
some
exciting
stuff
that
we
can
share.
I
mean
we've,
obviously
been
working
on
a
bunch
of
features
which
are
aligned
with,
like
other
package
managers
like
the
isolate
mode
rfc,
which
has
been
like
a
year
and
a
half
in
the
making
and
his
collaboration
with
folks
from
microsoft,
365.,
so
it'll
probably
be
like
coming
up
on
like
two
years
by
the
time
we
actually
land
it.
But
you
know
that's
a
different
reification
mode
that
other
than
sort
of
the
whole
hoisted
strategy
which
we've
traditionally
supported.
E
We
also
have
like
a
legacy
like
non-hoisted
mode
in
npm,
but
this
sort
of,
like
you,
know,
unlocks
the
same
kind
of
capabilities
like
package
managers
like
pmpm
have
and
rnv2,
and
three
I
think,
have
like
configurable
reification
modes
as
well,
so
we're
hoping
to
be
able
to
demo
that
for
folks
at
the
conference
and
and
then
also
we've
got
a
few
other
things
that
we've
been
working
on
command,
specific
configuration
is,
is
one
of
the
things
we're
expecting
at
right
now,
which
we
think
is
really
really
kind
of
important.
E
I
think
going
forward.
It's
been
something
we've
been
asked
for
for
a
long
time
and
then
also
empty
of
query.
Actually
is
like
this,
this
very
large
rfc
that
I've
put
together
with
the
help
of
roy
and
some
other
folks
on
our
team,
and
it's
it's
probably
a
command.
E
You
didn't
know
you
need,
but
it's
gonna
be,
I
think,
really
important
going
forward
just
getting
insights
out
of
like
your
your
dependencies
right
so
come
up
with
like
a
using
css
as
sort
of
like
a
baseline,
like
we've
come
up
with,
like
a
selector
syntax.
E
That,
like
feels
very
natural
to
to
write
like
a
selector
query
and
be
able
to
then
get
get
dependencies
back
from
what's
inside
your
project
and
then
essentially
iterate
over
those
to
do
actions
is
kind
of
a
hope,
so
we're
hoping
that
we
can
potentially
demo
that
by
the
conference,
but
I'm
sure
roy
is
shaking
in
his
boots,
because
I
think
it's
on
his
his
radar
to
pick
up
soon.
E
A
Oh
yeah,
that
was
that
was
like
the
blocker
for
us
switching
to
npm,
whereas
now
like
we,
we
actually
a
couple
weeks
ago,
we
had
like
the
the
metrics
like
crossed
where,
like
yarn,
usage,
dropped
and
npm
usage
eclipsed
it.
So
we're
well
on
our
way.
E
E
E
I
don't
think
there's
like
anything
else,
but
we're
hoping
that
a
few
folks
are
going
to
come
to
that
working
session
if,
if
it
gets
approved,
to
have
a
npm
collab
summit
session
with
some
new
ideas,
because
I
know
a
few
things
have
been
flowing
around
in
terms
of
like
pipelines
like.
I
know,
turbo
repo
and
lag
have
like
sort
of
introduced
like
the
idea
of
like
dependent
like
build
dependencies,
and
that
kind
of
was
a
thing
in
in
learn
as
well.
E
So
we're
trying
to
support
those
types
of
use
cases
going
forward
as
well,
so
yeah,
just
essentially
trying
to
take
what
we're
seeing
as
trends
in
the
ecosystem
and
pulling
that
back
into
core
the
core
experience.
E
E
Yeah,
we
just
need
to
get
blarne,
I
think,
into
core
somehow.
A
D
A
D
A
D
You
don't
need
to
be
I'm
just
adding
facilitators
to
the
issue
here
and
you
didn't
come.
I
mean
feel
feel
free
to
add.
A
Me
there
goes
your
inbox,
never
never
had
occasion
to
be
in
there
before.
Okay
ben
is
here
hello,.
B
A
B
A
A
B
F
Yeah,
I
don't
think
I'm
gonna
go,
I
don't
know.
D
Summit
with
it
yeah,
which
is
something
we've
been
talking
about
too-
I'm
about
to
create
two
session
proposals
for
the
tooling
group
and
what
we
had
discussed
while
we
were
waiting
for
you
is
the
first
session
would
be
like
intro
and
get
new
people
interested
look
at
what
we've
done
in
the
past
year
and
then
like
sort
of
a
current
state
of
affairs,
including
like
discuss,
parse,
args
or
any
other
work
that
is
currently
active
and
then
the
second
session
would
be.
D
What
do
we
want
to
do
next
and
and,
as
darcy
suggested,
include
the
discussion
of
the
scope
of
the
work
of
this
group.
But
I
know
we've
talked
about
sort
of
road
map
ideas
as
well,
so
that
was
sort
of
a
future
looking
session.
Gotcha
makes
sense.
A
Into
the
agenda
yeah,
I
was
gonna
say
now
that
now
that
you're
here
the
couple
things
that
you
know
you
were
kind
of
the
last
person
to
comment
on.
We
looked
at
some
of
these
issues.
One
was
around
recursive
reader,
any
any
contacts
you
can
share
on
that.
One.
F
F
F
F
The
tldr
is
the
codes
kind
of
challenging
to
work
in.
So
I
think
that
they're
trying
to
figure
out
how
to
do
an
asynchronous
implementation
of
the.
F
A
F
Yes,
I
think
that
we
could,
we
could
say
it's
not
experimental,
we
could.
We
could
remove
the
experimental
from
the
docks,
perhaps
see
if
there's
any
bugs
open
for
it.
A
Yeah
that
sounds
reasonable
yeah.
I
think.
F
A
F
Yeah
he
said
it's
an
edge
like
a
an
edge
case
that
requires
the
being
able
to
handle
a
buffer
in
some
of
these
edge
cases,
if
I
recall
and
and
he'd,
be
open
to
removing
experimental
since
there's
not
too
many
bugs
open
for
it,
but
we
should
just
be
cognizant
of
if
we
then
have
issues
with
people
in
japan
or
china
who
are
using
complicated
characters,
character
encodings,
because
because
that's
what
the
buffer
is
for,
it's
for
non-english
character
sets
but
unicode
works.
So
I'm
not
quite
sure
it
must.
Just.
F
A
All
right
and
then
yeah
I
mean
last
but
not
least,
on
the
agenda-
was
argument
parsing,
so
yeah
ben,
I'm
sure
you've
got
stuff.
On
your
mind.
There
darcy
you
mentioned
in
slack
that
you
also
maybe
had
an
update
or
something
who
wants
to
go.
First.
F
So
I
mean
there's,
there's
the
pr
that
has
a
lot
of
that
has
approval,
which
is
just
like
bringing
over
the
community
work.
That's
happened
on
parsegs
and
the
parsers
repo,
pretty
much
just
with
like
community
driven
consensus
on
a
bunch
of
this
stuff
so
like
that
was
how
we
landed
on
like
strict
mode
was
there
was
a
few
people
arguing
for
it
kind
of
maybe
interesting,
and
that's
the
reason
that
so
first
of
all,
I'm
advocating
that
that
pr
is
that
we
land
is
experimental.
F
So
so,
with
with
like
an
openness
to
change
the
api
surface
so
like
I
don't
think
it
has
to
be
the
final
api
surface
at
all,
and
I
think
if
we
find
it
clunky
or
you
know
you
know,
I
want
to
say
there
is-
continues
to
be
room
to
extend
extending
it
to
since
there's
some
points
of
contention.
You've
brought
up
darcy,
but
the
other
interesting
thing
I
think,
is
there's
currently
a
vote
happening
with
the
tsc
just
to
say
where
the
the
parsers
method
should
live,
which
is
why
we
can't
land
that
pr.
F
A
Just
commented
in
the
chat
that
apparently
the
vote
happened.
What's
the
result
process
seems
to
be
the
consensus.
E
F
E
No,
it's
encrypted
it's
an
encrypted,
I
think
not
encrypted,
but
it's
obfuscated
vote
voting
mechanism.
I
think
it
could
be
wrong.
They
use
that
funny
enough.
They
use
mpx
and
gh.
So.
F
A
You
know
the
best
argument
I
heard
was
stuff
in
util
is
meant
to
be
like
really
shimmable
in
a
browser
so
because
of
that
process,
kind
of
seems
to
make
sense.
To
me,
I
mean.
E
E
F
E
No,
I
think
I
I
misspoke
there.
I
think
I
don't
know
yeah
the.
F
Yeah,
so
we
should
have
consensus
soon,
yeah,
so
I
don't
know
where.
Where
do
we
stand?
Where
do
you
stand?
Darcy
like,
I
feel
like
there's
not
much
happening
on
the
repo
right
now.
I
feel
like
the
contributors
kind
of
for
better
or
worse,
got
to
a
point
of
consensus.
I
know
it's
not
as
simple
as
you
were
hoping
so
like.
E
Yeah
I
mean
like
I'm
I'm
open
to.
I
don't
want
to
stop
progress
like
I
know
I
wasn't
able
to
contribute
over
the
last
several
several
months
like
like
probably
six
six
plus
months
so
I
know
was
a
lot
of
discussion
and
like
a
lot
of
work
that
went
into
the
implementation
as
it
is
now,
so
I
don't
want
to
like
like
come
in
and
voice
a
strong
opinion
against
it.
I
think
it
is
weird
that
we've
gone
towards
strict
by
default
and
like
requiring
configuration.
E
If,
if
we're
talking
about
something,
that's
going
to
error
like
without
options
by
default,
because
I'm
for
sure
it's
strict
is
true-
is
the
default
right
now
in
the
way
that
it's
set
up
yeah,
and
I
you
know
the
approach
I've
been
taking
with,
like
the
the
separate
library
that
I
I've
been
sort
of
using
as
a
way
to
sort
of
ideate
and
think
about
you
know
on
the
most
minimal
set
is
like.
E
Can
we
get
away
with
like
almost
no
options
and
like
continue
to
pair
back
and
and
in
fact
after
having
a
conversation
with
roy,
like
it
might
be
impossible
to
remove
the
last
few
options
that
I've
got
in
the
min
arcs
library
there?
So
essentially,
what
you'd
be
returning
is
like
a
and
and
roy
put
this
adequately.
I
think,
like
a
is
like
a
glorified
split
which
gives
you
extra
information
about.
E
You
know
the
parse
and
provides
enough
detail
that
you
could
essentially
do
unique
things
with
it
right
or
at
least
with
main
arcs-
the
way
that
god
now
does
have
a
rv
array
that
that
has
index
like
indices,
that
help
you
map
back
to
the
original,
like
array
that
you
provided
or
a
process
rv
array.
So
I
saw
that
as
a
comment
in
the
parseart's.
E
You
know
discussion
and
I
thought
that
that
was
really
important,
especially
if
we're
talking
about
something
that's
going
to
be
extensible
but
yeah
like
I
don't
want
to
be
a
blocker
like
if
you
want
to
keep
on
you
know
it's
an
argument,
parser
and
core,
like
I
think
everybody
will
celebrate.
E
So
I
I
just
don't
think
that,
in
terms
of
any
in
terms
of
any
like
function
that
I've
seen
like
I
being
able
to
run
it
and
have
it
do
something
by
default
without
having
to
configure
it,
I
feel
like
kind
of
like
harms
it's
like
it's
hard,
tough.
I
think
for
like
a
newcomer
to
like
use
and
not
be
tripping
over
their
feet.
I
don't
know
it's.
E
Like
you
sort
of
have
to
know
exactly
what
keys
and
configuration
to
to
use
before
you
can
play
with
like
the
functionality
so.
F
I
think
that's
a
oh,
my
muted,
I
think
that's
a
fair
critique
like
I.
I
don't
like.
I
think
just
trying
to.
I
also
don't
disagree
with
the
direction
that
parsers
went
in
either,
like
I
think
it's
it
had
a
bunch
of
community
input.
People
were
pushing
towards
like
a
slightly
more,
I
guess
explicit
configuration.
F
This
is
what
it
comes
down
to
and
I
like
minnows
like
I've,
certainly
used
minimus
to
write
clies,
where,
like
I
literally
just
run,
minimus
give
it
process.rt,
run,
minimist
and
and
then
do
stuff
with
whatever's
pumped
out
from
it
right
also
having
maintained
yards
for
years
and
years
and
years.
I
know
how
just
making
assumptions
can
like
actually
ultimately
be
a
pain
in
the
neck,
like
you
like,
like
you're,
it's
hard
to
make
an
assumption.
F
F
Let's
go
with
this
document
that
at
least
has
existed
forever
and
give
some
guidance
about
how
you
should
write.
Unix
command
line
applications.
So
that
was
my
my
two
cents
I
would
say
without
I
would
argue,
not
to
add
any
more
features
onto
parse
rx,
if
at
all
possible
and
just
do
exactly
what
you're
saying
and
add
the
parser
positions.
So
we
know
exactly
what
point
in
the
string.
E
E
Tooling,
like
the
current
tooling,
I
wanted
to
have
a
standard
like
like
core
piece
that
could
be
used
by
yards
or
minimus
or
anybody
else
and
like
if
we
had
somehow
a
way
to
find
all
the
commonalities
or
to
introduce
something
that
like
gets.
You
like,
80
percent
of
the
way
there
and
and
at
least
I've
completely
paired
back
on
usage
like
like.
F
Or
maybe
there's
a
compromise
there,
which
is
like?
I
honestly
I
do
like
the
kind
of
extended
functionality
of
sparse
args,
but
to
your
point,
maybe
we
make
a
strong
case
to
put
it
in
util
and
then
maybe
there
is
an
underlying
library
which
literally
just
parses
the
string
and
puts
the
positions
that
puts
the
positions
that
the
parsing
happened
and
then
you
could
use
that
you
could
use
that
directly
or
you
could
use
the
thing
that's
been
built
on
top
of
it.
F
E
That
feels
like
totally
like
a
path
forward
for
me
because,
like
I
feel
like
there's
like
a
set
of
users
that
want
to
configure,
I
don't
know
like
there's
the
minimal
set
and
and
then
there's
like
the
like,
extended
use
cases
etc.
Like
I
don't
know,
I
just
always
thought
that
as
soon
as
you
get
into
usage
and
like
throwing,
then
you
get
into
like
formatting
of
that
messaging,
right,
formatting
and
that's
like
so,
and
and
validation
as
well,
which
can
be
like
takes
you
off
into
like
a
whole
another.
F
I
I
agree-
and
I
think
we
have
to
like-
I
think
we
have
to
stand
our
ground
on
the
experimental
status
of
it
like,
like,
I
think,
like
being
like.
This
is
experimental
like
let's
see
how
the
community
takes
to
it,
but
I
actually
really
like
this.
I
like
the
idea
of
the
even
more
like
more
bare
bones
than
we
even
initially
suggested
for
parse
args,
the
like
positional
parsing
information.
I
think
that
could
be
really
interesting
and
then,
if
we,
I
think,
don't
throw
it
the
higher
level
thing.
E
E
D
I
guess
I
just
sort
of
generally
feel
good
about
the
fact
that
more
people
got
involved
and
it
sort
of
took
a
natural
shape.
I
suppose,
because
I
didn't
really
have
a
horse
in
the
race.
You
know
in
terms
of
any
of
the
the
details
really,
I
just
wanted
to
help
and
build
something
so.
A
Ian
yeah,
I
I'd,
have
to
think
more
about
the
kind
of
the
two
I
mean
I.
A
Like
the
I
was
kind
of
in
favor
of
as
simple
as
possible,
initially,
and
that
is
a
little
that
was
one
of
the
things
with
the
current
implementation
again
with
the
strict
mode
and
needing
to
like
provide
a
bunch
of
options
to
use
it
that
I
feel
we
got
a
little
bit
away
from,
but
yeah.
I
don't
have
like
a
strong
opinion
at
this
point
in
time.
A
I'd
actually
like
to
use
the
two
of
them
a
little
bit
more
as
well
to
kind
of
get
a
little
bit
better
of
an
idea
from
you
know
actually
trying
to
write
like
a
cli
with
the
two
of
them,
which
I
actually
do
have
to
do
for
the
thing
that
I'm
doing
for
openjs
world,
so
maybe
that'll
be
an
opportunity
for
me
to
get
a
little
bit
hands-on
with
both
of
them
and
and
form
a
stronger
opinion.
At
that
point,.
F
All
right
is
anyone
opposed
to
going
forward
with
getting
an
experimental
version
merged
and
see
how
the
tires
meet
the
road
with
users
and
with
the
understanding
it
might
be
a
little
rocky.
If
we
start
to
change
our
minds
a
little
bit
over
the
next
year,
her
courage
decided
to
go
in
a
slightly
different
direction.
It's
like.
D
Yeah,
no,
I
I'm
I'm
all
for
it.
I
wonder
too.
Let
me
just
ask
this
as
a
group
generally,
I
agree
with
what
ian
said
too,
but
if
we
wanted
to,
if
we
landed
as
as
experimental
and
and
people
had
differing
opinions
once
you
know
it
was
maybe
being
used,
do
you
do
you
think
there'd
be
any
problem
in
backing
off
a
bit
and
going
back
towards
minimal?
If
that
was
what
you
know,
consensus
was
I
wouldn't
think
so.
Right
I
mean
that's,
why
it's
experimental.
That's.
F
What
we
call
experimental
it
might
mean
it
would
be
worth
here's
where
I'm
of
two
minds
I
mean
we
should
put
a
runtime
experimental
warning
in
just
to
make
it
like
really
just
make
it
clear.
It's
it's
a
toy
that
might
change
a
little
bit
yeah
where
that
kind
of
stinks
is
it
makes
it
hard
to
write
a
like.
It
makes
it
hard
to
write
a
command
line.
App
people
would
actually
use
because
you're
going
to
get
a
runtime
warning
every
time
you
try
to
run
your
command
line.
F
F
We'd
have
to
roll
it
out
in
a
future
release.
We
would
take
out
the
runtime
deprecation.
It's
just
like
it's
relatively
easy
to
do.
I
think
you
just
wrap
the
method
in
a
runtime
deprecation
and
then
I
started
I'm
using
deprecation,
experimental,
runtime,
experimental
and
but
it
does
mean
we
could
maybe
land
as
experimental
like
it's
very
clear,
like
someone
can't
pick
it
up
and
not
know,
it's
experimental
yeah
mean
that
gives
us
room
to
poke
at
it,
see
if
people
like
it,
but.
F
D
A
Yeah
from
my
point
of
view,
I
worry
a
little
bit
about
releasing
something
a
little
more
full-featured
and
then
like
pairing
that
back,
but
I
think
on
the
other
hand,
if
we
don't
kind
of
land
this
soon
I
I
I
don't
know
I
don't
wanna.
I
don't
wanna
like
go
back
to
the
drawing
board
on
this
again
because
I
feel
like
that
might
never
happen.
You
know.
F
I
think,
like,
I
think,
like
the
reason
that
the
configuration
is
needed
is
partially
because
we're
starting
strict
mode
by
default
like
like.
If
we
said,
if
we
said
strict
mode,
was
false
by
default,
I
think
it
would
for
for
some
usage.
You
could
just
call
the
method
directly
on
on
your
arguments
and
we
could
get
back.
So
you
know
what
I
mean
like
it's
not
like.
I
don't
know
if
we
have
to
completely
revisit
the
drawing
board
to
get
to
a
simpler
design.
F
Really-
and
personally,
I
do
like
the
idea
that
we
can
differentiate
between
booleans
and
things
that
take
values.
So
it's
like
that's
useful,
like
like
I
I
would
say
if
we
were
going
to
go
simple.
I'd
want
to
go
the
simplest
possible
where
it's
literally
just
a
parser.
That
gives
you
positions
and
no
no
user
would
ever
use
it.
They'd
only
ever
use
it
to
build
it.
Their
own
command
line,
argument,
parser
right
so
so
parsers
becomes
a
reference.
Implementation
of
how
you
would
build
on
top
of
this
parser
is
what
you're.
F
I
think
that
and
and
the
idea
that,
if,
if
when
we
do
introduce
this
idea
of
like
ast
based
parsing
positions,
we
expose
that
as
well
and
then
darcy
could
potentially
use
that
to
underpin
many
args.
If
you,
if
you
you
know,
if
you
wanted
to
just
use
an
ast
based
one,
you
can
use
that
to
build
yards,
you
can
use
that
to
build
many
many
arcs,
whatever
parse
args
is
a
tiny
little
argument.
Parser
built
on
that
reference
implementation.
I
think
that's
what
I'm
arguing.
A
Okay,
yeah,
that
that
seems
like
an
okay
plan
to
me,
I
am
a
little
bit
on
the
fence
too,
on
the
runtime
experimental
warning
I
I
would
be
in
favor
of
making
it
more
experimental,
but
then,
like
you,
said,
if
you're,
if
every
time
you
use
it,
you
get
that
warning
in
your
terminal,
like
probably
no
one's
gonna,
actually
build
like
a
real
thing
with
it.
F
Maybe
we
don't
have
to
as
long
as
like
it
sounds
like
a
pretty
good
plan,
which
is
we
try
to
slice
out
the
ast
part
from
the
parsers
part
and
the
sd
part?
Maybe
some
people
just
use
that
I
don't
know
what
the
naming
is,
but
to
me
that
makes
it
to
seem
less
likely
to
me
that
we-
I
guess
I'm
saying:
maybe
we
wouldn't.
Maybe
we
don't
need
to
do
runtime
experimental
if
the
main
change
might
be
strict
switches
to
false
by
default
from
true
to
default.
That's
not
a
massive
change.
Yeah.
A
I
guess
I
guess
it's
easier
than
going
the
other
way
too
yeah
yeah
yeah.
Okay,
I
mean
this
does
seem
like
a
reasonable
plan
to
me.
Then
I
think
we're
actually
over
time.
Oh
sorry,
yeah
and
I
have
an
another
meeting,
so
I
gotta
I
gotta
drop,
which
I
think
means
that
will
end
okay.
This
is
a
meeting
this
plan,
sound
okay
to
you,
darcy.
E
Yeah
I
mean
like
district
by
default
and
like
heavy
need
for
configuration,
was
like
a
problem.
I
think
the
the
thing
that
changed
my
mind
for
sure
is
just
like
the
idea
of
like
the
index
references
like
the
like
est
sort
of
portion
of
this.
I
think
it's
kind
of
like
where
I'm
driving
out
with
like
main
arts.
F
Okay,
cool
cool
cool,
we'll
work:
let's
talk
off
like
well
as
much
as
to
your
ability,
but
let's
get
kevin
looped
in
and
try
to
make
that
the
next
thing
we
start
doing
so
all
right.
Thank
you
so
much
sorry,
sorry
to
go
over
okay,
yeah
yeah.
Let's
continue
the
conversation
with
slack.