►
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
Yeah
I
think
for
now
maybe
just
leave
the
show
open.
Maybe
some
simple
to
do
comment.
Maybe
we
have
new
API.
Is
that,
like
kind
of
so
better
that
that
particular
use
case
in
a
different
way,
so
yeah
I
just
have
that
in
our
Rader
I
think
it
would
be
nice?
You
know,
yeah
I
would
like
to
just
solve
that
issue
right
like
having
a
better
way
from
Reed's
standard
input
just
using
core
without
that
I
user,
like
modules
right.
A
A
A
Don't
know
if
I
did
that
quite
right,
but
you
know
it's
not
as
simple
as
let
us
delete
stuff
from
the
cache,
because
we
need
to
know
what
exactly
to
delete
and
when
we're
gonna
do
something
like
that
and
without
a
a
proper
graph
of
some
sort.
Like
a
you
know,
a
module
graph
that
you
can
traverse.
We
don't
necessarily
know
what
that
is.
A
A
A
You
have
to
delete
the
one
thing
because
you
only
have
like
one
one
reference
to
it
and
so
I
don't
know
it
seems
like
we
can
work
around
that
I
I
would
hesitate
to
call
it
like,
like
a
requirement,
I
think
if
we
had
the
ability
to
just
remove
things,
just
some
ability
to
remove
things
from
that
cache.
I
think
that
would
be
like
that
would
be
progress
and
we
could
start
working
with
that.
A
A
B
A
C
There's
I've
been
kind
of
waiting
patiently
for
because
part
of
it
is
that
ease
as
we
can
apps
I
think
for
the
implementation
for
modules.
If
I
would
call
are
they
like
they?
You
can't
iterate
over
the
data
structure
that
the
models
are
and,
if
I
recall
the
way
it's
implemented
right
now,
but
there's
like
a
new
there's,
a
new
JavaScript
feature:
that's
in
the
pipeline.
That
would
actually
give
you
the
ability
to
observe
a
data
structure
like
this
and
actually
actually
delete
things
and
add
things
to
it,
using
a
different
method.
C
So
I've
been
waiting
for
that
for
the
source
map
stuff
because
it
means
I
could
like
load
and
unload
source
Maps
more
easily.
But
what
is
you.
C
Me
find
the
I'm
like
blinking
and
it's
random
ECMA
proposal,
and
it's
it's
not
like
stage
3.
Already,
though
you
you.
C
C
Weak
references,
I,
don't
know
if
this
is
directly
applicable,
but
when
I
was
implementing
source
maps
like
one
of
the
annoying
things
was,
it
wasn't
easy
to
remove
and
add
entries
to
a
Google
stock.
It
wasn't
easy
to
remove
entries
because
it
was
using
a
weak
map
and
there's
actually
a
way
to
get
around
this
with
with
weak
references
which
are
feature
that's
on
its
way
into
JavaScript.
C
C
How
do
I
find
a
chat
in
this
web
URL
here
where
this
is
so
this?
This
gives
you
more
Matt,
at
least
maybe
I'm.
Being
it's
tangent
until
about
the
saturation
is
pinning
when
this
lands.
It
will
give
me
more
flexibility
around
referencing
things
that
have
been
put
into
the
ESM
logical
cache,
because
it
can
actually
like
track
them
and
remove
them
as
needed,
so
they
can
build
kind
of
more
complicated
data
structures
on
top
of
it.
Maybe
we
can
use
a
similar
approach
for
figuring
out
if
things
need
to
be
loaded
or
unloaded.
A
C
A
C
It
was
I,
maybe
get
well
you've
got
the
you've
got
the
issue
on
github.
Was
it
that
you
shared
51
or
whatever
err?
Sorry,
yeah,
51
I
know
that's
shop,
that
around
to
a
few
more
people
and
see
like?
Is
there
any?
What
are
the
technical
reasons
why
we
can't
do
this
like?
Who
do
we
talk
to
you?
Probably
like
Bradley,
redneck
and
probably
guy
right,
yeah.
A
C
A
A
A
C
A
A
B
A
That's
my
fault,
and
it's
still
on
my
to-do
list,
and
so
I
will
do
that
and
so
then,
once
we
do,
that
I
think
would
the
next
step
be
to
probably
just
open
a
pull
request.
That
does
that
after
you
know,
maybe
a
next
meeting
I'll
say:
okay,
this
is
what
I
came
up
with
for
all
the
things
that
needs
to
track
and
run
it
by
the
group,
and
then
we
can
say
yeah
that
sounds
right
or
no,
it
needs
other
stuff
and
then
then
it
would
be
ready
for
a
pull
request.
I.
A
C
Dump
a
new
way
to
tech
this
and
then
have
three
ways
to
detect
exiting
in
an
app
like
you
know,
on
signals
upon
process,
not
exit
and
then
we'll
have
this
new
event.
I
just
want
to
make
sure
we've
documented
in
the
right
places
right,
cuz,
cuz,
it's
like
six
months
I
would
think
it's
actually
the
most
predominantly
I
advertised
one
for
people
who
have
this
use
case
right.
A
C
So,
if
you'd
like
did
this
midway
through
your
application,
loading
you
know
have
source
maps
for
50%
of
your
application
and
not
the
50%
that
floated
early
on
so
I
see
the
value
for
a
tool
like
mocha,
that's
instrumenting
an
app
but
then
I'm
like
well.
If,
if
you
only
need
to
kick
off
a
sub
process
like
as
you
bootstrap,
that's
like
adding
what
like
50
milliseconds
to
run
time
or
something
like?
Is
it
worth
the
confusion
of
having
this
new
API
method
for
50
milliseconds
or
a
hundred
milliseconds
of
performance
improvement?
A
It's
its
complexity
right
because
now
you
have
to
manage
the
sub
process
and
manage
the
signals
being
sent
from
you,
know
the
terminal
and
manage
this
standard,
I/o
options
and
all
sorts
of
things.
So
it's
just
like
I
want
to
avoid
that.
If
I
can-
and
you
know,
there's
lots
of
stuff,
you
can
use
a
node
shoot
yourself
in
the
foot
with
it.
C
Yeah
I
think
my
ultimate,
my
other,
like
architectural
concern,
was
like
random
program.
Food
turns
this
on
because
they
like
the
feature
and
now
three
layers
up
in
the
dependency
graph.
You
have
it
turned
on
in
mocha,
even
though
you
didn't
know
it
was
turned
on
like
it
like.
You
can
imagine
some
some
issues
or
it
would
be
weird.
A
C
C
Make
the
pr
for
discussion
if
you
want
to
what's
the
way
there
I
mean
yeah
I
mean
if
you
want
to
make
a
strong
case
for
it
like
like
it's
the
codes
there.
It
works
pretty
well
I
even
found
he
could
turn
it
on
in
the
same
file,
and
we
come
in
as
long
as
it's
synchronous.
It
will
actually
collect
it,
so
the
file
was
turned
on
on
yeah.
So
can.
C
Would
be
there's
an
API
for
doing
that,
so
you
could
potentially
I
mean
it
would
be
kind
of
hard
you
you
would
have
to
like.
You
could
delete
your
module
cache
and
then,
as
things
are
loaded,
you
would
get
stuff
cuz
that
because
it's
now
going
to
hit
the
module
cached
again.
But
you
can't
do
that
for
GSM
right
now,
because
ESM's
immutable.
A
C
A
C
I'm
like
right
in
the
middle
like
like
it's
not
like
a
strong,
no
or
a
strong,
yes,
but
the
thing
that
it
actually
enables
so
part
of.
Why
I
did
this
feature?
First,
was
it
meant
I
had
like
it
gives
me
a
little
more
tracking
I
built
out.
It
meant
I
built
it
to
cover
the
source
map
function,
a
little
more
with
with
making
it
kind
of
lazier.
So
it
happens
later
in
the
flow.
This
I
won't
get
into
details
as
to
why,
but
this
actually
made
it
easier
to
replace
that
custom.
C
You
know
if
you
get
as
an
exception
and
nodejs,
it
tries
to
show
you
the
line
of
code,
that
the
exception
happened
on
and
it
puts
like
a
little
arrow
underneath
it
and
says
your
exception
happened
here
and
on
line
22.
Does
anyone
know
what
I'm
talking
about
I
do
so
you
have
like
the
stack
trace
and
then
above
the
stack
trace.
You
have
like
the
kind
of
snippet
in
context
mm-hmm
with
source
Maps
right
now.
C
We
don't
rewrite
that
little
snippet
in
context,
because
it's
it's
done
in
C
and
it's
it's
done
as
C++,
and
it's
like
a
pain
and
there's
an
appendix
source
maps
aren't
exposed
to
C++
right
now,
because
the
stack
trace
is
actually
in
JavaScript
ran
long
story
short.
This
refactor
I
did
to
make
it
so
you
can
turn
on
source
Maps
later
in
the
flow
made
it
easy
to
replace
that
C++
stack
trace
with
one
in
JavaScript,
where
we
actually
show
the
where
the
exception
happened
in
your
original
source,
not
to
transpile
the
source.
C
A
So
that
is
that
line
that
that
little
thing
that,
as
he
said
as
in
C++
I
did
not
so
I
was
trying
to
write
this,
this
little
library
and
so
the
what
the
library
does
is.
It
is
essentially
like
an
exception
handler
for
intended
for
libraries,
and
so,
if
you
are
consuming
the
library
and
there's
something
in
and
an
exception
originates
in
that
library,
it
will
it'll
it'll
try
to
determine
if
it
happened
in
this
library,
but
B
it'll
go
and
we'll
show
information
to
the
user,
and
it
will
say:
hey
this
exception.
A
This
uncut
exception
handled
in
this
library
that
you're
using.
We
would
love
information
about
what
you
did
to
cause
this
exception,
because
we
want
to
handle
it
or
we
want
to
say:
oh
no,
that
isn't
our
fault
or
something,
but
at
minimum
we
should
know
about
it,
and
so
just
like
dumps
all
this
information
to
the
terminal.
It
gives
a
URL
and
you
click
on
the
URL
and
it
goes
in.
A
A
That's
the
line
that
will
show
right,
because
that's
like
the
last
place
that
the
exception
so
anyway,
but
I
didn't
want
to
do
that.
I
wanted
to
show
the
original
error,
but
that
little
that
little
idiot
line
was
getting
in
my
way
and
I
wanted
to
be
able
to
change
what
that
line
was
because
it
should
I
want
to
show
then
we're
in
context
that
happened
in
that
library,
not
in
my
repartee.
C
In
a
way
we
would
have
to
refactor
it
a
bit,
but
what
I
did
is
basically
I've.
Now
in
C++,
C++
knows
that
source
maps
have
been
turned
on
so
before
this
C++
didn't
have
a
clue.
What's
the
source
maps
where
now
it
at
least
knows
that
source
maps
are
turned
on
and
when
it
gets
to
that
line,
that's
about
to
add
that
weird
arrow
line
in
in
C++
it
just
says
if
source
maps
are
turned
on.
C
Don't
do
that
and
then,
when
you
get
to
your
prepare
stack
trace
step,
which
is
how
source
maps
are
applied,
which
you
could
override
yourself.
You
can
then
add
your
own
thing
at
the
top
of
the
source
map.
You're
preparing
are
sorry
at
the
top
of
the
stack
trace.
You're
preparing.
You
could
add
your
own
error
message.
Basically,
so,
basically,
what
I'm
suggesting
is
some
way
to
turn
off
that
that
annoying
one
line
in
C++
from
happening
so
that
you
can
instead
do
it
in
JavaScript
land.
With
your
prepare
stack
trace,
you.
A
C
A
C
Anyway,
so
that
aside,
maybe
this
is
a
baby
step
towards
what
you
want,
but
that
aside,
like
is
this
behavior
that
we
want
to
maintain
into
a
world
with
source
maps,
or
do
we
just
like
get
rid
of
that
entirely
and
call
it
done
like
you
know,
just
have
the
stacktrace,
don't
have
a
contextual
error?
How
much
value
does
that
contextual,
your
ad
versus
the
pain
and
implementing
it
I.
C
Actually
have
working,
yeah,
I
haven't
working,
but
I
was
frustrated
because
the
line
I
don't
know
I'm
so
frustrated
by
this
errors
file,
because
these
the
original
code
points
to
the
start
of
the
throw
and
when
you're
throwing
an
error
and
the
line
reported
by
the
stack
traces.
The
start
of
the
error
object,
not
the
start
of
the
throw
and
I'm
like
this
works
perfectly.
What
if
I
opened
up
a
pull
request?
The
first
thing
anyone's
gonna
say
is:
why
is
your
arrow
underneath
the
error
message?
C
Instead
of
underneath
the
throw
message
and
I'm
either
gonna
have
to
like
write
code
that
screws
around
trying
to
move
the
arrow
over
figure
out?
Why
I
have
a
different
position
in
the
stack
trace?
Then
C++
has
or
argue
that
it
doesn't
matter
because
who
cares
at
bureaus
under
the
error
or
the
throw
so
I'm
like
I'm,
already
picturing,
how
annoying
of
a
pull
request?
This
is
gonna,
be.
C
A
A
C
I
still
say,
like
I'd
love
to
still
work
on
it
with
Roy
in
and
you
or
anyone
else,
who's
interested
I
had
a
thought
and
I.
Don't
know
if
this
relates
to
argument.
Pricing
or
not,
but
I
was
thinking.
There's
like
a
type
of
there's,
a
type
of
thing.
People
use
a
rim
raft
for
McGuirk
for
in
know,
Jess
applications
which
is
in
there
NPM
scripts
to
create
a
folder
when
the
appropriate,
when
they're
test
run
or
something
or
to
remove
a
folder.
C
For
me,
I
was
having
the
stock
like
if
we
have
this
lovely
command
line
parser
in
node,
now
like
as
we've
built
one
like
it's
one
of
the
things
that
we
should
be
a
reach
goal
or
like
something
that
we
can
then
build
on
top
of
it.
The
ability
to
make
it
these
really
short
little
NPM
scripts,
that
instead
of
using
rim
Raph
to
remove
a
directory
after
testing
or
instead
of
victor,
to
create
a
directory,
as
you
run
your
init
script.
C
A
Right
I
just
put
in
the
chat
there,
Python
Jim,
simple
HTTP
server,
so
you
install
Python
and
it
comes
with
all
these
like
little
tools
right.
It
comes
with
these
built-in
modules
that
aren't
part
of
the
the
like
standard
library
but
they're,
one
of
them
just
like
simple
HTTP
server,
and
it's
just
it's
always
there
and
it
sits
in
the
system
and
you
can
go
into
any
directory
and
you
say
Python,
gem,
simple,
HP,
sorry
I
think
that's
what
it's
called
anyway
and
it
opens
a
port
and
it
serves
the
files
in
that
directory.
A
D
D
So
we
have
a
really
nice
to
be
able
to
like
you
could,
in
theory,
do
this
right
with
node
we
enough.
It's
like
if
you
like,
no
dashiki,
like
in
poorer,
remember,
was
just
a
little
awkward,
so
yeah
I
mean
I
mean
that
would
be
pretty
cool
I
like
the
idea,
but
maybe.
A
In
this
would
probably
make
more
sense
as
a
node
thing,
not
an
NPM
thing,
because
you
have
something
like
that
in
your
package.
Scripts
and
and
it's
no
longer
portable
other
package
manager,
see
lies
right
because
the
script
expects
something
like
Rimmer
have
to
be
there
and
om
cams,
like
oh
I,
didn't
make
this
available
to
all
these
scripts.
Just
to
save,
say
people
headaches
right,
but
if
the
other
ones
don't
do
it,
then
that's
more
headaches
for
everybody,
so
it
makes
more
sense
to
it.
C
A
A
A
C
And
I
think
like
as
we
maybe
things,
I,
don't
know
if
it's
like
I
would
say:
I
don't
know
if
it
exactly
fits
in
or
we've
been
talking
around
around
command-line
parsing,
but
maybe
it
does
maybe
we're
like
I
don't
know.
Maybe
maybe
it
helps
us
make
more
of
a
case
as
to
why
this
would
simple.
Command-Line
person
would
be
nice
and
know
that
you
can
say
well.
Look
we've
written
these
we've
made
Rim
Ralph
recursive.
C
A
C
C
C
So
I've
been
at
Google
working
on
this
release.
Please
thing
for
quite
some
time,
but
I
haven't
I,
don't
really
know
the
process
for
making
it
so
I
could
host
it
for
the
wider
community
I'm
like
the
github
app,
we
wrote,
but
it
turned
out
that
with
the
newer
types
of
github
actions,
you
can
do
this
in
a
github
action.
I
don't
need
to
host
anything
for
you
and
it
works
really
well,
so
you
can
create
a
actually
switched
yards
to
this.
You
create
this
really
minimal.
Get
have
action.
C
So
I
have
this
minimal
github
action
called
release.
Please
oh
wait:
crap,
maybe
I
didn't
or
I
left
some
craft
in
there.
It
looks
like
so
have
this
minimal
get
have
action
called
release,
please
that
listens
on
pushes
to
master
and
then
on
a
push
to
master.
It
runs
the
my
be
Co
release,
please
action
and
with
the
this
secret,
then
this
github
action
magically
creates
these.
C
These
release
PRS,
like
that.
Look
like
chor
release,
one
point:
4.0
these
get
updated
as
you
land
new
work
to
your
default
branch.
So
you
know
if
you
land
for
more
things
those
start
get
to
get
glommed
in
over
everything,
features
and
bug
fixes.
When
you
merge
this
thing,
it
results
in
a
release
being
created
for
you,
so
this
release
will
be
automatically
created
here.
C
Here's
like
more
my
last
releases
yeah.
So
that's
it
like
automates,
your
creation
of
github
releases
based
on
conventional,
commit
messages,
and
you
don't
have
to
run
service
for
it.
It
just
runs
as
a
github
action,
which
I
thought
was
really
cool
and
awesome
week
it
doesn't
publish
to
npm,
for
you,
like
I'm
still
I
still
don't
quite
know
a
good
way
to
do
that
and
get
like.
You
still
have
to
Kiev
a
we
at
Google.
We
have
the
wombat
dressing
room
which
we
open
sourced
for
publishing,
10
p.m.
C
B
B
D
A
I'll,
look
at
using
us
in
report
toolkit
a
guy
I
wish
I
could
use
it
in
mocha
Wow,
but
can't
to
keep
get
to
commit
miss
just
to
look
like
that,
so
just
just
thinking
outside
thinking
outside
box.
No
so
like
I
know,
we
know
about
this,
this,
the
OTP
stuff
in
in
publishing.
What,
if,
like
you,
had
a
tool,
and
so
you
would
have
this
I
don't
know.
Maybe
you'd
have
an
account
that
was
like
an
NPM
account
and
that
NPM
account
was
for
your
releases
and
you
own
a
package.
A
You
are
the
owner
of
the
package.
You
are
you,
and
so
when
it's
time
to
release
you
have
some
script.
That
adds
this
other
bot,
for
whatever
adds
that
NPM
user
to
the
owners
of
your
package
and
then
with
with
no
there's
no
2fa
enabled
there
and
then
you
and
then
that
bot
publishes
your
package
and
then
as
soon
as
it's
done,
it
removes
them
from
the
owners
list
or
something
I
don't
know.
Would
that
would
that
like
work,
I
mean
it
seems
flimsy
as
hell,
but
I
mean.
C
C
C
A
B
Yeah
the
thing
is
that
yeah,
the
security
folks
are
not
gonna
like
it.
The
idea
of
having
to
FA
is
is
to
have
some
way
to
authenticate
like
having
it's
a
second
factor
right
like
not
not
have
just
because
if
your
users
compromised,
if
that
user,
even
though
you're
adding
and
removing
it
gets
compromised
on
the
way,
there's
nothing,
you
can
do
right.
So
security
folks
will
definitely
not
approve
yeah,
I.
Guess.
A
A
Like
a
temporary,
what
about
that
like
just
like
temporary
keys
but
hey
this
key,
this
publish
key
lasts
three
minutes:
I,
don't
know,
I
mean
how
I
don't
know
I.
You
know
it's
just
like
I
I,
don't
know
of
any
other
service
that
like
where
you're
trying
to
automate
stuff
and
it
makes
you
go
through
the
OTB
dance
like
is
that
does
anybody.
You
know
another
thing
that
that
that
makes
this,
because
every
places
I
see
works
with
keys
and
tokens
and
right
I've.
C
Been
to
and
other
big,
like
enterprise
companies
that
do
you
have
it
as
part
of
like
everything
to
do
and
there's
like
an
app
on
your
phone,
where
you
just
have
to
hit
your
thumb
on
it.
Occasionally
when
some
times
when
you're
doing
something
on
a
web
app,
that's
automated
mm-hm,
so
I
definitely
seen
it.