►
From YouTube: Node.js Tooling Group Meeting 2020-11-13
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).
B
Cool
great
well
welcome
anyone
who's
watching
on
youtube.
This
is
friday,
the
13th
of
november
2020-
and
this
is
the
no
tooling
working
group
meeting.
B
So
I
guess
we
can
just
dive
right
into
the
agenda
if
folks
want
to
add
their
name
who
are
in
attendance,
so
we
know
who's
know
who's
here,.
B
B
I
might
have
one
thing
that
I
spent
so
I
was
traveling
across
the
u.s
the
other
week.
Two
weeks
ago
I
got
node.js's
test
coverage
working
again,
so
the
kind
of
cool
new
thing
I'm
going
to
just
share
a
link
here.
B
It
actually
now
runs
coverage
every
time
someone
commits
to
any
branch.
It
doesn't
comment
on
the
pr,
but
if
you're
working
on
a
pr,
you
can
go
out
and
check
out
what
the
coverage
looks
like
without
running
it
yourself
locally,
and
it
includes
c
c
and
javascript
coverage.
Now
it
used
to
only
include
javascript
coverage.
So
if
you're
looking
for
a
place
to
contribute
to
the
project,
our
c
plus
plus
coverage
is
pretty
low.
Actually
so
maybe
a
low-hanging
fruit.
A
B
A
C
B
C
B
A
B
A
B
B
A
Yeah,
for
sure
poppy
was
definitely
happening.
Is
that
somehow
not
on
the
agenda.
B
A
A
Working
on
some
kind
of
like
an
rfc
for
that
which
I
actually
I
didn't,
I
wanted
to
start
on
this
week,
but
I
didn't
have
a
chance
to,
but
I
did
kind
of
start
looking
at
like
like
I
looked
at
the
npm
rfc
repo,
they
have
a
pretty
good
template.
It
looked
like,
and
I
started
like
putting
together
some
links
to
like,
for
example,
the
fs
extra
implementation
of
recursive
copy.
A
So
my
intention
was
to
spend
some
time
this
week,
but
maybe
maybe
I'll
I'll.
Do
it
this
weekend
to
at
least
start
writing
an
rfc
cool?
A
C
Might
be
good
too,
and
if
you
want
to
meet
the
next
week
or
something
and
work
on
it
together,
I'd
be
happy
to.
A
Sure
yeah,
I
could
try
and
like
cobble
together
like
a
really
rough
outline
or
something
and
then
maybe
we
could
kind
of
walk
through.
That
sounds
good.
B
Chris
sent
me
a
message
out
of
the
blue,
where
he
said
that
he
basically
said
he'd
been
reading
a
document
that
talked
about
that
security
issue,
that
isaac
had
brought
up
and
for
chmod
and
joan
type
stuff
and
one
way
to
mitigate
it.
He
says,
is
to
something
to
do
with
file
descriptors,
I'm
going
to
butcher
it,
but
there's
I'm
going
to
paste
the
text.
B
A
A
I'm
not
sure
I
haven't
looked
into
any
of
the
details
about
like
if
this
is.
This
is
going
to
be
an
issue
with
the
recursive
copy.
I
also
part
of
the
reason
I
was
looking
at
like
fs
extras.
A
B
A
So
that's
I
mean
that's
going
to
be
a
good
thing
to
discuss
right
now,
like
I
yeah,
I
did
look
at
their
implementation
of
it
and
it's
like
it's.
You
know
pretty
long.
It's
like,
I
think,
like
400
lines
of
code
or
something
like
you
said,
lots
of
like
edge
cases
and
things
like
that,
so
that
yeah
I
mean
that
does
kind
of
raise
the
question
right
like
do.
We
want
to
basically
just
copy
that
functionality
in
versus,
like
writing,
something
ourselves,
I'm
not
really
sure.
B
C
A
And
then
we
can
also
discuss
the
possibility
of
like
yeah
just
using
the
code
they've
already
written
or
see
if
they
have
any
concerns
with
with
that
or
or
if
they
think,
it's
better
to
start
fresh
or
whatever.
B
A
There
there
was
that
one
issue
that
someone
raised-
I
don't
think
it's
on
here,
but
the
way
that
we
handle
the
some
of
the
flags.
A
I
was
actually
unsure
how
to
handle
this
like
at
first,
I
thought.
Oh,
we
can
just
update
the
docs,
but
then
I
also
started
thinking
maybe
longer
term.
We
should
make
some
of
these
flags
work
like.
I
think
the
argument
was,
for
example,
we
we
don't.
We
only
use
the
the
retry
mechanism
when
you're
removing
something
recursively,
but
maybe
that's
a
nice
thing
to
have
just
in
general.
A
So
then
I
thought
maybe
we
implement
some
of
that
in
the
rm
command
and
then
I'll
be
wondering
like
do.
We
maybe
want
to
avoid
like
documenting
that
doesn't
work
a
certain
way,
then
changing
it
to
work
a
different
way
and
then
changing
the
docs
again
like
that
churn
is
maybe
just
more
confusing
than
just
fixing
it.
B
Yeah
I
mean
the
I'd
almost
rather
not
expose
the
well.
I
don't
know
the
right
answer,
like
a
big
big
reason
for
the
retry
is
purely
that
windows
behaves
weird
on
rimroaf,
so
so
it
would
feel
like
a
little
weird
to
me,
adding
the
feature
to
rm
when
that's
not
an
issue.
So
much
on
ram
like
the
issue
is
that
windows
gets
an
exception
on
an
unempty
folder
because
of
a
race
condition
deleting
files
from
folder.
B
So
then,
like
one
in
a
once
in
a
while,
you
need
to
run
rm
twice
on
the
folder,
which
I
think
is
like
the
motivation
of
a
lot
of
the
retry
logic.
A
B
A
Yeah
yeah,
maybe
we
can
just
put
that
on
the
agenda
for
next
time
or
something
I'm
not.
I
did
initially
look
at
this
and
then
I
kind
of
like
I
said
kind
of
thought
about
it,
a
bit
and
realized
like
I
don't
actually
know
what
the
correct
way
to
handle.
This
is.
B
Cool
okay:
if
we
move,
I
I've
got
a
kind
of
hard
stop
at
1
30ish
today,
so
okay
yeah
sure
see
how
far
we
can
get
through
the
agenda.
I
don't
want
to
rush
it,
but
if
we're
good
with
that,
I
mean.
B
A
B
Yeah,
let's,
why
don't
we
take
like
some
of
the
like
when
we
consolidate
some
of
the
recursive
file
stuff
just
into
the
one
issue?
That's
like
come
up
for
a
plan
for
the
recursive
stuff
and
get
rid
of
some
of
the
agenda
items
for
like
chima,
chmod,
r
and
stuff.
Maybe
I
think
that'd
be
the
smart
way
to
go
moving.
I
feel
like
there
is
no
real
champion
for
foreign
function.
Interface
right
now,
right,
yeah,
that's
kind
of
been
the
case.
B
B
Here,
oh
crap,
look
in
the
wrong
place.
Where
did
I
put
that
util.
B
B
B
I'm
talking
about
the
agenda
item.
Esm
module,
reloading
and
module
graph
number
51.
B
I
know
chris
hiller
had
a
like
a
prototype.
He
worked
on
with
someone
where
they,
I
think,
would
have
been
leaking
module
names,
but
they
were
able
to
kind
of
replace
them
real
time
or
something.
So
you
could
have
something
resembling,
like
you
know,
what's
the
name
of
that
library,
I'm
blanking
today,
like
just
basically
being
able
to
replace
like
require
spy
kind
of
thing
like
whatever
that
require
module.
B
B
That
would
be
worthwhile
yeah
like
like,
I
think,
if
we
can
show
it
working
really
well
and
if
it
feels
like
something
there
should
be
a
user
land
tool
that
does
this.
If
you
delivered
it,
it
would
probably
drive
more
people
using
esm
and
we've
just
been
sitting
on
this,
not
doing
it
forever
so,
but
I
feel
like
we
could
just
build
a
prototype.
B
B
B
A
proxy
choir,
it's
called,
which
is
like
a
way
to
replace
something
at
require
time,
instead
of
like
mocking
it.
Instead
of
mocking
it
using
something
like
sign
in.
B
B
A
Yeah,
okay,
I
mean
that's
worth
adding
to
the
issue,
maybe
or
maybe
we
make
a
new
issue
about
like
yeah
bill
either
either
like
trying
out
one
of
these
libraries.
If
this
is
a
thing,
that's
been
published
or
building
something
yeah.
B
B
Right,
oh
and
maybe
it
could
use
weak
references
so
that
it's
not
actually
a
memory
leak,
and
that
would
be
a
way
to
get
around
one
of
the
main
concerns
we
had,
which
was
that
you'd
be
leaking
these
modules
when
they
were
not
used
any
longer.
Of
course,
as
long
as
you
you're
not
holding
reference
to
them,
you
don't
even
need
to
use
a
weak
map
like
it
should
just
should
just
be
garbage
collected
cool.
B
A
Which
I
could
definitely
see
being
a
big
issue
for
testing
like
we've,
been
fighting
some
issues
where
we
have
unit
tests
that
are
leaking
memory
and
then
the
more
tests
you
had
the
more
memory
the
node
process
uses
to
run
them
all
so
yeah.
I
could
definitely
see
that
potentially
being
a
problem.
B
Okay,
so
what's
the
action
item
there?
I
guess
it's
read
a
bit
more
into
that
library
and
and
and
if
it
turns
out
there's
like
good
options
in
user
land,
maybe
we
can
drop
it
from
our
agenda.
B
Better
way
to
these
are
the
ones
that
we
were
saying
these
hooks
here
were
ones
we
were
saying
we
stephen
bellinger,
was
even
on
the
call
last
week
and
he's
said,
he's
continuing
to
look
into
this
occasionally,
and
I
think
I
had
an
action
item
to
talk
to
him
about.
B
B
B
And
we
need
to
I'd
love
to
get
these
agendas
auto
generated
by
like
hack
pad
or
something
it's
these
note.
These
stocks
end
up
looking
a
little
janky
by
the
time
you
put
some
notes
in
them.
A
That
is,
that
is
in
progress
right,
a
pr.
B
A
I
looked
at
it
the
other
day,
though
it's
it's
waiting
on
some
other
functionality
that
they're
they're
trying
to
get
in
there
for
the
I
forget,
reported
that
from
like
the
package,
maintenance
group
or
something-
or
he
has
a
note
in
the
note
of
the
priority
waiting
on
this
one
thing
to
be
fixed.
B
Cool
anyone
else
have
anything
like
the
the
reason
I
was
interested
in
the
hook
stuff
and
corey
when
he
was
joining
more
often
was
interested
was
that
we
have
to
do
this.
B
We
do
this
with
some
of
our
test
coverage
stuff
to
inject
variables
and
whatnot
as
a
sub
process
happens,
so
that
you
collect
coverage
for
the
sub
process
as
well
as
the
top
level
process
and
and
there
I
think
there
was
just
and-
and
we
yeah
we
do
this
for
test
coverage
and
node
itself,
and
we
do
it
kind
of
with
some
hacky
approaches
in
nyc
the
that
old
test
coverage
library
I
used
to
work
on,
so
it
would
be
kind
of
neat
if
there
was
a
blessed
way
of
doing
this,
and
so
I
think
the
idea
was
just
that
we
make
sure
we
talk
to
stephen
bellinger
and
make
sure
if
possible,
our
use
case
is
covered
if
he
does
start
to
take
on
some
of
this
work.
B
B
I
refactored
our
source
map
support
so
that
it
now
works
for
the
one
edge
case
where
it
didn't
work,
which
is
what,
if
an
exception
happened
before
your
program
actually
finished
its
first
tick.
So
if
there
was
so,
if
your
module
was,
if
your
module
uses
typescript
or
something
and
then
fails
before
it
actually
gets
to
the
next
tick
of
execution,
then
there
was
no
way
for
us
to
track
source
maps
which
now
there
is
so
that
fixes
one
of
those
edge
cases,
and
we
also
had
a
bug
where
we
weren't
parsing
webpack
properly.
B
It's
finding
the
issue
there's
one
other
really
finicky
bug
that
is,
that
source
maps,
don't
if
you're,
using
webpack
or
uglify
source
maps,
don't
give
you
the
stack,
there's
no
way
for
us
to
go
from
the
stack
trace
back
to
the
original
function,
name
because
the
function
name.
We
have
the
line
and
column
position
of
where
the
exception
happened,
but
we
don't
have
the
line
and
column
position
of
where
the
enclosing
function
exists,
that
the
exception
happened
within,
so
we
can't
show
it
in
the
stack
trace.
B
So
I
just
started
a
conversation
with
the
folks
on
v8
to
see
if
we
could
expose
that
information
from
v8,
which
which
should
mean
without
any
like
backflips
or
parsing
or
anything
weird.
We
should
be
able
to
give
a
better
stack
trace
than
we
have
today
when
you're,
using
typescript
or
uglify
webpack,
anything
with
source
maps.
B
B
All
right,
I'm
talking
too
much
moving
down.
Moving
on
to
the
last
one
argument:
parsing,
I
don't
know
why
this
is
the
end
of
our
agenda,
because
this
is
a
really
exciting
one.
C
But
I
I
pulled
down
chris's
branch
and
am
processing
some
of
the
blocking
comments
in
the
pr
so
that
I
can
open
up
a
new
pr
make
sure
those
comments
are,
you
know,
brought
over
to
the
new
pr
and
then
start
to
address
them
and
try
to
get
that
over
the
finish
line.
B
C
B
C
Yeah,
if
I'm
not
mistaken,
I
think
there
are
three
blocking
comments,
so
I
will
figure
that
out
bring
bring
them
over
to
a
new
pr.
A
What's
that,
I
think
the
biggest
one
was
around
the
exception:
handling
like
yeah,
throw
or
return
an
errors
object.
I
think
that
was
from
what
I
remember
like
the
the
sort
of
the
biggest
outstanding
issue
where
there
was
kind
of
you
know,
people
in
both
camps.
C
Yeah
when
I
before
I
opened
the
pr
and
and
bring
all
that
stuff
over
I'll,
maybe
ping,
the
group
and
slack
and
just
reiterate
where
I
think
we're
at
and
and
you
know
we
can
discuss
a
little
bit
and
you
know.
B
Awesome
I
feel
like
like
that
feature
would
be
super
exciting.
The
the
stuff
you're
doing
with
recursive
fs
is
super
exciting
and
then,
like
I'd,
be
stoked.
If
I
could
get
the
stack
trace
stuff
to
a
more
mature
point
and
then
like
pretty
good,
delivered
some
pretty
cool
stuff
as
a
group
in
2020
at
that
point,
which
is
exciting,
awesome.
Well,
I
hate
to
rush
the
meeting,
but
I
think
I
have
like
two
meetings
that
overlap
with
it.
So
I
think
I'm
gonna
go
join
one
of
the
other
ones.
That's
okay
sounds
good.