►
From YouTube: 2020-08-19 Node.js Diagnostics WG 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
A
A
The
scope
of
the
deep
dive
meetings,
which
we
have
every
other
week,
two
other
topics.
Besides
the
deep
dive
journeys
that
we
were
doing
before
we
have
some
items,
fly
to
the
deep
dive
as
suggestions
and
last
week
we
actually
deep
dived
into
one
of
these
other
topics.
A
So
this
is
on
the
agenda,
basically
for
awareness
of
folks
that
were
not
aware
yet
that
we
are
going
to
start
discussing
other
topics
on
those
zip
types.
A
We
still
have
to
finish
the
journeys,
but
maybe
it's
good
to
take
a
break
and
look
at
other
topics,
especially
now
that
someone.
So
we
started
to
have
some
topics
to
tell
to
discuss
again.
A
A
So
right
now
we
have
meetings.
Every
week
we
alternate
between
the
diagnostic
meetings
and
the
dip
type
meetings
and
the
main
reason
we
were
having
every
other
week.
Meetings
was
to
work
on
the
user
journeys,
but
since
we
are
almost
finished,
I
was
wondering
if
we
should
get
back
to
every
other
week
meetings
and
maybe
alternate
between
deep
dives
and
working
group
meetings.
A
Meetings
every
other
week
and
we
alternate
one
week-
we
are
having
the
working
group
meetings
and
the
other
week.
We
have
the
flip
types
and
then
we
go
back
to
the
working
group
meetings
and
we
go
back
and
forth
and
if
needed,
we
can
schedule
one
of
deep
dives,
which
I
think
makes
sense
for
some
topics,
because
we
need
to
involve
folks
that
are
not
regular
attendees
from
the
working
group
for
certain
topics.
C
Yeah,
so
what
we
had
before
it
was,
I
think,
this
regular
meeting
every
two
weeks
and
then
the
deep
dive
we
kind
of
like
added
in
the
spaces
between.
C
So
if
we
change
it
to
be
alternating
between
these
two
every
two
weeks,
then
this
one
would
only
be
once
a
month.
So
I'm
not
sure
how
people
would
deal
with
that.
We
wanted
that
infrequently.
C
It
seems
like
we
haven't,
had
a
ton
of
stuff
to
say,
but
it
would
still
be
good
to
be
able
to
give
some
updates
relatively
frequently
so
like
we
could
maybe
do
like
this
one.
That's
like
full,
just
update
everything
and
then
maybe
the
like
deep
dive.
We
could
do
like
a
quick
update
at
the
start
or
something
we
wanted
to
do
it.
That
way.
A
So,
instead
of
alternating
every
other
week,
we
alternate
in
the
meeting.
C
A
And
if,
for
some
reason
for
one
meeting,
we
have
more
topics,
we
can
use
most
of
the
time
for
normal
meetings.
I
think
that's
fine.
A
A
A
C
A
A
A
Okay,
next
topic
is
not
just
contributor
survey.
Jury
chatted
this
topic,
so
I'm
not
sure
if
we
can
discuss
it
today.
Anyone
read
this:
you
don't
have
context.
C
A
A
C
A
A
A
A
C
Not
too
much,
but
there
was
the
the
request
I
made
to
v8
fixed
then-ables.
Now
we
were
looking
at
backboarding
that
to
12,
but
there
was
apparently
some
issues
with
blob
watcher,
not
passing
tests.
It
was
breaking
for
some
reason
which.
C
C
No
so
especially,
I
think
that
should
get
backported
soon,
no
well,
there's
a
backpack
pr.
I
think
I
think
it
should
be
landed
soon,
probably
other
than
that.
I
haven't
really
seen
too
much
basic
hooks
progress
and
I've
been
working
on
other.
C
A
A
Unless
we
decide
that
facing
hooks
will
never
be
stable,
but
until
then
I
think
we
should
keep
it
on
the
agenda.
A
A
A
A
So,
moving
to
the
last
one,
improving
automated
remote
connection,
the
inspector
protocol,
so
I
did
this
issue
to
the
agenda
to
present
a
demo
of
a
common
line,
so
I
implemented
that
achieves
most
of
the
requirements
I
raised
on
the
issue,
which
are
no
need
to
make
changes
to
application
code
or
install.
So
the
plan
depends
on
the
fly,
trigger
inspector
tools,
all
already
running
processes
and
now
to
do
that
not
interactively
so
existing
tools
they
achieved
most
of
it,
but
none
of
the
tools
I
tried.
A
I
achieved
all
of
those
four
points
which
usually
they
achieve
one
or
two,
for
example:
usually
either
you
get
to
trigger
tools
and
already
running
processes
or
you'll
get
to
trigger
non-interactively
the
scripts.
A
One
example
is
the
inspector
client
that
we
ship
with
node.js.
It
allows
you
to
run
inspector
tools
on
already
running
processes,
but
it
doesn't
allow
you
to
run
non-interactively.
It
will
always
prompt
for
the.
A
A
And
the
idea
was
to
come
up
with
something
that
was
small
enough
in
scope,
that
it
might
eventually
be
parted
into
a
car,
but
that
fitted
those
three
requirements
and
this
last
one
the
only
way
to
really
fit
this
one
is
by
having
it
on
car,
so
it
all
depends
on
if
we
get
the
two
encore
or
not
so
I
wrote
a
tool
that
achieves
those
three
requirements
and
I'm
gonna
demonstrate
it
now,
so
I'm
gonna
run
a
server
here.
A
We
pass
a
duration
for
like
10
seconds
like
five
seconds
since
its
presentation
and
it's
going
to
run
for
five
seconds
and
then
it
outputs
the
to.
B
A
H4
hit
profile
means
after
how
many
bytes
allocated
it
will
sample
the
application
by
default.
It's
like
32
kilobytes,
I
think
so.
If
we
reduce
to
512
bytes,
it
should
capture
a
lot
more
frames.
A
A
A
And
there
we
go,
we
have
a
hip
profile.
We
didn't
have
to
make
any
changes
to
the
application
code.
This
can
be
run
with
npx,
so
it
has
to
be
installed
somewhere.
A
A
B
A
A
Let's
elaborate
so
when
you
start
inspector
protocol,
you
have
the
server
or
the
host,
which
is
the
application
you
are
inspecting
and
you
have
the
client
you
can
close
the
client
connection,
but
once
you
open
the
connection
on
this
server,
you
can't
close
it
unless
you
opened
it
programmatically
via
code
on
the
application,
which
is
not
the
goal
of
this
tool.
B
A
A
Inspector
comments
so,
for
example,
and
only
some
comments
so,
for
example,
if
you
you're,
if
you
could
start
the
connection
and
run
runtime
dot,
enable
it
should
not
incur
any
overhead,
but
if
you
run
hip
profile,
dot
enable
followed
by
hip
profile
takes
hip
snapshot.
B
B
C
If
we
get
this
into
car
and
does
that
give
it
a
path
to
be
able
to
turn
that
off.
A
C
No
looks
good
to
me
is:
is
there
anything
in
particular
blocking
it
from
like
opening
the
pr
trying
to
get
it
into
call
right
now.
A
I
think
before
we
add
the
apr
to
add
the
car,
we
need
to
try
to
move
it
to
the
foundation
and
after
that
we
can
probably
try
to
add
it
to
car.
It
already
has
tests.
I've
been
trying
it
on
production
from
time
to
time.
So
far
it's
been
fine,
so.
A
A
Do
you
issue
which
tools
because,
for
example,
nodding
spec,
doesn't
happen
dependencies?
They
implemented
their
own
websocket
client
to
connect
to
the
inspector
protocol.
C
C
A
C
A
C
A
And
next
week,
then
we
don't
have
a
meeting,
and
after
that
we
have
the
first
15
minutes
for
updates
and
rest
with
the
meeting
for
deep
types.
Just
checking
again
does
that
sound
good
for
everyone.