►
From YouTube: Node.js Diagnostics WG - 2018-10-03
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
C
C
That's
Friday!
Okay,
now
there
is
some
read
like
the
I
know.
They
were
like
that
survey
that
you
mentioned
Joey
I,
think
that
was
partially
to
see
where
people
in
which
sessions
people
are
going
to
figure
out
West
sized
rooms.
I
think
there
might
have
some
effect
on
the
schedule,
but
when
I
looked
a
while
back,
that
was
what,
where
we
were
scheduled.
A
D
A
F
C
You
know
I'd
say
if
there's
things
people
really
want
on
the
agenda
and
specific
topics
that
people
say:
hey
it's
a
great
thing
to
cover
when
we're
face
to
face
getting
those
in
the
agenda
and
upfront
would
be
good.
Otherwise,
you
know
I
think
we
can
see
based
on
the
people
who
were
there
what
the
interest
is
to
go
through.
C
G
And
go
ahead:
okay,
I
would
be
happy
if
we
could
talk
about
post-mortem
diagnosis
in
the
future.
So
that's
great
that
we
have
other
node
now
and
we
can
analyze
condoms,
but
with
promises
potentially
and
acing
hoax.
Probably
it's
not
the
future,
or
this
is
something
that
I
don't
know,
so
it
would
be
nicely
to
be
talked
about
that
which
direction
we
are
expecting
from
from
post-mortem
debugging.
C
G
It
would
be
great
to
have
a
direction
about
what
what
we
think
in
OGS
would
be
useful
in
the
future.
So
is
it
something
like?
Is
it
the
core
dump
analyzes
the
future,
for
no
J's
or
is
some
something
different?
So
it
would
be
nice
to
agree
about
this
and
and
come
up
with
some
plans
how
we
can
achieve
that
future.
G
G
Async
hooks
would
be
in
an
alternative,
but
it
seems
to
be
that
we
completely
stuck
with
a
performance
overhead,
so
I
don't
have
a
really
good
idea
now,
but
it
can
be,
for
example,
a
plan
that
we
want
to
use.
They
sing
hoops.
So
we
have
to
fix
the
performance
issues
around
there.
That
could
be
one
potential
issue,
but
I
don't
have
anything
concrete
so.
G
G
It's
more
like
it's
more
like
I
would
be
interested
to
see
the
current
state
of
the
protest
when
I
was
brushing
and
because
which
promises.
We
cannot
be
sure
that
the
state
doesn't
change
between
between.
We
are
crashing
the
process
of
not-so-bright
ability
to
to
being
able
to
see
what
was
but
I
about
there
yep,
mainly
this,
because.
C
C
And
in
that
case,
ideally
what
we'd
want
is
a
core
dump
to
occur,
or-
or
maybe
you
know
a
note
report
or
some
sort
of
set
of
Diagnostics
information,
but
that's
where
we've
been
challenged,
I
haven't
I
haven't
come
up
with
any
alternatives
with
the
only
thought
I
did
have
was
tonight.
I
don't
know
anyway.
The
only
thought
I'd
had
was
like
if
we
could
be
smart
enough
to
learn
about
and
handled
rejections.
C
H
So
and.
C
H
Go
ahead
at
the
moment,
v8
was
and
pushing
something
out
and
to
actually
have
spec
traces
or
async/await
and
and
that's
with
zero
cost
over
half.
So
we
are
going
to
get
these
in
general
and
from
then
on
and
debugging
async/await
is
going
to
be
easier.
It's
also
going
to
happen
for
promise
all
and
from
its
race.
It's
not
going
to
happen
for
promise
them
all
promise
reject,
and
but
this
would
be
possible,
but
it
would
not
mean
zero
cost
over
hand.
C
G
A
I
think
like
for
me
the
thing
that
would
be
helpful.
There
is
like
getting
to
some
level
of
precision
about
what
we
want
to
be
able
to
answer
in
those
cases.
So,
for
example,
it's
like,
if
we
say
I
have
a
unhandled,
rejection
or
I
have
a
promise
that
was
rejected
and
there's
no
handlers
on
it
and
the
thing
got
GC
and
you
know
I
want
it.
I
want
information
about.
You
know
the
promise
chain
or
something
along
those
lines
called
diagnose.
The
problem.
A
G
Recently,
I
had
a
meeting
with
young
and
based
on
that
conversation
seems
to
be
that
for
specific
promise,
use
cases
so
detecting
when
we
have
an
unhandled
rejection
or
doing
something
when
we
are
trying
to
resolve
the
promise
twice,
it
seems
to
be
that
now
there
are
solutions.
You
know
J
space
like
we
have
some
even
time
it
goes
or
hooks
or
we
are
handling
those
edge
cases
suffer
seems
to
be
that
the
promise
use
case
is
the
new
direction.
Is
that
it's
not
a
sin
hoops?
G
We
are
just
adding
specific
features
to
do
no
J
score
that
can
handle
these
edge
cases
or
candy.
You
notify
the
user.
So
I
think
this
is
also
something
that
we
could
agree.
That
is
this
a
new
direction
and
we
we
don't
want
to
build
on
18
hoops
or
it's
still
the
dream
that
one
day
we
have
a
singles
and
we
use
those
I
know
it's
not
completely
related
to
the
post-mortem
debugging.
Just
this
is
also
something
that
I
I
see
that
we
are
thinking
about
lots
of
different
solutions.
G
A
C
It's
a
long-running
discussion
because,
like
some
people
are
feel
strongly,
we
shouldn't
throw
or
cause
a
core
dump
because
it
would
break
like
it's.
Not
spec
compliant
would
break
some
code.
That's
perfectly
within
this
back.
Other
people
are
happy
to
actually
do
that
because
it
would
provide
more
information
and
I.
Don't
think,
there's
any
clear
answer
like
there's.
No
good
right
answer
right,
which
makes
it
difficult
for
people
to
coalesce
on
one
thing:
I
agree.
G
A
A
C
H
C
F
A
A
A
C
So
there's
sort
of
if
you've
got
new
people
in
the
room,
and
you
want
to
like
give
them
a
snapshot
of
what
going
on
and
get
them
encouraged
to
join
in.
It
might
be
worthwhile
going
through
it
a
more
surface
level
and
then
as
soon
as
you,
the
sort
of
working
sessions
was
the
second
part
like
as
soon
as
you
say.
Okay,
now
we're
going
to
go
into
really
nitty-gritty.
Then
you
really
do
need
somebody
who's
going
to
lead
that
particular
session
right.
A
A
C
G
A
E
D
G
By
the
way,
I
don't
know
how
many
people
knows,
but
mad
house
will
hold
a
workshop
about
a
low
note
and
it
seems
to
be
that
more
than
60
people
is
interested.
So
probably
we
can
collect
some
good
feedback
after
that
work.
Group
I,
don't
know
that
it
was
there
any
discussion
around
that,
but
I
think
it
would
be
useful
to
send
out
a
survey
or
find
a
way
to
to
collect
some
feedback
from
that.
Sixty
people
who
is
interested
in
in
post-mortem
debugging,
that's.
C
G
C
And
then
maybe
yeah
I
think
your
idea
of
like,
if
you
you
probably
need
their
like
email
address
or
whatever,
but
like
we
sent
around
a
sheet
saying
hey
we'd
like
to
send
out
a
summary
or
sorry
a
survey
afterwards,
you
know
if
you're,
okay,
with
that
add
your
add
your
name
here,
would
be
that'd,
be
really
handy.
Okay,.
A
A
So
I
think
what
it
is
now
is
what
we
probably
want
ride
safe,
feel
free
to
shoot.
This
down
is
for
those
the
the
first
set
of
bullets
right
the
goals
and
the
current
initiatives.
You
probably
want,
like
a
quick
side,
slide
deck
that
somebody
can
go
and
talk
through
that
and
say
here's
where
we're
at
with
tracing
your
route
with
profiling,
right
here's
where
we're
at
with
memory
analysis.
You
know
this.
D
A
Progress
that
was
maybe
this
is
where
we're
trying
to
get
to.
This
is
where
you
can
help
me
I.
Think
for
Michael
comment
that
it's
like.
Can
you
get
other
people
involved
and
excited,
and
hopefully
that
can
go
quickly
right,
definitely
and
then
I
think
what
we
do
is
we
then
have
time
box
working
section
since
we've
identified
to
you
and
I
think
you
know,
one
can
be
post-mortem
of
promises
and
another
comedian
saying
determine
all
of
your
stuff
and
I
think
that
and
I'm
I'm
happy
to
drive
acing
terminology.
A
So
if
you
got
a
talk,
they
got
accepted
for
this
on
Wednesday
on
the
async
terminology
stuff,
so
the
deck
that
I've
had
and
kind
of,
like
you
know
this
mildly
pro.
Actually
it
wasn't
even
mildly.
It
was
like
I,
don't
know
the
b-league
form
is
getting
polished
on
and
that's
gonna
we're
gonna
present
that
on
Wednesday
sometime
I
can't
remember
that,
though
deductive
yeah
I'd
know
it
interactive
yeah
on
it's
a
1140
I
won't
say,
and
so
what
I
can
do
there
for
the
acing
terminology.
A
E
So
I've
written
down
like
I
needed
to
create
a
slide
deck
to
talk
about
the
car
initiatives
and
the
current
state
of
Diagnostics
just
as
an
overview
and
then
hope
with
I
shouldn't
be
more
than,
and
it's
maybe
and
then
we've
done
the
rest
of
the
time.
Divided
between
talking
ways
and
nology.
We're
like
you'll,
be
giving
like
an
abridged
version,
every
presentation
and
then
it
would
be
your
heading.
The
post-mortem
part
of
this
perfect.
D
A
C
Know
what
I'd
almost
suggest
is
that,
after
the
after
the
conference,
we
might
want
to
set
like
a
dedicated
meeting
just
to
focus
on
that,
at
least
as
a
kickoff
or
something
like
cuz
I.
Think
what
it
could
just
use
is
sort
of
the
high-level
view
of
what
we
want
to
write
up
so
that
it
would
be
easily
and
clear
to
say
you
know
we
want
these
six
guides
right
and
then
it's
easier
for
people
to
say.
C
For
example,
you
know
I'm
gonna
get
us
his
name
wrong,
but
so
I
Zack
has
done
some
work
on
like
flame
graphs
and
I'm
sure
he's
you
know
it
sounds
like
he's
interested.
If
we
had
another
one
that
said,
here's
it
here's
another
one.
He
would
also
jump
into
that
one
right
so
I
think
just
having
that
focus
thing
where
we
just
try
and
put
together
the
best
first
guess
of
what
the
best
practices
would
look
like
in
what
areas
to
cover
we
could
enable
other
people,
then
jump
into
specific
ones.
B
D
C
Just
gotta
say
any
any
any
chance
you're
willing
to
set
up
that
meeting.
He
yeah
my
good
blue
that
you,
like
basically
an
open,
an
issue
saying
hey.
This
is
what
we're
gonna
do
in
this
meeting.
If
you're
interested,
let
me
know,
and
then
you
know,
come
coordinating
the
people
to
to
the
meeting.
That
would
be
really
good,
so
yep
that.
A
C
Yeah,
like
it's
basically
get
together
for
now
or
on
line
orbit
at
some
point
just
to
talk
about
that
right.
Cuz,
like
these
meetings,
we
have
enough
different
things
and
we're
not
really
sort
of
coming
with
the
headspace
I.
Think
if
you
just
said
this
is
the
hour
people
might
think
about
it.
Before
we
come,
we
hash
something
out,
we'll
probably
have
like
a
good
enough
to
start
working,
and
then
we
have
volunteers
and
we
can
say
well,
we
don't
have
volunteers
for
these,
and
you
know:
okay,.
A
Okay,
deep
dive,
so
I've
got
this.
This
I've
been
saying
we're
gonna,
do
one
on
the
async
contact
stuff
and
so
now
I
actually
have
this
forcing
function
that
you
know
I've
got
to
go
in
and
get
actually
polished,
slides
and,
like
you
know,
talk
about
so
maybe
I'll
schedule.
One
of
these
is
like
a
practice
run
for
me.
I,
don't
know
like
people
have
like
the
only
time.
I
can
think
that
this
would
work
would
be.
C
C
A
Maybe
that's
fair
I
was
hoping
to
get
a
practice
run
and
get
like
yeah
you
a
up
and
out
from
a
you
know
a
smaller
and
you
know
potentially
more
friendly
audience.
Maybe
but
that's
fine.
Let's
do
that.
Then
I'll
I'll
plan
on
and
we'll
see
what
falls
out
of
the
stuff
at
the
summit
and
then
I'll
schedule.
Some
some
follow
up
there
and
at
least
I'll
have
the
deck
as
a
collateral
thing
to
drive
conversations.
A
A
H
Coffin
profit
process
and
we
should
probably
support
it
and
because
that's
what
I'm
broke
recently
in
the
clinic
and
I
actually
went
ahead
and
also
open
a
PR
to
fix
the
note
internal
tests
that
we
have
for
this.
For
the
ticket
processor.
It
is
a
bit
ugly
because
we
test
it
on
all
the
M
different
systems
and
not
on
all
systems,
and
we
actually
get
out
a
good
data
back.
H
H
C
H
C
H
See
if
something
and
became
defined
with
meantime
and
removed
elder
as
skipping
part
first
and
remnant
on
all
the
systems,
and
but
it
turns
out,
it's
still
an
issue
pretty
much
everything
there.
You
skipped
and
also
I
had
to
add
a
couple
of
fall
banks,
because
it's
not
that
easy
to
to
get
it
on
some
systems
that
we
run
our
tests
on
right.
C
H
So
if
I
must
and
correct
what
you
would
like
is
pretty
much
a
list
of
systems
that
it
fails
at
yes
and
then
we
might
see
what
we
could
improve.
So
we
actually
improved
the
data
output.
C
Exactly
so
like
you
know,
if
we've
said
that
that
this
prof,
you
know
profit
profit
process
flag,
is
one
of
our
key
debugging
tools.
If,
for
example,
it
doesn't
work
on
regular
Linux,
we
should
prioritize.
You
know
publicizing
that
that's
an
area
somebody
could
contribute
in
right.
If
it
just
doesn't
work
on
AIX,
for
example,
then
you
know
I
can
maybe
highlight
internally
that
well,
this
is
becoming.
C
H
It's
maintained
by
v8
itself.
We
just
use
that
feature
from
v8
right,
but.
C
But
we
can
change
things
in
v8
too,
right
of
course,
yeah
and
and
are
like.
We
have
people
who
work
on
that
I.
Just
I
just
was
thinking
it's
when
it's
like
when
we
have
partial
testing
until
we've
got
like
hey.
This
is
good
good
testing
across
the
node
platforms.
We
should,
if
we're,
emphasizing
and
working
on
it,
we
should
highlight
where
there
the
gaps
are
because
then
we
can
maybe
line
up
something
to
help
fix
those
or
or
document
it
somewhere.
That
says,
but
it
doesn't.
C
H
C
I
just
pasted
the
page,
so
there
was
a
PR
and
it
did
land
to
move
a
number
of
things
to
tier
3.
Okay,
which
is
the
like,
will
run
them.
If
we
can-
and
you
know
it-
we
see
that
we
have
like
partial
v8
cpu
profiling
tests.
The
question
I'd
have
from
that
is
like.
Okay.
Is
that
partial
to
me
that
sounds
like
well,
we've
got
some
of
what
we
want,
but
not
what
else.
So
we
should
probably
make
an
issue
somewhere.
That
says:
what's
the
something
else
we
want
to
or
think
we
should
do.
C
C
H
C
C
I
C
G
G
C
G
A
A
A
Perfect,
okay,
anything
else
here,
I
think
this
is
a
good
sort
of
testing
thing
to
try
and
you
know,
figure
out
the
best
process
for
moving
these
things
up
right,
profiling,
okay,.
A
A
F
I
I
I
A
G
Them
things:
okay,
okay,.
A
So
a
sink
context
like
I,
said:
I've,
there's
a
talk
with
God
in
wait-listed
and
then
last
week
they
were
like
you
know:
you've
been
accepted,
so
I
read
to
do
that.
So
it's
sort
of
it's
a
good,
forcing
function,
sort
of
bubbled
down
on
my
stack
and
so
so
yeah
so
I'm
all
be
at
Summit
and
talk
is
like
11:40
on
Wednesday.
So
so
hopefully
some
of
these
ideas
are
coalescing
and
again,
I
think
the
goal
there
is
to
just
you.
A
In
the
forum,
I
think
the
goal
there
is
really
to
get
feedback
again
from
a
broader
audience,
and
you
know
particularly
around
terminology
and
around
you
know
the
the
next
steps
forward
or
something
like
this
and
it's.
C
A
I
Don't
have
any
updates
on
that
one
other
than
I've
been
working
on
fixing
some
of
the
trace
event,
issues
and
there's
a
whole
cascade
of
bugs
that
actually
currently
waiting
on
a
thread.
Initialization
issue,
one
windows
and
once
that
is
fixed
I
am
hopeful
that
trace
events,
so
the
some
of
the
race
conditions
will
be
fixed.
I
still
owe
this
working
group
a
detail,
write
down
off
what
are
the
concrete
things
that
need
to
happen
for
treatment
and
I
haven't
been
able
to
do
that.
Yet.
Okay,.
A
Okay,
that
was
the
last
thing
on
the
agenda.
I,
think
anything
else
from
anybody.
A
Okay,
I
guess
we're
done.
Okay,
great.