►
From YouTube: Diagnostics WG meeting, April 22, 2020
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
C
We
could
just
add
to
the
calendar
like
it
could
be
same
time,
the
alternate
weeks
and
just
say,
like
Diagnostics
dives.
The
only
thing
that
wouldn't
happen
is
that
wouldn't
necessarily
create
an
issue
in
the
repo
I.
Don't
know
if
that
like,
if
we
name
it
differently,
wouldn't
create
an
issue
necessarily
but
I'm,
not
sure
we
need
one
or.
A
A
A
A
C
A
Okay
and
welcome
in
the
working
group
and
I
see
Harshita
that
you
are
on
the
call
actually
now
that
you
are
becoming
part
of
the
working
group,
I'm
wondering
if
just
you
wanna
share
in
couple
of
sentences
like
what
we
are
doing.
What's
the
main
interest.
Sorry,
if
you
earlier
said
it
I'm,
really
bad
with
name.
E
E
A
B
The
point
is
we
originally
we
discussed
this
and
thought
of
opening
an
issue
and
see
if
people
are
having
business
as
usual,
we
could
just
close
it,
but
then
now,
if
you
see
the
situation,
it's
quite
dynamic
and
it
it's
not.
If
you
see
the
graph
right,
it's
still
relevant
things
can
change,
so
it
makes
sense
to
keep
it
up
while
keep
it
open
for
a
while.
Okay.
A
B
So
the
process
was,
we
first
proposed
in
the
diagnostic
working
group
got
the
recommendation.
Then
I
raced
as
a
PR
in
the
core
apoo.
There
was
couple
of
commands.
I
address
that
I
got
one
approval,
ideally
based
on
the
process.
I
could
land
that,
after
waiting
for
seven
days
with
one
approval
but
in
the
best
practice
say,
is
that
at
least
to
approval
so
I'm
just
waiting
for
waiting
to
see
the
next
second
approval
to
come
in
was.
B
A
G
A
A
A
Okay,
so
this
looks
like
it's
also
going
report
version.
Semantics
are
not
fine
if
I
remember
correctly
and
like
the
last
face,
was
that
Mattel
said
that
he's
going
to
propose
something
around
how
the
schema
could
be
mapped
to
versioning
I'm,
not
sure
if
there
anything
happens
instead,
what
else?
If
you
have
a
time
to
put
any
thoughts
into
that,
make
sure
he's
not
in
the
corn
in
the.
A
A
A
A
A
A
E
E
H
H
H
H
The
thing
is
CPU
profiling
is,
we
will
always
happen
on
v8,
it's
not
something
that
we
can
implement,
because
you
don't
have
access
to
the
information
needed
to
generate
a
CPU
profile
and
because
of
that,
if
we
want
to
improve
the
situation,
make
the
two
more
production
safe
speaking
of
QT
profile,
you
want
to
make
dat
profile
more
production
safe.
We
need
to
engage
with
the
v8
team
again
and.
H
C
A
Before
we
jump
into
the
weird
collaboration,
the
first
points,
what
you
raced
around
like
confusion
and
in
with
people
like
which
was
has
supported
how
they
work
those
kind
of
things,
I
think
that's
actually
the
next
week.
Conversation
will
be
a
great
starting
point
and
about
the
v8
engagement
totally
I
am
open
to
any
ideas.
A
C
C
That's
the
point
where
it
would
be
good
useful
to
go
back
to
the
v8
team
right
mm-hmm,
so
probably
out
of
them.
You
know
the
deep
dive
and
maybe
any
follow-ons.
We
should
clearly
try
and
clearly
identify
like
here's.
What
we
think
you
can
do
today
here.
The
shortcomings
here
are
the
things
we'd
like
to
work
on.
Then
we
can
get
the
v8
team.
You
know
find
the
right
people
to
talk
to
on
the
v8
side.
B
Yeah,
for
me,
the
profiling
is
one
of
the
good
candidate
for
the
B
use
a
journey
discussion,
because
at
the
moment,
if
you
see
there
is
no
standard
way
of
telling
whether
you
should
be
whether
you're
going
to
collect
the
profiling
information
and
throughout
the
life
cycle
of
the
application
and
caddy
data,
only
at
the
end
of
it.
But
is
there
any
way
you
can
stream
the
data
and
being
able
to
generate
inside
from
instantaneous
views
of
data,
etc.
B
Those
kinds
of
things
are
not
necessarily
coming
from
the
user
journey,
but
from
the
developers
as
they
provide
the
tooling
capability
and
then
retrofit
into
these
specific
use
cases.
So
definitely
I
believe
this
should
run
through
the
user
journey
and
come
up
as
a
set
of
recommendations
or
set
of
concerns
etc,
which
we
can
interlock
with
a.
A
C
A
C
A
H
H
Start
also,
since
we
are
talking
about
Diagnostics,
there
might
make
sense
to
reach
out
to
the
chrome
dev
tools
team,
since
the
underlying
underlying
features
that
we
need
for
Diagnostics
are
the
same
that
they
need
for
dev
tools,
for
example,
deviates
if
you
profile
is
used
on
dev
tools,
so
is
the
heap
profiler.
So
is
the
heap
snapshot
and
they
might
have
more
bandwidth
to
work
on
that
and
the
heating?
H
A
Okay,
so
it
sounds
like
the
plan
will
be
that
we
made
this
usually
next
week
we
are
going
to
probably
identify
some
gaps
and
there
are
some
other
gaps
and
open
items
or
earlier
that
we
know
it's.
We
want
to
change,
so
maybe
we
should
write
them
down
in
I,
don't
know
some
easily
consumable
format
and
then
pinging
both
the
v8
team
and
contactus
team.
Maybe
we
could
invite
them
to
to
run
out
for
weekly
meetings
and
figure
out
how
we
are
going
to
collaborate
on
that
and
who
is
going
to
be
the
champion?
A
If
there
is
no
more
what
you
can
discuss
about
CP
of
a
filing,
then
we
can
go
to
a
singles
and
Steven.
Do
you
want
to
talk
about
it?
I
know,
last
time
you
shared
a
lot
I,
don't
know
if
you
have
a
changes,
yeah.
D
D
D
G
A
C
G
C
G
C
C
C
Completely
yeah,
as
long
as
as
long
as
it's
basically
exposed
as
a
like
a
handle,
which
you
know
it's
just
whatever
you
you
get
something,
and
then
you
know
you
get
it
from
one
call
and
you
pass
it
to
another
call
that
touch
me.
That
is
fine.
That
changes
behind
the
scenes
we've
done
that
before,
for
example,
and
in
other
things
as
well,
but
it's
good
to
bring
up
and
we'll
you
know,
take
double
check
that
we're
all
comfortable.
If
that's
the
case.
D
D
Javascript
implementation
of
the
promise
hug
handler,
which
doesn't
do
any
GC
tracking
and
doesn't
emit,
destroy
events
so
that
that's
I
have
PR
put
together.
It
seems
to
basically
I'll
be
functional.
There's
one
comments
I
just
got
since
yesterday:
I
haven't
had
a
chance
to
respond
to
you
yet,
but
the
functionalities.
Basically,
all
there
there's
some
issues
with
the
test
suite
not
actually
running
my
tests
for
some
reason
this
it
seems
to
pass
the
test
now
are.
D
D
Like
it,
it's
at
a
hit
time
will
create
I
get
if
it's
in
there,
like
hook
with
a
destroy
mode,
then
it'll,
and
this
they're,
like
promise,
wrap
on
the
C++
side
and
like
that
that's
class,
has
the
destroy
ability.
If
it
does
the
in
this
in
the
JavaScript
side,
then
it
it's
just
a
regular
promise
and
it
just
text
some
properties.
D
Specifically,
if
you
have
it
enabled
with
no
destroy
hook,
and
then
you
enable
with
a
destroy
hook.
So
it's
enabled
in
both
in
both
situations
it's
just
you've
now
added
as
it
is,
but
yep
until
that
point
it
didn't
have
the
disk
drive
hook.
So
it
wasn't
attaching
any
of
the
stuff
to
do
the
GC
tracking
to
be
able
to
do,
destroys
right
right.
C
D
D
C
Kind
of
like,
what's
the
use
case,
where
you
add
in
a
hook
later,
on,
like
a
destroy
hook
later
on
right,
and
if
you
weren't,
if
you
weren't
tracking
the
destroyed
before,
do
you
really
need
to
destroy
the
track
it
for
those
ones
either?
Right
maybe
like
it
may
be
perfectly
fine
that
you
only
track
new
stuff.