►
From YouTube: WebPerfWG call - 2022 08 04 - User timing conventions
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
So
for
background
in
easy
timing
level,
one
the
they
had
a
recommended
mark-
name
section-
I
pasted
it
here,
but
basically
sites
could
mark
when
they
were
fully
loaded,
fully
visible
above
the
fold
and
the
time
to
user
interaction
and
back
in
those
days
like
events
like
page
load,
which
were
kind
of
used
for
generic
track.
Generic
measurement
of
of
these
things
didn't
fit
all
the
cases
and
we
still
do
think
even
with
core
vitals
that
there's
gonna
be
like
a
lot
of
product.
A
Specific
cases
that
will
sites
will
always
want.
I
mean
that's
why
we
have
user
timing
in
the
first
place.
They'll
always
want
to
have
their
own
points
to
track,
so
some
some
things
that
the
reason
we've
been
looking
at
into
it
in
chrome
is
because
there
are
a
lot
of
sites
that
would
like
to
tell
us
like
specific
points
in
the
page
load
where,
like
a
very
large
javascript,
has
finished,
and
things
like
that
that
help
us
optimize.
A
But
we
thought
this
might
be
useful
for
other
use.
Cases
like
rom
analytics
large
platforms
that
have
to
monitor
a
lot
of
sites
and
automated
testing
tools.
A
B
So
I
have
vague
memories
of
the
history
of
this
and
I
believe
they
were
removed.
Just
for
lack
of
usage.
I
think
there
was
either
some
telemetry
data
that
was
gathered
from
the
chrome
team
way
back
in
the
day
or
one
of
the
rom
providers
looked
at
it
and
we
just
saw
like
extremely
little
use
cases
of
it
and
then
once
we
moved
to
level
two,
it
was
just
a
question
of.
Should
we
continue
bringing
this
forward
or
is
it
just
kind
of
extra
verbiage
in
the
spec
and
then
so?
A
B
From
a
rum
provider
point
of
view,
I
was
always
in
favor
of
keeping
them
in
just
because
I
felt
like
it
could
be
a
useful
thing
for
us
to
standardize
around
and
then
mark
in
similar
ways
and
dev
tools
and
rum
and
stuff
like
that.
B
C
Thinks
it's
really
helpful,
michael.
D
So
one
that
isn't
on
the
list,
although
the
fourth
one
was
similar,
maybe
which
might
be
interesting
to
add-
which
is
something
like
mark
time
to
interactive.
So
there
is
a
metric
called
time
to
interactive,
which
has
a
sort
of
browser-defined
somewhat
artificial
definition.
But
I
know
that
a
lot
of
sites
are
working
to
optimize
the
way
they
define
interactive.
D
So
maybe
there's
like
a
a
fallback
until
hydration
completes,
and
then
you
have
a
more
interactive
mode
or
something
like
that.
So
that
would
be
an
interesting
time
point
because
I
know
it's
like
a
a
hot
topic
of
competition
and
optimization
right
now,
and
it
would
be
interesting
to
see
that
I'm
glad
dan
is
willing
to
share
there.
A
Yeah-
and
I
and
my
team
works
very
very
hard
on
exactly
that-
right,
like
a
standard
implementation
of
measurement,
but
you
know
years
into
the
core
world
vitals
program.
We
are
still
seeing
like
from
internal
performance
teams
at
google,
like
the
v8
team,
for
example.
They
they
really
want
to
know
like
exactly
this
point
in
time
that
it
kind
of
makes
sense
as
a
topic
like
when
can
somebody
edit
a
google
doc
or
something
like
that?
But
it's
not
necessarily
like
some
pages.
Don't
have
the
concept
at
the
same
time,.
E
D
E
C
Thanks,
I
think
I
think
pat's
next
in
the
queue
sorry
I'll
open
it
up.
F
Yeah
thanks,
I
think
part
of
that
can
be
avoided.
If-
and
I
don't
know
that
we
necessarily
want
to
have
like
time
to
x
but
like
the
not
just
the
naming,
but
the
description
beyond
like
mark
app
is
useful
or
first
app
is
first
useful
or
something
like
that
as
long
as
they're
reasonably
scoped,
I
don't
know
that
you'd
ever
want
to
necessarily
use
these
as
comparisons
for
like
site
a
versus
site
b,
but
within
the
scope
of
a
given
site
or
selfishly
on
the
browser
side.
F
If
at
scale
we
want
to
optimize
for
metrics,
and
we
say
hey
if
you
instrument
the
things
you
care
about,
we
will
try
our
darndest
not
to
regress
them
or
even
improve
them.
If
we
can,
if
you
mark
things
that
you
don't
care
about
as
if
you
cared
about
them
well,
sorry,
I
I'd
encourage
thinking
about
this
beyond
just
performance,
though
it'd
be
really
nice.
F
If
we
could
start
to
get
into
things
like
user
engaged
or
user
did
something
that
I
consider
a
business
useful
metric
in
some
way
that
we
can
start
correlating
performance
and
business
metrics
and
optimizing
for
both
rather
than
just
one.
C
B
What
pat
just
said
really
resonates
with
me,
bringing
it
outside
of
just
the
performance
focus
obviously
having
something
like
that
can
give
indicators
to
a
rum
system
that
what
something
that
happened
was
interesting
or
important
to
the
site
in
a
standard
way
that
we
could
pee
off
of
or
take
an
action
or
measure
more
or
report
on,
or
something
like
that.
B
I
think
I
understand
dan's
concern
around
these
being
undefined
intentionally.
Almost,
I
think,
as
long
as
there's
no
carrots
associated
with
how
fast
these
are,
then
it's
okay,
right,
like
I'm
assuming
the
major
search
engines,
aren't
going
to
prioritize
this
just
because
you've
marked
fully
loaded
at
zero
milliseconds,
but
that
is
certainly
a
consideration
here
is
making
sure
that
these
are
intentionally
left
vague,
almost
intentionally
not
comparable
across
sites
usable
by
somebody
that
wants
by
within
a
site
for
any
tools
that
they're
using
to
automatically
key
off
of
devtools
or
rom.
B
Or
what
have
you
yeah?
The
whole
conversation
here
is
very
interesting
because
it
was
part
of
the
original
user
timing,
spec
design,
and
we
thought
back
in
2010
like
oh,
this
would
be
fantastic.
Everybody
is
going
to
use
these
it'd
be
great
if
we
all
standardized
around
them
and
unfortunately
we
just
didn't
see
that
that
usage
back
then
now
way
back
then
before
user
typing
was
a
thing
and
it
certainly
picked
up
a
lot
since
then.
So
I
I
am
in
favor
of
this.