►
From YouTube: 2022-11-09 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
A
B
A
A
Thank
you,
Martin
said
he'll
be
late.
I
put
one
question
there
based
on
a
message
on
slack
Channel,
I
I
wasn't
sure
about
it.
So
I
thought
I'll.
Ask
you
folks.
Do
you
know
if
the
performance
timing,
API
in
any
way,
can
tell
you
that
a
page
was
loaded
from
Cache
or
a
resource.
A
D
A
20
milliseconds
also
sounds
High
to
me.
C
I'm,
pretty
sure
that
it
was
something
like
that
it
was
based
on
our
the
size
of
the
CSS
and
the
JS
packages
were
downloading.
Okay.
C
Got
it
but
yeah
it
really
yeah
if
you've
got
small
bundles,
then
you're,
probably
gonna
change
that
but
for
identity
at
the
time
it
was
like
quite
large
bundles
and
we
tried
to
aggressively
make
them
teachable.
Okay,.
A
C
Yeah
I
didn't
get
back
to
those
yesterday.
Sorry
I
haven't
looked
any
comments.
A
Yeah,
okay,
so
we
we
could
discuss
the
event.data.
C
Was
the
plan,
so
it
get
some
feedback
from
them.
One
of
the
things
we
talked
about
was:
do
we
want
to
add
the
schema
originally
I
started,
adding
event,
dot,
schema
or
event.data
schema,
and
then
I
decided
to
pull
it
out
thinking
it
was
going
to
be
caused
too
much
churn.
C
But
if
you
look
at
the
cloud
of
it,
they
have
a
a
data
schema
which
explicitly
describes
the
data
so
that
third
parties
can
Define
their
own
and
then
provide
it
actually
like
a
lot
of
the
stuff
they've
done
in
that
cloud.
Events.
D
C
Yeah,
but
they
do
have
a
bunch
of
additional
stuff
which
we
haven't
discussed
but
would
be
useful,
like
you
can,
instead
of
schema,
you
can
see
what
type
it
is,
or
you
can
say
how
it's
encoded,
so
you
can
specify
that
the
data
is
encoded
in
a
completely
different
way.
A
Okay,
I
think
we
might,
you
know
I
I
feel
like
we
might
need
to
put
something
saying
you
know
what
should
not
go
into
event.data.
A
So
some
of
these
parameters,
additional
high-level
attributes
like
the
schema,
URL
or,
let's
say
Cat
event,
category
anything.
That's
already
present
in
in
the
cloud
event.
Spec
are
the
elastics
common
schema
I
think
did
they
could
still
go
at
the
top
level.
C
Well,
really,
the
way
I
worded.
It
is
depending
on
the
domain
name
combination.
So
the
domain
really
defines
where
the
data
goes,
and
this
is
because
we
Slide
the
open
question
about
exception.
C
Do
we
just
do
we
say
we
want
an
exception
event
and
that
exception
event
puts
it
into
Data,
or
do
we
say
the
exception
event
puts
it
into
the
top
level
attributes
and
even
in
the
cloud
events
definition
they
actually
Define
that,
depending
on
the
the
type
of
event,
you're
you're
doing
that
you
could
have
a
couple
of
or
required
attributes
as
well
as
optional
ones.
C
A
But
they
they
better
be
consistent
right,
like,
for
example,
for
if
you
take
the
exception
event
today,
all
the
exception
attributes
are
top
level
attributes
yep.
A
So
do
you
think
they
should
remain
as
is
or
they
should
move
into.
Some
of
them
should
move
into
the
event
on
data.
That's
just
actress.
C
I'm
in
two
minds
for
this,
so
it
would
be
nice
to
have
everything
consistent
and
say
everything
is
in
data
but
I'm,
mindful
that
exception
on
span
has
already
been
defined
and
even
the
logs
I
think
has
already
been
defined.
So
if
we
did
this,
that
would
be
a
breaking
change
for
some
Downstream
systems,
so
I'm
always
decline
like
I,
even
Define.
The
vendor
data
as
optional
for
that
purpose.
So
we
could
say
domain
is
browser,
name
is
exception
and
then,
for
an
exception.
You
just
have
the
top
level
stuff.
C
If
we
want
to
do
enhancements
to
any
to
reporting
that
in
you
know,
browser
or
mobile
or
whatever,
that's
potentially,
where
we
could
move
that
stuff
into
Data,
because
it's
currently
not
defined
the
whole
point
being
withhing.
Well,
what
I'm,
trying
to
say
is
semantic
conventions
for
event.data
are
defined
at
the
domain
name
combination,
not
at
the
Global
semantic
conventions,
level.
A
Okay,
we
can
discuss
this
further
in
the
last.
Let's
see
what
they
say.
C
A
Okay,
we
can
talk
about
the
next
topic
say
that
I
think
you
put
that
Android
client
Library
update
from
Splunk.
E
Update
we
we've
basically
been
using
the
Splunk
Android
GitHub
code
base
and
sort
of
refactoring
it
to
really
pull
out
the
non-bender
specific
things,
and
the
goal
is
to
Upstream
it
early
next
year
to
the
hotel
community
so
that
we're
that's
kind
of
the
plan.
So,
basically
the
idea
is
to
donate
a
part
of
that
Library
as
we're
refactoring.
It
I
think
the
timeline
we're
sort
of
looking
at
is
January
or
February
I
think
we're
pretty
confident
we
could
hit
those
timelines.
E
I
did
speak
to
Morgan
yesterday,
I,
don't
know
if
he's
almost
called
yet.
Oh,
he
is
okay.
Oh
team
working,
oh
yeah,
just
wanted
to
give
an
update
on
the
Android
piece
yeah.
If
there's
any
feedback,
I
can
leave,
I
can
leave
the
library
link
here.
We're
happy
to
continuously
get
feedback
on
on
the
approach.
We're
taking.
A
Okay,
yeah
thanks
for
the
update,
so
one
related
topic
is
that
I
there
is
a
swift
Sig.
D
A
In
open
Telemetry,
so
there
I
have
started.
You
know
discussing
the
same
discussion
that
we
are
having
here
about
identifying
some
common
events
and
then
defining
the
schema.
We
we
just
started
last
week
so
I
think
if
it's
helpful
to
Android
too
I
think
they.
You
know,
you
know
either
you
or
you
know
more
folks
could
chime.
D
A
There
any
yeah
yeah
so
so,
basically,
the
the
idea
is
that
there
are
events
that
are
not
necessarily
relevant
in
a
span,
so
think
of
a
random
error
are
in
the
case
of
a
mobile
lab.
You
can
detect
the
network
change
right,
you,
you,
people
move
from
Wi-Fi
to
Cellular
or
the
other
way
and
then
the
applications
crash.
A
So
you
want
to
report
a
crash,
so
these
are
not
modeled
as
spans,
so
they
need
to
be
modeled.
As
you
know,
independent
events.
A
E
Okay,
thank
you
for
that
directions.
We'll
have
to
sort
of
explore
if
we
can
also
attend
that
as
a
team
yeah.
Thank
you
are.
A
There
any
yeah,
yeah
I,
think
you
can
attend
this
I
mean
they
can
attend
this
meeting
too,
but
I
think
yeah
I
think
either
this
meeting
or
these
the
Swift
segmenting
either
is
fine.
E
Okay
and
okay,
thank
you.
We
will
talk
internally
and
see
how
we
want
to
pull
people
in.
A
You
can
you
can
find
out
if,
on
the
Android
team
in
Splunk
have
they
started
thinking
about,
let's
say:
reporting
errors
crashes,
things
like
that
which
don't
fit
very
well
into
spans.
Oh
I
think
last
time
very
long
time,
John
Watson.
He
mentioned
that
you
already
have
the
concept
of
events
which
you
know
internally
represent
as
zero
duration,
spans.
E
A
E
Right
so
just
given
sort
of
our
Focus
right
now,
just
to
make
sure
that
we
do
have
some,
we
do
have
the
Android
Library
sort
of
Upstream,
that's
our
core
Focus
right
now,
but
yeah.
We
do
have
a
lot
of
event-like
data
that
we
do
send
as
zero
iteration
spans
today,
including
errors
for
that
matter.
Okay,.
D
A
Do
you
also
report,
let's
say
in
the
middle
of
a
network
request
if
the
network
changes
do
you
report
that
okay.
E
E
The
reason
for
that
is
like,
as
you
mentioned,
you
know,
does
the
network
changes
something
there's
usually
some
unexpected
behavior
on
the
application,
so
we
do
have
Network
change
events
and
I.
Think
we've
added
this
to
the
data
model
that
we've
shared
in
the
document
that
Nev
started.
A
So
it
includes
the
events
for
the
mobile
as
well.
E
Yeah
I
thought
that
was
the
ask
from
nav
some
time
ago
right
to
fill
up
the
document
that
I
think
Ram
was
going
to
review
nav.
Is
that
correct.
C
Yeah
of
the
spreadsheet
and
stuff
was
to
try
and
is
there
anything
that
could
be
shared
across
all
of
them
so,
rather
than
defining,
you
know,
always
only
browser
specific
or
your
domain.
Specific
events
is
there
any
event
that
we
could
share
across
the
board,
so
that
was
the
the
idea
of
the
spreadsheet
is
to
get
browser
mobile.
Anything
we
can
and
see.
If
there's
some
commonality,
I
don't
think
we
are
end
up
achieving
that.
Yet
I
think
we're
just
defined
browser
domain
level
ones.
At
this
point.
E
Sure
I'm
so
sorry
I'm
having
a
lot
of
Zoom
issues
on
my
laptop.
So
that's
why
I'm
not
able
to
share
my
screen,
but
we've
actually
added
everything
into
that
document
and
we
actually
have
public
documentation
on
each
of
these
events
as
well.
So
this.
E
F
E
C
A
Okay,
so
I
think
in
addition
to
sending
as
an
independent
event,
you
know
this
could
also
make
sense
as
as
a
event
on
the
spam.
So
you
know
that
you
know
if
something
failed,
you
know
your
your
request
to
your
network
request
failed
because
there
was
a
network
change.
It
is
easier
to
correlate
but
yeah
that
could
their.
C
I
guess,
except
if
it
goes
into
an
offline,
State,
never
recovers.
You
never
get
expand.
E
Yeah
I
was
about
to
say
that
too.
Instead,
what
we've
done.
A
No,
let's
say
think
about
retries
right.
If,
if
your
client
is
is
retrying
because
it's
unable
to
connect
so
I
think
recently
the
I
think
there's
a
PR
talking
about
adding
a
retry
account
attribute,
and
as
part
of
that,
you
could
say
that
hey
you
know
I
had
to
retry,
because
there
was
Network
failures
and
then
eventually
you
know
the
network.
The
network
actually
changed
from
Wi-Fi
to
cylinder.
E
No
but
I
mean
sure
but
I
think
that's
a
sequence
of
events,
not
necessarily
a
part
of
the
actual
request.
The
request
should
know
whether
you're
on
Wi-Fi
or
cellular,
but
if
the
network
change
has
occurred,
that
might
not
happen
during
a
request.
It
might
happen
before
it
might
happen
after.
A
Yeah
I
think
we
let's
discuss
when
when
we
really
get
into
it,
I
think
yeah.
This
came
up
in
the
Swift
Sig
I.
Think
the
the
iOS
folks
suggested
this
so
I
I'm
not
sure
either.
A
Anything
else
Ram,
you
have
anything.
G
Oh,
apologies
for
John
and
late
I
had
a
quick
drop
off
this
morning
and
then
to
take
care
of,
but
not
yesterday
we
talked
about
how
Mev
issued
a
PR
and
referencing
that
I'm
going
to
be
shooting
a
PR
pretty
soon
on
how
we
want
to
structure
the
events
in
semantic
conventions.
G
G
First,
you
know
draft
of
one
of
the
events
that
we've
already
worked
on
page
view,
for
example.
So
that's
that's
the
plan
so
I'm
working
on
that
other
than
that.
Don't
have
much
updates.
D
G
Yeah
we
could
use
it,
I
mean
yeah
folks.
Here
we
could
use
the
time
to
continue
our
schema
discussions
for
other
events
that
we
haven't
looked
at.
If
people
are
up
for
it,
if
not,
you
know
we
can
go
back
and
work
on
getting
the
missions
that
appears
out
and
and
stuff
I'm
open
for
either.
A
G
There
probably
are
some
open
questions
on
fetch
timings.
I
think,
do
you
remember,
you
know
you
and
I
talked
about
fish
timings.
Thursday
did
we
did
we
decide
we
needed
to
go
back
and
he
was
suppose
with
the
team
or
with
the
grouper.
C
Yeah
I
think
I
think
we
should
we
talk
about
it.
So,
from
my
perspective,
I
think
we
can
actually
drop
fetch
timing
altogether
and
just
say
it's
a
resource
timing.
G
Yeah
thanks
for
helping
jungle,
memory,
yeah,
I,
I,
agree.
Also
I
got
a
little
bit
confused
when
I
saw
fetch
timing
in
that
spreadsheet,
Santosh
I
guess
I,
wasn't
there
for
a
couple
of
I
mean
there's
something.
That's
probably
when
the
group
worked
on
that
so
I
had
the
question
I
opened
up
the
topic
in
there
give
some
response
here
just
a
second
ago.
G
Is
that
something
that
we
can?
You
know
open
up
talk
about
you
know.
I
would
like
to
hear
your
thoughts
Satish
and
are
you
drove
the
pitch
timings.
C
Maybe
I
can
rephrase
so
the
French
tummy,
like
this
event
that
we've
defined
it
so
far
is
saying
it's
got
all
the
fixed
details
which
is
like
you
know,
four
through
eighteen,
which
I
think
is
an
event
on
its
own,
and
then
we
had
20
onwards
or
the
green
block,
which
effect
is
all
the
resource
timing
here,
so
yeah
20
through
to
28,
which
I
think
is
really
just
a
resource
timing
attendant.
C
So
really
what
I'm
saying
is
we
actually
split
these
into
those
two
events,
because
we
already
have
a
resource
timing
which
will
already
encompass
all
that
stuff
and
because
of
the
way
effect
xhr
works.
We're
not
going
to
have
the
resource
timing
for
a
little
bit
of
time
afterwards,
and
we
actually
may
not
get
it
at
all,
because
once
the
resource
list
gets
full,
you
stop
getting
resource.
So
for
us
for
a
long-running
spa.
A
I
I,
if,
if
I
remember
correctly,
what
we
said
is
there
is
a
there.
A
Is
this
timing
events,
so
we
create
one
generic
timing,
event,
object
and
then
Define
all
the
possibilities
in
in
the
form
of
a
table,
so
so,
depending
on
another
attribute
to
indicate
which
type
of
I
mean
what
its
timing
event
refers
to,
whether
it's
for
the
navigation
or
the
resource,
what
type
of
resource,
and
then
you
know
each
of
these
parameters
you
know,
may
or
may
not
apply
so
that
way,
you
know
we
could
further
combine
the
resource
timing
and
the
page
navigation
timing
too,
and
have
just
one
timing
event.
G
Yeah,
okay,
I
think
that
makes
sense.
Then
I
can
take
this
time
of
it.
You
know
move
some
of
the
things
you
know
the
the
non-research
timing,
specific
things,
the
research
standing
things
out
of
here
and
put
them.
We
can
do
it
right
now.
Also
we
can
you.
A
Know
yeah
yeah:
let's
we
can
do
it.
Yeah
I
I
think
with
respect
to
what
goes
into
the
research
timing,
I
think
maybe.
G
H
Since
then,
I
would
I
I
do
Wonder
like
if,
if
it,
if
it
makes
sense
to
combine
into
a
single
timing
event
at
this
point
like
if
you
look
at
the
examples,
so
there
is
so
I
think
I
think
what
I
was
looking
at
is
like
the
common
common
fields
between
the
apis
from
the
w3c,
like
user
timing
and
largest
continental
paint,
paints
the
resource
and
there's
there's
a
bunch
of
them
and
I
think
we
would
have
to
like.
If
you
combine
them
all
into
one
like
you
would
have
to.
H
First
of
all,
you
would
have
to
Define
like
our
own
type
attribute,
and
then
I
also
wonder
like
if
it
just
makes
it
a
little
bit
unnecessarily.
Complicated
like
like
you
know
like
if
we
were
talking
about
schemas
so
like
I,
think
it
might
be
easier
to
just
have
schemas
for
individual
events.
But
that's
just
my
opinion.
So.
A
I
I
I
think
maybe
I
think
you
might
be
right.
I
I
I'm,
suggesting
that
without
you
know,
good
knowledge
on
this
topic,
but
I
I
based
it
on
the
assumption
that
most
of
these
have
you
know
most
of
the
attributes
common.
But
if
that's
not
the
case,
then
I
agree
with
you.
Yeah.
D
A
Will
be
separated,
so
we
could
start
with.
You
know
this
table
Martin
if
you
could
fill
up
this
fill
this
up
further,
because.
G
D
G
A
Stuff,
the
First
Column
Ram.
There
are
more
parameters,
I.
H
Think
you
know
I
think
I
think
I
as
far
as
if
I
remember
correctly,
it's
been
a
while,
but
if
I
remember
correctly,
this
table
was
filled
in
to
this.
My
knowledge.
H
There
might
be
like
so
and
I
think
for
the
most
part,
I
was
going
by
again
by
the
w3c
apis
or
like
navigation
resource
LCP,
and
then,
like
navigation,
then
has
like,
like
a
long
list
of
additional
attributes
that
are
not
in
this
table.
I
think
they're,
probably
like
listed
underneath
somewhere.
H
A
Fetch
start
and
yeah.
H
Those
the.
H
I
think
yeah
I
think
if
you
go
like
they're
below
yeah
I
I
did
not
include
all
of
these
in
the
same
table.
Yeah
yeah.
H
A
G
We
do
we
want
to
do
that
or
you
know.
Are
we
all
in
agreement
that
we
you
know,
we
would
essentially
want
to
create
multiple.
You
know
different
schemas
for
each
one
of
these
fetch
timings.
I
mean
this
is
one
of
these
timings.
A
A
C
Well,
the
resource
coming
specifically
like
a
lot
of
these
are
resource
timings
that
they're
their
own
thing,
but,
like
the
resource
time
is
really
the
the
SSL
stuff
that
is
Santosh
called
The
Connect
domain
whenever
so,
the
only
ones
that
really
I
think
makes
sense
to
potentially
combine
would
be
the
navigation
and
the
resource
timing.
All
these
other
ones
know
so.
Yeah.
C
But
you
know
from
a
back-end
processing
perspective.
It
probably
makes
sense
to
even
keep
the
navigation
and
general
resource
timing
separate
like
for
the
for
the
fetch
resource
timing.
It's
going
to
make
sense
for
a
fetch,
the
instrumentation
to
have
a
have
a
span
covering
all
the
events
for
linking
so,
but
we
probably
also
want
to
have
a
unique
name.
So
it
knows
that
that's
the
resource
timing
for
now
for
the
fetch,
as
opposed
to
a
general
resource
timing,
for
you
know,
loading
up
the
Javascript
file
or
an
image
or
a
CSS.
C
C
G
Yeah,
so
here's
Here's
a
thought
physician
right
in
a
modern
instead
of
going
through
this
table
in
a
public
in
this
table
and
then
transferring
it
over
to
our
schema
and
things.
Perhaps
we
create
a
new
dabs
for
each
of
these
events
that
we
want
to
Define
and
start
listing
them
there.
That
could
be
a
long
list.
You
know
each
even
separated
without
some
things.
You
can
use
this
as
a
starting
point
and
then
put
them
in
the
Excel
spreadsheet
yeah
that
works
for
you,
I
think
you
can
start
there.
C
A
bit
like,
like
we
haven't,
talked
about
web
vitals,
but
it'll
include
some
of
some
of
these
as
well
that
we
might
have
a
you
know,
a
single
event
that
says.
Okay,
we've
got
all
this
data,
especially
when
there's
no
crossovers
for
things
like
that,
but
yeah
we
can
start
with
defining
them
separate,
and
then
we
can
look
at
and
say
well,
it
doesn't
make
sense
for
to
send
FCP
and
LCP
separately,
for
example,.
H
Yeah
I
think
I
think
the
way
when
I
was
putting
this
table
together.
I
was
looking
at
the
w3c
apis
and
there
is
like
the
like:
the
LCP
and
the
resource
and
navigation.
They
have
like
different
subclasses
of
of
like
the
performance
entry
class
and
so
like
the
performance
entry
has
some
common
common
attributes
that
apply
to
all
of
them,
although
some
of
them
may
not
use
them.
So
it's
that's
where
this
came
from.
D
A
Yeah
in
this
section,
maybe
let's
put
each
one
in
each-
you
know
in
a
separate
tab
and
then
we
can
do
the
same
kind
of
exercise
that
we
did
yeah.
C
Because
it'll
probably
be
the
point
of
when
we
start
creating
interpretations
to
to
do
this,
it'll
it'll
probably
then
become
more
logical
about
which
you
know,
because
you
have
a
an
instrumentation.
That'll
require
multiple
events,
and
once
we
get
to
that
point,
we
can
look
at
it
and
say
well.
Does
it
make
sense
to
keep
these
as
separate
events.
G
That's
right,
yeah
I,
agree
now
combining
events,
though
you
need
to
start.
You
know,
I
do.
G
The
the
if
we
combine
we
have
to
come
up
with
a
type,
and
you
know,
for
you
know
the
type
and
things
like
that.
That
might
be
a
little
bit
of
a
complicated
thing.
It
might
make
sense
to
just
have
them
distinct.
Also.
C
G
Okay,
I
just
have
a
quick
thing:
I
haven't
checked
that
or
you
know
you
probably
talked
about
it
already
now.
Was
there
any
feedback
on
your
APR
for
the
data
attribute?
How
is
that
looking.
C
I
I
didn't
get
to
that
yesterday.
After
the
meeting
with
that
internal
team
I
found
a
bug
which
I'll
talk
to
that
with
you
later
so
I
took
took
ages.
I
was
online
with
them
for
ages
and
then
okay,
just
to
dig
into
a
point
out.
Why.
C
I'm
planning
to
bring
it
up
there
in
terms
of
get
feedback
from
them
about
event,
data
and
potentially,
whether
we
want
to
talk
about
event,
schema
or
event,
data
schema,
which
is
as
defined
in
the
Cloud
events,
whether
to
include
that,
because
there's
a
lot
of
stuff
in
the
cloud
events
which
I
agree
with
and
yeah,
potentially
we
could
just
say.
Well,
it's
a
morph
events
into
Cloud
events,
which
we
sort
of
already
are.
We
just
haven't
defined
all
the
fields
and
all
the
different
combinations
they
have
yet.
C
Well,
if
you
look
at
my
yeah
I've
got
a
link
to
the
cloud
events
in
there
yeah
how
the
data
event.data
or
the
you
know
they
just
have.
Data
is
encoded,
so
they've
already
done
a
lot
of
the
heavy
lifting
I
think.
C
So
I
think
I
used
to
love
the
wedding
from
here,
but
instead
of
saying
may
contain
domain
specific
I'm
saying
the
data
is
domain,
specific
for
us
right,
but
yeah.
The
fact
they've
got
the
data
content
type
attribute,
so
they
can
Define
how
it's
encoded
because
they
don't
find
the
transport
level
at
this.
They
just
say
this
is
what
a
cloud
event
would
contain,
but
we're
sort
of
the
next
level
down
from
consuming
this
in
my
PR
I've
defined
how
this
might
get
mapped.
C
As
an
example
I'm
not
saying
that's
the
only
way
I
can
get
mapped,
it's
just
an
example
of
how
it
could-
and
this
is
a
little
bit
more
restrictive
to
than
us,
because
they're
saying
the
attributes
could
only
be
a
to
z,
0
to
9
and
lowercase.
Well,
we
have
camel
case
as
attribute
names
and
I've
been
trying
to
create
as
an
opportunity.
C
So
for
us,
my
PR
is
really
just
taking
their
their
event.
Data
from
this
we
already
have
our
domain
name,
which
effectively
infers
our
schema
at
this
point,
but
yeah
there's
a
lot
of
things
in
here.
That
would
be
nice
to
extend
and
include
them.
C
But
doesn't
have
to
be
anything
so
I.
Think
I,
don't
remember
what
they
say.
The
default
is,
if
you
don't
specify
the
content
type,
the
data
content
type
is
optional.
G
G
G
This
this
would
be
our
the
thing
that
we're
defining
its
schema
version
right,
yeah.
C
We're
referring
this
by
our
combination
of
domain
and
name
today,
but
it
would
also
be
nice
to
have
this
so
that
third
parties
can
Define
their
own,
so
that
back
ends
can
then
know
how
to
infer
or
how
to
decode,
validate
or
whatever
the
the
event.
G
Right,
I
I
think
at
some
point
we
talked
about
it.
It
probably
is
in
Pages
scheme
also
the
version
for
the
data.
You
know
that
we
defined
the
necessary
activities
to
be
defined.
We
needed
that
right,
I
think
we
we
added
that
somewhere.
I
can't
remember.
C
I
know
we
talked
about
it.
Can
you
come
back
to
the
is
the
lack
of
the
version
number
version
one
enough,
but
it
would
always
be
nice
to
have
a
defined
first
yeah.
G
C
D
C
G
C
Not
block
getting
event
data
in
while
we
have
a
big
discussion
about.
Do
we
really
want
schemas,
because
there's
a
already
been
lots
of
discussions
about
schema
and
then
you
have
to
get
into
the
discussion?
Okay.
Well,
what?
How
is
that
schema
defined?
What's
it
looked
like
I?
Don't
want
to
get
into
that
discussion
just
yet.
C
A
C
Yeah,
if
you
got
the
top,
where
they
limit
the
yeah
names
to
A
to
Z,
0
or
the
time
they
do
talk
about
that
it's
to
do
with
back-end
compatibility,
which
is
also
why
they
don't
have
uppercase
in
there
I
see.
G
Yeah
a
lot
of
the
backends,
don't
like
the
dock,
so
they
convert
that
to
underscores
or
dashes
or
something
you
know
we
have
that
going
on
a
lot
done.
A
Okay,
so
yeah
we'll
we'll
discuss
this
in
the
logic
further.
G
The
last
thing
is
at
10
o'clock
right,
10
o'clock,
yeah
I'll,
try
and
see.
If
I
can
you
know
attend
that?
Also,
today,
okay,
just
when
they're
listening
to
what's
going
on
there.
B
A
B
G
A
Okay,
yeah,
because
I
think
we
we
are,
we
are.
We
are
relying
on
that
for
a
whole
bunch
of
attributes.
Okay,.
A
A
I
I
think
it
will
also
be
helpful
to
get
one
TC
members
to
help
us
on
that.
Okay,
I,
don't
know
if
Ted
is
part
of
the
current
TC.
A
A
So
in
this
I
think
Josh
mentioned
that
raise
it
in
the
next
PC
meeting.
F
Were
there
any
updates,
if
you
keep
scrolling
down
anything
since
then?
No,
no,
no!
Okay!
Let's
do
two
things.
Let's
ping
Josh
I
guess
he
already
did
Pink
Josh
there
again
for
an
update
and
then
secondly,
we'll
go
to
the
specsig,
although
I
can't
make
it
next
week.
But
if
you
go
to
the
specsig
Tuesday
morning
at
8,
A.M
Pacific,
that's
the
best
place
to
ask
for
re-review
of
this
or
to
get
more
eyeballs
on
this.
C
I
go
to
that
every
week,
so
perfect,
yeah
yeah!
Do
you
want
to
drop
this
link
into
the
our
notes
here
so
that
I've
got
a
quicker
access
to
it?
A
H
Sounds
good
thanks.
Everyone
yep.