►
From YouTube: 2023-02-07 meeting
Description
Instrumentation: Messaging
A
B
C
All
right,
we
can
probably
go
ahead
and
get
started
at
this
point,
so
just
for
the
agenda
just
kind
of
any
carryover
conversation
from
the
EU
Sig
we'll
do
a
quick
layer,
assessment,
update,
I
think
we're
making
progress
there,
but
node
and
python
are
still
open.
C
A
quick
update
from
the
spec
I
think
we're
getting
closer
to
a
point
where
we
can
make
a
proposal
to
like
actually
change
the
spec
attributes
and
things
like
that.
I
think
Tyler
custom
notes
around
there
and,
if
there's
any
specific
items
we
want
the
EU
Sig
to
address.
So
we
can
add
those
kind
of
as
we
go
any
other
top
of
Mind
things
for
anyone
here.
D
Okay,
hello
yeah-
this
is
Carlos.
Sorry
I
may
have
to
leave
in
20
minutes,
but
something
that
I
wanted
to
raise
is
that
we
need
somebody
from
this
group
to
become
an
approver.
Probably
one
of
the
things
is
that
we
need
more
reviews
from
Tyler
Benson
Casa
PR,
to
update
the
specification,
and
even
though
a
few
members
of
this
group
have
approved
that
we
need
more
and
I
were
trying
to
get
more
people
to
review
that,
and
you
know
like
they
feel
that
they
are
not
experts.
D
So
one
possibility
is
to
add
somebody
else
from
the
TC
besides
me,
so
we
can
do
reviews
faster,
but
I
think
it
could
be
also
good
to
try
to
get.
Somebody
from
this
group
become
an
approver.
D
C
Yeah
more
TC
members
would
be
nice.
I
know
we
kind
of
have
an
ongoing
project.
Blockage
issue
from
the
lack
of
TC
members,
so
I
guess
I'm
not
super
hopeful
that
we'll
get
others
from
that
in
any
kind
of
group.
Thoughts
around
that
thanks
for
raising
that
shoe
Carlos.
A
Are
you
just
looking
for
more
people
to
comment
on
that
PR
or
are
you
specifically
needing
spec
approvers?
Yes,.
A
D
Was
trying
to
one
of
the
issues
regarding
that
is
that
people
are
not
experts
in
these?
This
is
not
like
a
theory
common
stuff
like
traces
like
SDK
stuff,
so
it
seems
that
for
some
people
this
is
not
super
specialized.
What
kind
of
specialized
item
and
that's
why
they're
not
enthusiastic
or
they
are
putting
that
at
the
end
of
their
backlog?
You
know.
C
Yeah
I'd
say
for
Tyler's
PR
to
I'll,
go
ahead
and
open.
It
I
think
it's
relatively
Lambda
specific,
so
maybe
the
expectation
is
I
think
Anthony
yeah
Anthony
reviewed
it.
So
do
we
do
we
need
more
spec
approved
reviewers
to
get
like
to
the
merge
point
or
Carlos.
What
are
we
looking
to
do?
It
seems
like
most
of
the
contributors
have
looked
at
it.
Yeah.
D
I
mean
that's
a
like
I
think
that
that's
similar
will
depth
and
the
idea
is
of
we
can
once
we
get
approval
approvals
from
members
of
the
working
group.
That's
a
good
signal,
but
then
we
may
get
stock
forever.
You
know,
and
tomorrow
the
TC
has
a
Converse
will
have
a
conversation
about
this,
trying
to
send
more
TC
members
or
adding
more
TC
members.
D
So
we
can
have
more
people
involved,
but
yeah,
there's
always
a
small
I,
don't
know
whether
it's
small
or
not,
that
we
may
need
to
add
people
here
and
experts.
This
is
how
we
move
faster
with
metrics
as
well.
We
are
we're
adding
you
know,
people
who
are
experts
in
metrics.
D
D
So
for
the
repo
itself,
we,
you
have
to
be
an
experica
Brewer
and
actually
I
could
say
that
Tristan
or
Anthony
are
technically.
You
know
qualified
to
be
that.
So
you
know,
because
you
have
been
involved
very
involved.
Also
Alex,
but
Alex
probably
is
more
involved
in
the
collector
these
days,
but
yeah.
So
just
somebody
actually
has
their
approver.
D
You
know
qualification,
let's
say.
A
D
You
actually,
we
can
talk
offline,
but
basically
we
can
request
that
either
I,
don't
remember
whether
me
or
so
like.
You
need
two
sponsors
who
are
in
the
approvers
and
we
can
recommend
you
to
become
one
and
mentioning
what's
your
area
of
expertise
and
that's
it.
Let's
talk
offline,
if
you're
interested,
are
you
interested
in
this
company.
A
Yeah,
it's
like
yeah,
you
know
it's
going
to
be
important
to
move
this
forward
and
it's
also
a
thing
that
I
think
I've
discussed.
We've
discussed
previously
with
with
jagrin
as
well.
Please
suggested
it's
an
area
where
it
could
be
more
involved
in
the
project.
C
Yeah
so
speaking
of
subject
matter,
experts
I
guess
it's
kind
of
tough,
but
should
we
get
Azure
and
Google
folks,
as
part
of
this
I
feel
like
it's
weird
to
make
them
a
spec
approver
when
they
only
care
about
one
specific
piece?
So
maybe
there's
just
like
a
separate
group
that
needs
to
be
set
up.
That's
more
subject
matter.
Expertise
focused
because
I
yeah
yeah.
D
I
mean
like
such
people
like
the
ones
you
mentioned,
that
are
all
interested
in
a
small
piece
of
open
Telemetry.
They
can
perfectly
be
part
of
this
working
group
and
they
can
give
their
symbolic
approval
and
that's
all
you
know
so
they
don't
I,
think
that
as
long
as
they
are
providing
actual
value
like
feedback
or
PR's,
that's
totally
fine,
it's
just
at
by
the
end
of
the
day.
We
need
yeah
people
who
are
actually
you
know
that
are
have
been
part
of
open,
Telemetry
or
not
having
been
part
but
like
they.
D
A
D
A
C
Yeah,
no
totally
aligned
with
that
I
guess
it
just
seems
like
I'm,
not
sure
the
Sig
expectation
here,
I
think
what
we
can
do
is
get
like
Azure
and
gcp
and
AWS
symbolic
approvals,
But,
ultimately,
I,
don't
like
I
I'll
I'll
I'll
I'll
work
offline
to
find
TC
members
that
could
potentially
approve
this,
but
I
feel
like
this
is
potentially
an
Hotel
spec
issue
of
where
they
need
to
be
doing
more
I
mean
yeah.
A
I
think
it
is
I
would
say
that
the
approval
from
the
invited
experts
is
more
than
symbolic,
though
right.
It's
it's
not
just
that
yeah
they've
given
their
thumbs
up.
That's
a
it's
a
signal
that
carries
a
lot
of
weight
for
the
the
people
who
are
non-experts
back
approvers,
who
have
the
context
of
all
the
rest
of
this.
If
they
can
say,
okay
I
could
see
how
this
works.
It
doesn't
seem
to
impede
on
the
rest
of
the
spec.
A
The
the
experts
who
know
this
area
specifically
have
said
that
this
is
good
for
for
that
area.
So
that's
that's
a
strong
signal
to
the
spec
approvers
that
it
is
indeed
good
for
this
particular
use.
D
Yeah
yeah
I
agree
with
that,
and
some
more
context
about
this
is
can
be
found,
for
example,
in
the
probabilistic
sampler
stop.
That
was
added
a
cenotype
and
basically
a
lot
of
people
who
are
experts
in
these,
like
Voldemort
from
Dana
Trace.
They
are
not
involved
in
open
Telemetry,
but
they
did
a
lot
of
reviews
and
once
they
were
happy
even
if
they
were
not
approvers,
they
just
approved
the
pr
saying,
like
from
an
expert
perspective,
that's
good
to
go
so
once
we
have
like
you
know,
because
that
part
that
was
long
and
complex.
D
Once
we
had
a
pair
of
like
two
or
three
approvals
from
such
experts,
then
we
can
actually,
you
know,
touch
it
down
to
to
Hotel
two
Hotel
members.
You
know,
so
it
could
be
something
like
that.
Like
first
we
get
agreement
with
Azure
or
gcp
people,
for
example,
especially
for
bprs
for
long
PRS,
and
once
that's
done,
then
we
just
like
me
and
Anthony
or
anybody
you
for
me
to
see-
would
just
you
know,
double
check
that
it's
looking
good
and
we
just
merge
that.
C
Okay,
I
wonder
if
there's
something
that
can
be
done,
maybe
from
like
a
labeling
perspective
to
say,
like
SME
and
put
you
know,
received
or
not
not
present
yet
or
something
like
that.
So
you
straw
a
signal
to
say
that
this
is
you
know
the
sigs
ready,
I
guess
for
full
approval,
but
okay
yep
thanks.
Thanks
for
raising
that
issue,
I'll
talk
to
Tad
about
it,
a
bit
as
well,
I,
guess,
Tyler,
a
quick
question
for
you
or
any
Azure
gcp
approvals
needed
on
this
current
one.
E
In
yeah,
this
one
in
particular
is
very
AWS
specific,
so
I
don't
think
that
gcp
or
Azure
need
to
weigh
in
on
it.
I
mean
they're
welcome
too,
but
yeah.
C
Okay,
that's
qualifying
comment:
cool
yeah
thanks
for
it.
Thanks
for
raising
that
issue,
Carlos
definitely
follow
up
with
Anthony
and
see
if
he
can
get
him
set
up
and
I
guess
others.
If
you're
interested
reach
out
to
Carlos
as
well
now
ping
Tristan
I
think
he
said
he
wasn't
gonna
be
able
to
make
this
meeting
so
I'll
see
if
he's
interested
in
it
too.
C
E
So
the
EU
meeting
was
pretty
brief.
E
It
was
mainly
just
me
and
the
epsagon
folks,
the
I'm,
trying
to
remember
I
think
that
there's
a
little
bit
of
concern
about
how
the
the
current
architecture
is
around
reporting
about
how
The
Collector
blocks
on
Lambda,
reporting
and
I
wasn't
aware
of
how
the
architecture
was
and
so
I
suggested.
We
take
that
to
slack,
and
so
then
we
had
that
discussion
in
slack
a
little
bit
over
the
past
couple
days.
E
Yeah
I
I'm,
not
sure.
If
there's
a
better
solution
there
other
than
you
know,
if
there's
an
option
to
reduce
blocking
by
The
Collector
potentially
via
you
know,
offloading
the
data
collection
onto
something
like
sqs
or
something
like
that,
but
yeah
that
that
was
the
the
main
discussion
there.
C
So
quick
kind
of
note
on
the
layer,
assessments,
Java
and
net
have
been
done,
but
node
and
python
are
still
open.
I
think
Cossack
just
posted
a
link
to
the
python
equivalent.
So
we
should
just
be
missing.
Yeah
JavaScript
at
this
point,
so
making
really
good
progress
there
and
we
can
start
doing
some
of
the
unification
work
as
well.
C
Any
any
comments
on
that
the
the
pr
for
JavaScript
is
in
the
repo
as
well.
Oh
awesome,
the
PRS
open,
perfect.
C
Okay,
spec
update
so
Tyler
ever
feeling
you.
You
have
some
some
notes
here.
If
you
want
to
either
share
your
screen
or
just
I,
think
I
have
the
sheet
open,
I
heard
it.
E
Yeah
I
assume
you're,
referring
to
me
yeah,
so
the
past
week
you
know
I've
been
collaborating
on
this
spreadsheet
with
a
bunch
of
folks
I
there
there
was
a
separate
tab,
I
merged
into
this
one,
but
basically
the
the
column
on
the
far
left
are
the
existing
terminology
that
you
can
see
in
the
spec
right
now.
Column
C
is
the
stuff.
Are
the
new
names,
as
specifically
the
the
ones
in
bold
are
the
ones
that
are
a
change
from
the
column,
A
and.
B
E
I
think
that
for
the
invocation
ID
you
know
there
there's
alignment
there
platform
trigger
was
a
fairly
new
one
that
I
just
added.
Let
me
see
if
I
can
get
you
permission,
so
you
can
comment
on
that.
Really
quick.
C
I
was
requesting
edit
access,
which
might
be
different.
B
E
E
So
the
platform
trigger
is
one
I
just
added
this
encapsulates
like
the
Google
event,
type
and
I.
Think
Azure
has
a
a
notion
of
a
type
as
well.
Aws
does
not
make
that
super
explicit
I
mean
the
events,
have
an
implicit
type
but
I
think
and
Anthony
correct
me
if
I'm
wrong,
but
you
kind
of
have
to
into
it
what
the
type
of
the
event
is
based
off
of
the
structure
of
the
event
or
the
configuration
of
of
it.
E
And
then
so
the
the
Arn
I
feel
like
is
a
pretty
useful
thing
to
include,
but
there's
not
really
an
analog
for
it
in
the
other.
The
other
platforms
that
said
I
also
didn't
want
to
prevent
such
from
being
added
in
the
future,
so
per
Carter's
suggestion
I
just
labeled
that
as
a
resource
I,
don't
know.
If
that's
you
know,
gonna
confuse
or
conflict
with
generic
like
the
the
general
resource,
Hotel
terminology,
but
there's
that
and
then
I
think.
The
other
question
sorry
go
ahead.
C
Yeah
I
was
gonna,
say
this
is
probably
a
question
for
David,
so
does
gcp
have
an
equivalent
I'm.
Assuming
Azure
has
something
like
this.
It's
essentially
like
a
resource
keyword
saying
you
know
this
is
the
type
of
resource.
This
is
the
availability
zone
or,
like
you
know
the
region
that
it's
in
and
then
it
also
has
like
the
function
name
and
some
additional
kind
of
info.
So
this
could
probably
be
split
out
into
multiple
kind
of
columns,
like
maybe
like
region
function.
B
Yeah
so
for
us
I
think
it's
all
like
a
URL
so
like
you
could
actually
go
visit
the
URL
and
look
at
your
resource,
so
you
might
have
like
I'm
more
familiar
with
the
GK
UI
and
stuff,
but
like
yeah,
there's
something
like
this:
you
can't
get
it
directly.
You
would
have
to
assemble
it
from
the
function,
so
we
wouldn't
like.
We
probably
would
never
populate
it,
but
like
in
theory
something
similar
could
exist
for
gke
or
sorry
for
gcp.
C
Interesting
so
maybe
Tyler
we
might
need
to
deconstruct
the
AWS
equivalent
depending
on
how
weather
does
it
too,
if
they
both
kind
of
make
theirs
separate
kind
of
columns
or
something
like
attributes,
or
something
like
that.
C
C
Account
ID
so.
C
Like
an
open
question
for
like
how
to
treat
resource
metadata
since
and
maybe
decompose
it
or
not,
decompose
it,
okay.
E
And
then
the
other
one
on
the
last
one
with
regards
to
the
max
memory,
so
Alex
had
a
question
about
the
the
units
on
that.
Currently
it's
defined
in
megabytes,
but
that's
apparently
somewhat
inconsistent
with
how
the
broader
specification
generally
to
each
units.
E
The
main
thing
I
I
did
here
is
previously:
it
was
just
max
memory
without
the
units
being
defined
in
the
label
and
I
generally
prefer
to
have
units
included.
So.
A
A
I
mean
David
correct
me
if
I'm
wrong,
but
isn't
this
part
of
the
Prometheus
conversion
as
well
is
that
we
take
units
and
pull
them
out
of
the
name
or
put
them
into
the
name
when
converting
to
and
from
Prometheus,
but
we
don't
keep
them
in
the
name
when
it's
in
P
data.
That's.
B
Correct
so
the
open
metrics
form-
oh
good,
I'm,
still
unmuted
nice.
The
open
metrics
format
requires
that,
if
you're
going
to
attach
unit
as
metadata
to
your
metrics
that
that
unit
string
also
be
present
in
the
name
and
be
placed
as
a
suffix
but
before
other
suffixes,
so
it's
kind
of
confusing.
B
A
I
think
the
question
is
whether
we
should
have
the
units
in
the
name
in
our
semantic
inventions
for
open,
Telemetry
I.
Think
that's
not
the
the
practice.
We
typically
do.
I.
B
E
So,
if
that's
the
case,
then
the
the
question
so
okay,
so
we
don't
include
the
the
units
there.
What
should
we
Define
the
units
as
though
currently.
B
A
B
A
B
But
it
doesn't
include,
for
example,
the
version,
so
it's
it's
not
quite
like
a
super
unique
one,
but
it
does
identify
it
at
a
point
in
time.
C
B
B
C
B
David,
yes,
so
there
is
a
an
organization
Concept
in
gcp
as
well,
and
also
we
gcp
does
have
project
numbers
as
well.
That
probably
look
more
like
what
you
were
just
showing
for
AWS
but
generally
project
ID
is
used
more
because
it's
human,
readable
and
they're
both
unique
with
at
least
within
an
organization.
B
C
Okay,
well
Tyler,
it
seems
like
we
get
so
we
get
the
location,
information
and
the
project
ID
information.
So
maybe
we
can
decompose
that
since
we
already
have
the
function
name
as
well,
so
maybe
let's
just
gets
split
into
like
you
know
fast
dot,
account,
ID
and
then
maybe
faz.location
or
something
like
that.
E
So
I
I'm
not
clear
where
so
I
missed
the
the
part
where
you
said
that
this
this
URL
gets.
How
do
you
access
that
URL
again.
B
You
don't
I'm
just
saying,
like
the
concept
exists,
of
a
identifying
resource
in
gcp
like
this
is
how
we
identify
stuff
in
Google
uis.
But
it's
not
something
that
you
can
discover
from
within
your
function,
but
you
could
discover
all
the
pieces
of
it
and
reconstruct
it
if
you
really
wanted
to.
But
this
is
the
closest
equivalent.
C
B
E
Aws
user,
but
from
my
experience
having
the
Arn
errands,
are
fairly
common
through
AWS
and
used
throughout,
so
I
feel
like
at
least
in
AWS
land.
That
composed
string
is
fairly
useful.
E
So
that
that's
why
I
was
saying
include
it,
but
I
I
don't
feel
like
we
have
to.
If,
if
we
don't
feel,
I
mean
what
Anthony?
What
do
you
think
like?
E
Do
you
feel
strongly
about
including
Arn
as
as
a
complete
string,
or
do
you
feel
it's
better
to
just
leave
it
off
and
include
the
individual
elements.
A
A
Am
kind
of
torn
like
you,
I
see
the
value
of
it,
but
if
everything
else
is
already
available
and
has
to
be
made
available
in
separate
pieces
for
broader
compatibility,
maybe
it's
not
the
end
of
the
world
to
not
have
it
in
a
single
place.
C
Seems
like
it's
still
yeah
kind
of
an
open
question,
so
maybe
we
should
actually
I
can
make
an
action
item
to
reach
out
to
some
customers
and
get
a
sense
of
what
they
would
find
helpful.
I'm,
not
sure
if
we
would
want
to
do
something
similar
with
like
gcp
and
Azure,
because
I
can
talk
to
Rohit
about
getting
some
Azure
feedback.
C
David,
not
sure.
If.
E
You
have
some
questions,
another
option.
The
other
option
we
could
do
is
it's
much
easier
to
add
than
remove.
So
maybe
at
this
stage
we
say
we
just
remove
it
and
then
add
it
back
in
later.
If,
if
there's
enough
interest.
A
Yeah,
the
other
thing
I'm
thinking
is,
is
maybe
we
defined
it
in
the
alternative
and
say
in
instrumentation
should
have
one
of
either
this
unique
single,
unique
value
like
an
arm
or
these
sets
of
values
that
enable
uniqueness.
B
Well,
correct
me
if
I'm
wrong,
but
it
almost
feels
like
it's.
It's
a
convenience
for
users,
if
they're
foreign
I
sort
of
feel
like
it's
useful
for
everyone
to
have
access
to
the
the
common
set
of
like
yeah.
What's
my
ID
or
what's
my
function
name,
but
it
maybe
it's
just
like
a
helpful
extra
for
people
on
AWS,
where
Aaron
is
like
a
common
thing
to
have.
B
I
don't
know
if
I
have
a
strong
opinion
on
that.
Ideally,
it's
named
like
similarly
to
other
attributes
that
are
part
of
the
set,
but.
E
B
A
D
C
E
So,
on
the
the
first
tab
of
the
spreadsheet
I
have
a
link
to
it
down
at
the
bottom.
E
On
this,
it
actually
click
it.
You
have
to
click
the
the
first
one,
okay.
C
It
already
has
Regen,
that's
account
ID.
C
Were
you
missing
some
sort
of
like
resource
like
specific
attributes,
saying
like
hey,
this
is
a
Lambda.
Essentially
what
would
be.
B
Like
it
seems
like
clouds,
in
addition
to
providing
you
with
like
small
pieces
of
metadata
like
what's
my
function,
name
or
you
know,
what's
my
container
name
or
whatnot,
we'll
give
you
like
a
full
resource
ID
in
in
gcp
they're
called
resource
IDs.
What
does
RN
stand
for.
B
A
I
meant
it's,
my
understanding
is,
it's
meant
to
be
like
slanted
with
urn.
E
I,
like
it
I
I,
crossed
it
out
on
the
spread,
spreadsheet
and
updated
the
the
max
memory
to
bytes.
C
C
C
E
E
So
I
this
is
probably
a
better
question
for
for
David,
because
I
think
that
that
is
going
to
be
more
interesting
in
the
case
of
like
Google,
where
the
the
trigger
is
going
to
be
something
very
a
little
bit
more
specific.
If
you
click
on
the
gcp
tab,
I
think
it's
got
an
example.
E
But
anyway,
so
yeah
on
on
there,
it
shows
how
you've
got
like
the
the
google.cloud.pub
sub
topic
this
anyway.
So
it's
got
this
big
long
string
that
defines
what
type
of
a
message
it
is
in
a
little
bit
more
specific
way.
So
for
us,
we're
gonna
say
that
the
the
trigger
is
going
to
be
just
the
generic
Pub
sub,
for
example,
and
then
the
the
platform
trigger
is
going
to
be
that
whole
string
is
that
correct,
David,
I.
B
E
Okay,
I
I
was
adding
that
mainly
because
it
felt
like
it
would
be
a
useful
attribute
to
have
rather
than
just
Pub,
but
if
you
don't
think
so,
then
I'm
totally
fine,
just
dropping
it.
B
It's
funny
I
think
it's
possible.
It
may
be
useful
I'll
I'll
circle
around
with
I
think
their
name
is.
B
Sagar
yeah,
so
they'll
know
if
that
matters
or
not.
The
other
thing
we
can
do
is
if
we
see
an
event
type
string,
that's
well
known,
like
publish
message
and
that
fits
into
the
conventions.
We
can
map
it
to
the
hotel
string,
and
if
we
find
one
that
isn't
known,
we
can
just
stuff
it
there.
If
someone
has
like
a
really
unique
event,.
E
I
mean
I
I,
totally
expected
like
instrumentation,
to
use
that
the
platform
string
to
map
to
the
hotel
string,
but
there's
going
to
be
a
mapping,
and
you
know
one's
going
to
be
a
whole
lot
more
of
a
rich
definition
than
the
other.
Thank
you
one's
going
to
say,
hey.
This
is
a
pub
sub
message,
but
you
know
the
the
platform
string
might
say
you
know
this
is
a
a
publish
or
sorry.
This
is
a
topic
or
this
is
a.
This
is
an
HTTP
request.
E
You
know
that's
a
bad
example
right,
a
Kafka
Pub
sub,
or
this
is
a
an
sqs,
Pub
sub
or
SNS
like
something
that's
going
to
be
a
little
bit
more
descriptive.
E
Yeah
and
again
like
out
of
abundance
of
caution,
maybe
it's
better
to
just
leave
that
off
for
now
and
revisit
that
if
there's
interest
and
people
find
it
missing,.
B
I
would
say:
that's
my
general
preference
is
to
start
with
things.
We
know
users
are
going
to
care
about,
okay
and
maybe
add
on
as
needed,
but
I.
E
C
And
we'll
also
want
to
drive
the
the
cloud
change
too.
I
think
that's,
maybe
orthogonal
but
it'd
be
nice
to
have.
E
B
C
C
Cool
well
I
think
we.
We
had
a
really
good
discussion
today,
not
sure
if
we
have
any
open
items
for
the
EU
agenda
and
maybe
just
Cloud
attribute
I
mean
it's
probably
East,
we'll
discuss.
B
And
okay
and
we
should
switch
to
fast
dot
instance,
if
we're
using
it.
It
looks
like
because
I
think
we
were
using
fast.ib
or
we
are
using
fast.ib.
The
way
fast
dot
instance
is
currently
described.
E
Not
it
is
I
think
faster
instance
was
already
there
right.
That's
not
something.
E
Yeah
fast
instance
is
also
there
in
the
the
specification
for
it's
listed
in
resource,
so
so
part
of
the
actually
yeah
both
of
those
are
there.
So
this
is
where
I'm
looking
at.
E
So
this
lists
both
the
ID
and
instance
in
the
spec
here
ID
is
listed
as
the
Arn
instance
is
more
like
a
unique
identifier
per
to
like.
If
you've
got
multiple
instances
running,
each
will
have
a
different
instance.
Id.
B
E
So
I
noticed
that
this
ID
doesn't
have
the
at
least
the
the
Arn.
There
does
not
have
the
version
number.
E
Anthony
you'd
probably
know
more
about
this,
but,
like
I,
think
that
you
can
have
both.
You
can
have
a
an
Arn
that
references,
a
version
number
and
you
can
also
have
an
errand
that
does
not
and
they're
both
valid.
A
I
think
to
not
have
an
Arn.
You
have
to
have
an
alias
it's
like
you,
give
it
a
constant
name.
That
then
gets
pointed
at
a
particular
version
of
the
function
got
it.
E
So
I
guess
in
this
case
ID
is
like
a
combination
of
the
name
version
and
maybe
region
potentially.
C
C
C
E
I
think
the
cloud
specification
semantic
conventions
already
has
a
lot
of
those
defined
deconstructed.
So
you
know
it
seems
like
there
would
be
a
fair
bit
of
duplication
between
the
two.
It.
C
E
C
Yeah
yeah
I
guess
we
do
theoretically
just
get
the
information
kind
of
and
a
decompose
manner
already.
Then
we
have
to
region
ID
the
account
ID
the
function
name
then
I
guess
it's
just
the
usability
of.
If
users
want
this
readily
available.
C
E
E
Does
it
make
sense
to
start
working
on
spec
changes
around?
This?
C
Okay
and
we
can
also
draw
like
a
reviewer
eyes
to
it,
so
I
can
share
I
like
the
Azure
folks
and
get
their
feedback
on
the
pr,
and
you
know
get
that
in
GitHub
too.
So
I
think
it
probably
makes
sense
to
start.
You
know
kind
of
focusing
attention
using
a
spec
PR.
B
E
I
can
get
started
on
that.
Then
thanks
everyone
for
the
the
extra
eyes
on
this.
C
Yeah,
of
course,
I
think
we're
pretty
much
that
time,
so
we
can
go
ahead
and
call
it
I.
Think
we've
we've
covered
every
topic
unless
you
want
to
shout
suddenly
and
raise
an
answer.
C
Okay,
silence
is
acceptance,
so
that's
good
enough
for
me.
Yeah
I'll,
see
y'all
later
have.