►
From YouTube: 2023-02-08 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
B
B
B
I
think
right
now
we
have,
we
have
had
so
many
discussions
about
specs
and
what
we
want
to
do.
That's
not
the
bottleneck
anymore,
yeah.
Okay,
it's
more
just
getting
things
done.
A
Cool
well
I
had
I
had
like
one
thing
from
yesterday
that
I
haven't
had
a
chance
to
even
touch
just
as
far
as
doing
the
analysis
on
the
semantic
conventions,
but
that
should
happen
hopefully
soon
not
not.
Today,.
B
So
one
thing
that
we've
been
doing
in
this
in
this
group
is,
we
spent
quite
a
bit
of
time
going
going
over
the
semantic
convention
conventions
for
a
browser
for
browser
apps,
and
we
kind
of
settled
on
that
most
most
signals.
Most
style
battery
would
be
captured
as
events
and
we
I
think
we
expect
that
it's
going
to
be
the
same
thing
with
mobile
applications.
B
A
D
Yeah
I
think
in
that
Excel
sheet
I
have
put
most
common
events
event
types.
D
So
even
if
we
just
focus
on
those
I
think
you
know
that
might
be
sufficient.
I
think
the
crash
event
is,
you
know,
seems
to
be
the
critical
one.
D
Especially
if
we
were
to
standardize
across
Android
and
iOS,
we
could
discuss
that
further,
typically,
not
not
typically
sorry
I
think
sometimes
even
Bryce
joins
this
sig
price
is
one
of
the
members
of
the
Swift
Sig.
D
I
can
ping
him
to
join
next
time
if,
if
we
want
to
spend
next
week
on
on
the
mobile
side,.
A
C
A
D
Okay,
perfect
yeah,
then
I'll
bring
rise
to
join
this
next
Wednesday
cool.
D
What
do
you
want
to
discuss
today?
Any
topics
Martin
I
saw
your
PR
was
merged
yesterday,
so
I'll
bring
Bobby
to
you
know,
make
his
changes
further.
Yeah.
B
They
they
said
that,
probably
next
week,
oh
yeah,
yeah,
okay,.
D
In
the
meantime,
if,
if
there
is
any
further
feedback
we
can
get
on
spr,
you
know
that
will
be
helpful
to
avoid
further
delays.
So
maybe
maybe
you
can
try
it
out
yeah.
You
can
start
on
the
exporters
and
then
try
it
out.
That'll,
be
nice
yeah,
just
focus
on
the
logs,
initially
yeah.
A
C
D
Yeah
yeah
yeah
one
other
topic,
nav
I,
don't
know
if
you
saw
messages
from
Riley
on
the
separation
of
events
and
logs.
D
E
Yesterday,
where
he
blocked
the
pr
right,
I
think
Tigger
responded.
I
haven't
seen
any
more
messages
on
that
one.
Yeah
I've
been
off
doing
the
internal
stuff,
mostly.
D
Okay,
I
don't
know
if
he
would
join
the
lock
set
today,
but
otherwise
I'll
be
further
delayed,
yeah
yeah.
We
are
not
so
much
blocked
on
it.
As
of
now
I
think
we
we
need
progress
on
the
acid
attributes,
PR
yeah
more
than
this,
because
I
think
we
already
have
the
changes
before
even
the
spec
is
ready.
E
E
E
I
I
guess
we
had
some
slight
progress
yesterday
on
the
sandbox,
so
we
pushed
it
in
and
we
got
the
API
merging
I
did
notice
after
running
it
getting
open,
Telemetry
bot
to
run
it
that
it
is
deleting
the
main,
readme
and
I
think
that's
because
of
the
way
that
it's
merging
the
two
I
think
the
easiest
way
for
now
would
be
just
to
let
it
delete
it
and
then
I'll
follow
up
with
a
PR
to
put
it
back
and
then
subsequent
mergers
shouldn't
delete
it
and
if
they
do
I'll,
then
try
and
figure
out
a
solution.
E
That's
my
current
thought,
because
what
I
want
to
do
is
push
the
push
it
in
bring
back
the
the
readme,
because
it's
going
to
go
away
and
I
want
to
activate
the
CI
as
well,
so
it
actually
starts
building
the
stuff
that
emerges
so
I
may
or
may
not
get
back
to
that
today.
It's
probably
more
likely
later
in
the
week.
B
E
Yeah
but
yeah,
but
apart
from
that,
once
I
can
confirm
that
you
know
bringing
it
into
the
staging
and
then
bring
it
into
the
into
main
Works
cleanly,
which
it
should
it's
just
a
simple
I
update
that
config
TS
run
it
locally,
make
sure
it
works.
I
bring
quite
a
few
packages
in
so
most
of
the
packages
that
are
already
in
that
config
ts
are
work
for
running
the
build
in
the
test.
E
I
think
there
was
one
left
that
I
had
to
work
through
with
that
is
just
in
case
of
bringing
in
more
packages
or
you
know,
defining
more
packages
and
making
sure
the
automated
tests
work.
There
is
some
code
changes
we're
going
to
have
to
contribute
back
because
of
a
different
version
of
typescript
I.
E
Think
I
went
with
the
walkthrough
I
went
through
a
couple
weeks
ago,
they're
just
typings,
mainly
around
like
there
was
what
other
tests
I
assume
it's
always
running
in
a
browser,
but
the
the
merge
makes
it
run
in
a
worker.
So
therefore,
the
visibility
change
isn't
defined
when
it's
compiling
in
a
worker,
but
that
has
to
be
fixed,
so
that's
probably
contributing
back
the
worker
tests
into
JS
so
that
people
don't
go.
Why
are
you
just
changing
this?
Just
keep
it
the
same.
B
I'll
start
contributing
back
gonna
work
since,
like
the
it's,
not
an
exact
Fork.
E
Oh
yeah
you'll
just
have
to
go,
create
the
pr
in
the
other
ones,
so
once
you're
well,
once
we
do
the
the
mucking
around
in
sandbox,
we
say
yes,
this
looks
good.
It
is
then
a
case
of
going
and
duplicating
it.
No
in
theory,
you
could
because
they
are
shared
history,
you
could
try
and
merge
them
back
with.
E
C
E
Sandbox
has
Auto
linting
in
so
if
it
can
Auto
fix
it
it
just
does
it
doesn't
force
you
to
have
to
fix
everything,
there's
similar
linking
rules,
I
think
I
grabbed
it
rigid
from
the
Js,
but
there
is
likely
to
be
some
slight
differences
and
likewise
with
the
like.
Currently,
there
is
no
checks
on.
Have
you
named
the
pr
instance?
You
know
that
your
titles
correctly,
like
you,
don't
have
to
say,
chalk
or
Etc.
E
Like
as
long
as
we
keep
running
the
merging
fairly
consistently,
it
should
almost
be
the
same
code,
maybe
just
bring
up
Beyond
Compare
or
just
do
a
straight
file
copy,
and
it
should
work.
Okay,
yeah
the
Imports
will
be
different
because,
as
part
of
the
merging
it
does
rename
all
the
packages
to
have
sandbox
Dash
in
them.
E
So
it's
that
will
be
a
change
so,
instead
of
importing
from
open,
Telemetry
API
it
imports
from
open,
Telemetry,
sandbox,
API
and
stuff,
like
that,
the
package
jsons
are
changed
to
be
fixed
versions
rather
than
ranged
versions,
so
that
Rush
can
just
do
Sim
linking
with
everything
local,
so
you're,
always
playing
with
everything
with
all
the
current
version
in
the
sandbox
is
what
you're
compiling
with
not
all
the
versions.
So
all
that
additional
text
will
then
start
to
be
done
by
the
main
replays.
C
E
I
just
dropped
something
in
the
agenda
now.
This
also
came
up
in
the
spec
meeting
yesterday.
It's
part
of
the
HTTP
semantic
conventions
working
group,
which
Riley
is,
is
a
a
constant
member.
E
In
now,
I
haven't
had
a
chance
like
three
times
a
week
and
I'm
like
clashes
with
a
bunch
of
other
meetings,
but
trust
raise
this
they're
looking
to
align
with
elastic
common
schema
and
as
such,
specifically
this
one
there's
several
others,
but
this
one
here
is
they're
getting
rid
of
HTTP
user
agent
and
just
calling
it
user
agent
based
on
the
ECS
browser
user
agent,
was
called
out
as
part
of
this.
E
So
if
this
goes
through,
there
will
be
removing
browser.
User
agent
can
lead
to
making
conventions.
D
So
I
I
looked
at
it
briefly.
It's
interesting,
you
know
I
think
they
have
basically
reversed
the
ordering
of
the
items
in
the
in
the
attribute
name.
So,
instead
of
saying
http.url,
they
say
they're
now
saying
url.htv.
E
D
And
so
I
think
both
URL
and
user
agent.
You
know
we
use
in
the
browser
instrumentation.
So,
let's
you
know
talk
about
them
briefly:
URL
I
think
might
be
okay.
You
know
changing
HTTP
dot
URL
to
url.http,
but
user
agent.
If
I
look
at
the
like,
in
our
case,
we
are
sending
the.
Let
me
bring
up
our
you
know:
conventions.
E
Yeah,
so
so
really
could
the
change
here
instead
of
being
HTTP
user
agent,
it
would
be
user
agent
underscore
or
Dot
and
original
something
rather
from
memory,
yeah,
yeah,
yeah,
all
the
other
stuff
we
have
for
Brands
and
everything
that
would
still
be
the
same.
This
is
specifically
just
the
user
agent
and
so
on.
Yeah,
okay,
except
I,
guess
I
do
do
call
that
lucky
device
fields,
the
name
I
think
OS
was
down
the
bottom
as
well.
E
Yeah
trust
really
didn't
go
into
the
depth
of
it.
He
he
did
mention
that
it
wouldn't
be
everything,
so
it's
probably
just
going
to
be
useragent.original
if
this
gets
gets
through,
which
does
make
it
bigger
which
I
don't
like,
but
it
does
make
it
then
common.
So
instead
of
having
HTTP
user
agent
and
a
browser
user
agent,
which
are
effectively
the
same
thing
on
both
on
different
sides
of
the
coin,
it
just
simplifies
us
to
make
dimensions
a
little
bit
but
complicated
than
others.
E
E
De
facto
standards,
so
the
you
know
these
two
there's
three
of
them.
This
was
one
of
them.
If
you
go
back
to
the
notes
or
the
spec
meeting
this
week,
you
know
you'll
see
all
three
listed
okay,
but
yeah
children
was
against
because
of
the
de
facto
standards
the
state,
because
they
haven't
they've,
been
experimental
for
so
long.
E
Josh
I
can't,
remember,
I,
call
him
Josh
and
I
call
Jim
MCD.
So
sorry,
so
Joshua
was
for
it
and
I.
Think
Riley
was
one
of
the
ones
who
originally
proposed
aligning
with
the
ECS,
so
he
was
for
it
in
the
spec
meeting
as
well.
So.
E
So
anyway,
there's
there's
going
to
be
some
discussion.
The
HQ,
like
I,
think
the
HTTP
cementing
convention
working
group
second
meeting
right
now
as
well.
So
it's
like
they
have
lots
of
meetings
across
exactly
in
my
team
sort
of
I
can
talk
directly
to
him.
A
Do
we,
where
did
browser.useragent
even
come
from
I
mean
I
under
I,
can
appreciate
wanting
to
maybe
flatten
the
namespace
a
little
bit,
but
I
mean
the
user
agent
originates
with
HTTP,
mostly,
doesn't
it
so
like?
Where
did
browser
user
agent
come
from
like
the
the
name
browser.user
agent?
C
B
Yeah,
so
I
created
the
pr
the
change,
so
we,
the
browser
namespace,
has
a
bunch
of
different
attributes
on
it
and
most
of
them
come
from
a
browser
API
that
can
give
like
the
latest
the
new
browser
API.
That
can
give
you
the
name
and
version
separately,
but
there
are
older
browsers
that
don't
have
that
API.
So
we
added
that
user
agent
as
a
fallback
for
like
Legacy
browsers.
C
D
B
E
A
C
E
And
technically
it
is
possible
to
have
a
I
believe
a
span
processor
that
you
know
if
we
change
the
instrumentations
to
be
the
new
names,
we
could
also
create
a
Spam
process,
a
processor
for
those
systems
that
don't
like
the
new
names
that
can
effectively
double
pump
them.
So
we
can
pick
it
up
and
drop
it
into
HTTP
user
agent
or
whatever.
E
The
only
one
that
might
not
work
is
for
one
of
the
PR's
ECS
breaks
apart,
the
client,
the
crew
string
and
the
path
I
think
it
was
as
opposed
to
Target
or
vice
versa.
I,
don't
remember
which
around
it
was
now.
E
B
B
So
I'm
just
going
to
say
one
more
thing
for
next
week,
it
sounds
like
tigrant
can
join
next
week
on
Wednesday,
so
I
think
I
think
on
the
agenda
for
next
week
is
going
to
be
maybe
looking
for
a
different
time
for
the
Wednesday
meeting,
because
so
that
the
girl
has
a
conflict.
B
So
if
you
can
think
about
that
like
what
would
work
for
you
and
we
can
bring
up
but
I
think
that
the
things
that
we
were
blocked
on
from
a
TC
perspective,
I
think
is
the
the
nested
attributes,
the
session
ID
or
the
resource.
The
ephemeral
resources
and.
B
And
semantic
automations
for
the
events
right
like
we
still
have
Santosh.
You
have
that
PR
in
your
fork,
but
we
haven't
maybe.
E
I
guess
the
event
data
I
haven't
looked
at
it
this
week.
Did
you
get
a
chance
to
fix
the
typo
that
was
in
there
at
all,
so
Tosh.
D
No,
no
I
haven't
looked
at
it.
Okay,.
E
Sorry
so,
first
day,
most
likely
Friday
I
should
be
getting
some
time
back.
So
I
hope
that
get
back
to
this
stuff
then.
D
Yeah
yeah
I,
I
I
was
hoping
I
think
somebody
would
add
their
comments
or
you
know,
give
some
approve
I.
Think
in
the
last
meeting.
D
I
think
he
kind
of
gave
his
verbal
approval
right,
but
I
haven't
seen
him
putting
his
approval
on
the
pr
yeah.
E
B
Okay,
anything
else
yeah
if.