►
From YouTube: 2023-02-27 meeting
Description
cncf-opentelemetry@cncf.io's Personal Meeting Room
A
A
B
C
D
C
It's
just
the
end
of
summer
right
now.
It's
it's
still
pretty
hot.
Actually
it
was
actually
pretty
cool
summer
until
the
very
end
when
it
got
really
hot
I.
Don't
know
why
but
yeah
how
about
you,
how's
winter.
B
A
It's
very
cold
last
week
like
in
the
in
the
night
it
it
went
to
like
minus
10
Celsius.
It
is
surprisingly
low
for
a
comparing
to
previous
years.
B
C
C
Does
it
does
yeah
in
like
if
you
go
about
six
hours,
Southwest
of
Sydney
there's
like
mountains
and
it
snows
there
in
the
winter,
and
you
can
go
skiing
in
Australia,
so
yeah,
not
as
good
as
New
Zealand,
or
anything
like
that,
though,
yeah.
B
Really
soon,
two
weeks,
oh
wow,
didn't
he
oh.
C
B
Oh
cool,
we
don't
really
have
much
planned
out
for
Sydney
and
we're
gonna
go
to
Melbourne
and
definitely
like
found
some
things
to
do
around
Melbourne.
C
B
All
right,
I,
just
pinged,
Miller,
and
so
just
briefly
on
the
current
ones,
I,
don't
think
we
have
too
much
going
on
on
the
HTTP
side.
Most
of
the
stability
block
stability,
blockers,
problematic
stability,
blockers
right
now
are
General
semcon
issues,
but
there
is
this
one.
B
The
let's
see
this
is
probably
the
main
one
that
we
have
have
work
to
do
sort
of
on
our
side,
and
we
can
come
back
back
to
this.
B
B
We
want
to
drop
this
app
and
I
have
another
issue
about
that.
But
if
we're
going
to
ECS,
then
it's
different
still
so
we're
kind
of
in
holding
pattern
on
a
couple
things
just
until
we
find
out
about
ECS,
which
they
told
us
most
recently
they're
going
to
try
to.
Let
us
know
end
of
the
end
of
this
week.
If
okay,
we're
going
to
try
to
move
forward
with
that.
B
And
same
on
this
one,
this
is
pretty
straightforward.
Pr
has
reviews
already,
but
if
you
have
a
chance
to
look
at
that,
also
I
think
we'll
probably
hold
off
on
merging
a
couple
of
these
until
kind
of
all
in
a
one
batch,
probably.
B
No
worries
so
I
guess
the
things
we
could
talk
about
today.
One
would
be
if
we
want
to
talk
HTTP
the
HTTP
connection,
kind
of
the
HTTP
high
level.
Low
level
span
discussion
that
continue
that
that
we're
having
from
Friday
the
other
is
sort
of
what
we
think
from
The
General
semcon
that
either
we
can
help
with
or
that
we
should
escalate
to
the
spec
group.
B
B
D
A
B
D
B
A
So
so
it's
unclear
to
me
because
nobody
ever
replied
in
that
issue,
saying
they
disagree
with
the
conclusion
from
the
spec
meeting
right
I
concluded
that
folks
agree
that
we're
going
to
change
to
seconds
there
are
there.
There
are
some
complaints,
but
nobody
is
saying
they
disagree
with
the
final
decision.
They
were
saying
the
detail
is
going
to
hurt
them.
B
A
B
A
B
Okay,
cool
and
then
I
will
just
highlight
it
for
folks.
Just
so
that,
because
some
people
come
to
the
spec
meeting
in
lieu
of
following
the
all
the
issues,
foreign.
B
A
B
B
And
I
will
add,
I
will
add
our
oh
good.
Yes,
I
have
that.
B
B
I
think
we
have
pretty
much
resolved
this.
This
I
think
we
at
least
have
a
list
now
of
everything.
B
Yeah,
so
more
I
need
more
eyes
on
more
reviews
of
this.
B
So
if
you
happen
to
you
want
to
review
this,
that
would
be
great.
D
B
B
E
F
Oh
perfect,
yeah,
I'm,
sorry
yeah,
so
in
this
issue
the
septicon
HTTP
SIM
card
depends
yeah.
There
is
a
list
of
net
attributes,
I,
wonder
wonder
what
should
we
do
with
them?
Should
we
put
them
in
a
separate
yaml?
We
should
we
leave
them,
as
is
I,
think
we
should.
Since
we
have
time
it
would
be
better
to
put
them
in
a
separate
Yama
foreign.
F
No,
so
we
have
to
spend
General
symmetric
collections.
Well,
that's
that's
a
good
question.
So
in
theory
spend
General
it's
just
the
list
of
attributes,
but
socket
level,
attributes
and
logical
that
pearnet
hosts
are.
F
F
And
there
will
be
net
app
protocol
name
that
protocol
version,
so
there
is
more,
it's
not
given
the
full
list,
so
I
wonder
what
we
should
do
with
those
should.
We
take
the
part
that
we
are
going
to
depend
on
put
it
in
a
separate
document.
E
So
like,
as
we
discussed
previously,
we
have
two
options
here
right,
so
we
can
just
separate
them
to
a
sec
to
another
Dock
and
also
make
it
stable,
or
we
can
kind
of
note.
Every
single
attribute
is
the
status,
so
we
can
still
keep
them
in
the
same
kind
of
document
or
maybe
it's
the
same
even
memo,
but
we
can
treat
them
on
an
attribute
level.
F
Yeah
and
if
we
do
the
second
one,
then
we
probably
should
come
up
with
the
new
tooling
continue
and
a
new
property
here.
For
that
tribute,
say
in
the
state,
the
75
and
statuses
and
it's
it
feels
like
it's
in
the
scope
of
the
experimental
or
tab
Daniel
dealer
came
up
with
and
it
will
be
more
work
and
more
tooling
stuff.
F
F
F
Yeah
so
I
can
introduce
a
new
group.
I
can
extract
the
socket
attributes
there
and
other
attributes
would
depend
upon
maybe
in
different
groups,
and
they
can
just
add
a
note
saying
that,
okay,
this,
like
later
on
in
yamu,
we
can
just
add
a
note
saying
that
this
is
a
stable
thing.
Don't
touch
it
and
later
on
use
all
the
two
links
we
will
probably
have
for
individual
attributes.
B
One
interesting
feedback
we
got
in
the
JavaScript
about
HTTP
semantic
stability
was
that
there
was
a
lot
of
confusion
about
metrics
that,
if
we
had,
if
we
marked
of
marking
some
Metric
stable
and
some
metrics
not
stable
in
the
same
document
which
I
tried
to
explain
is
the
like,
the
the
pushback
was
saying
that
that
would
make.
Then
the
HTTP
wouldn't
metrics
wouldn't
be
stable,
since
there
were
some
things
in
there
that
were
not
stable
and
I
said
well.
B
B
How
much
we
want
that
has
not
been
the
case
for
other
things:
we've
done
the
mixed
and
marked
certain
section
as
as
stable
and
not
stable,
I'm,
not
sure
I
entirely
agree
with
that
feedback,
but
I
did
want
to
pass
it
along
in
case.
Anybody
else
has
heard
similar.
B
A
B
That's
what
I
figure
it's
just
I
know.
It's
not
kind
of
the
standard
of
splitting
things
out.
That's
kind
of
standard
in
the
spec
has
been
mixed
and
separating
by
section,
but
for
semantic
conventions.
It
fits
pretty
nicely
for
just
splitting
it
out,
because
it's
not
like
there's.
You
need
context
of
reading
some
a
doc
top
to
bottom.
A
We're
not
very
consistently
we
look
at
the
login
apis,
like
folks,
are
moving
the
events
to
a
separate
document,
basically
saying
we
want
the
knowledge
API
to
be
stable
and
we
don't
want
to
mess
up
with
events.
Let's
move
that's
when
the
other
document,
and
that
was
the
one
who's
calling
what's
the
difference
between
login
and
emails.
B
F
Okay,
let's
first
I
have
a
question
for
this
audience.
Assuming
we
do
nothing,
will
it
be
good
enough
like
if
we
just
let
if
we
put
a
stable
and
HTTP
markdown
file,
is
it
good
enough?
Do
we
care
where
the
net
attributes
are
defined
at
all.
B
E
E
E
E
B
So
I
have
a
question
about
that.
So
with
the
http.
B
Spec
right,
if
we
mark
this
page
stable,
that
is
because
some
of
these
are
overridden
sort
of
definitions
for
http.
B
F
Yeah
so
like,
regardless
of
what
we
do
with
yaml
and
how
we
separate
those,
we
would
be
the
factor
doing
the
same
thing
we
are
declaring
is
stable
in
any
change
to
this
file
will
have
to
be
non-breaking
right
and
we
cannot
enforce
more
by
removing
this
attributes
in
a
different
yaml,
all
right,
so
I
kind
of
tend
to
agree
with
trust
that
if
we
do
nothing,
we
want
it
will
be
not
worse
than
anything
else.
So
it's
it's.
We
are
okay,
doing
nothing.
B
I
couldn't
raise
this
tomorrow
in
the
spec
meeting
and
basically
because
my
my
thought
is:
it's
not
a
blocker
for
HTTP
stability,
but
I'd
like
to
get
a
broader
consensus
on
that.
B
So
this
one
we
still
have
I
think
I
glossed
over
that
there's
an
update,
but
it's
really
still
a
to-do.
B
B
B
B
E
B
E
B
Latest
is
we've
been
chatting
with
Alex
and
he's
they're
going
to
get
back
to
us
at
the
end
of
this
week
with
I
hope
is
decision
close.
You
know,
I'm
I'm
a
little
concerned
about
it
dragging
on
enough
like
well.
You
know
we
still
need
to
do
more
thinking
or
research
or
discussions.
B
So
at
some
point
we'll
need
to
take
a
firmer,
a
firm
kind
of
date,
time
time,
because
we
were
our
Target
is
March
10th
or
was
our
original
Target
for
sort
of
the
final
specs
being
available
for
review,
which
I
think
we
can
make,
given
that
we
have
it
if
we
have
a
decision
at
the
end
of
this
week.
But
if
we
don't
have
a
decision
at
the
end
of
this
week,
then
I
guess
maybe
we
just
proceed
forward
with
as
if
we're
not
doing
it
and
get
those
spec.
E
B
Cool
all
right:
well,
we
are
right
on
time,
and
that
was
helpful
with
triage
and.
B
Yeah
and
then
maybe
hopefully,
if
we
don't
have
too
much
else
to
chat
about
on
Wednesday,
we
can
talk
more
about
the
HTTP
layering
low
level.
High
level
stuff
sounds.