►
From YouTube: 2022-10-12 meeting
Description
cncf-opentelemetry@cncf.io's Personal Meeting Room
A
A
A
B
B
All
right,
I
guess
we'll
hop
right
into
it.
So
first
topic
is
the
metrics
ga
no
real
update
from
last
week.
The
TC
review
on
this
is
started,
it's
being
done
by
Riley
Yang
I
sent
him
all
the
various
packages
and
documentation.
It's
probably
a
lot
to
go
through
so
I.
Imagine
it'll
take
him
some
time,
but
yeah
that
started.
A
B
The
after
GA
section
here
is
largely
the
same
as
it
was
last
week,
except
that
two
of
the
items
have
pull
requests
now,
courtesy
of
Mark
for
both
of
them
I
believe
the
Prometheus
exporter.
Resource
support
was
open,
I
think
already
opened
last
week.
Actually,
oh,
no,
it's
not
okay
five
days
ago
and
is
just
in
need
of
reviews
should
be
relatively
straightforward
and
the
temporality
configuration
environment
variable
for
the
various
exporters,
I
believe
and
Mark.
Correct
me.
B
So
again,
please
review
that,
and
those
are
the
only
two
PR's
we
had
in
terms
of
of
metrics
topics
for
for
the
important
metrics
topics,
anyone
have
questions
or
anything
that
should
be
added
to
this
list.
Before
we
move
on.
A
B
Quick
update
on
merging
the
API
into
the
main
repo
I
migrated
all
of
the
existing
issues
from
the
API
repo
to
the
main
repo,
some
of
them
I
just
migrated
using
the
GitHub
migrate
button.
Some
of
them
I
closed.
Not
very
many
though,
and
some
I
actually
opened
new
issues
in
the
main
repo,
with
slightly
different
titles
or
Targets,
on
some
things
that
had
been
decided
or
where,
where
the
discussion
on
the
issue
had
diverged
from
what
the
original
title
was.
B
B
I
commented
on
all
of
them
to
let
people
know
what
was
happening.
I
know
it's
kind
of
a
a
pain,
not
an
ideal
situation,
but
it's
something
we
should
only
have
to
do
once.
B
I've
already
seen
a
couple
of
people
have
reopened
their
PRS
on
the
main
repo
and
we
haven't
had
anybody
complain
too
loudly
about
it.
So
that's
moving
along
the
only
thing.
That's
really
left
to
do
in
the
old
API
repo
is
to
Archive.
It
I
was
going
to
do
that
earlier.
This
week,
but
then
I
figured
I
would
wait
for
this
meeting
to
hear
anybody
has
a
reason
that
we
should
not
archive
it
or
anything
that
they
think
I've
forgotten
to
do
before
we
archive
it.
B
I
will
probably
create
an
issue
in
that
repo
for
the
archival
I
to
get
opinions
of
people
that
aren't
in
this
meeting.
I'll
leave
it
open
for
about
a
week
and
then
archive
it
before
the
meeting
next
week.
If
everybody
here
is
okay
with
that.
A
B
D
B
Okay,
just
out
of
curiosity,
how
much
actual
Behavior
change
was
required.
D
Yeah
I
can
just
quickly
summarize
it
if
that's
helpful.
The
main
Behavior
changes
that
were
required
were
to
the
Proto
base
exporter
and
the
proto-trace
exporter.
We
just
had
to
add
like
a
browser
platform
and
like
kind
of
rejig,
the
platform
stuff
and
then
the
main
like
functionality
change
was
that
the
xhr,
like
sendex
HR
function
also
had
to
support
sending
buffer,
arrays
or
blobs,
because
that's
how
we're
actually
sending
the
the
Proto
data
over
the
wire.
So
basically
it's
a
lot
of
the
same
stuff.
A
B
Just
out
of
curiosity,
the
browser
platform
for
the
base,
Proto
I,
was
a
little
bit
surprised
that
we
needed
a
browser
specific
version
of
this.
What
how
is
this
different
from
the
node
version.
B
D
D
Yeah,
it's
the
exact
same
as
the
HTTP
one.
We
do
this
there
as
well.
B
D
Yeah,
just
really
looking
at
the
send
mechanism
and
making
sure
that
that's
that's
what
how
we
expect
to
do
it.
The
one
thing
I
didn't
Implement
yet
and
I
like
I'm,
assuming
I'll,
get
comments
on
is
like
I
think
when
we
do
this
and
know
there's
different
compression
options
so
I
haven't,
you
can
see.
It's
commented
out
there
on
line
47
for
the
compression
algorithm,
so
I'm
wondering
if
that's
a
requirement
or
if
that's
something
that
we
can
push
in
later,
but
other
than
that
yeah.
A
Okay,
I
assume
the
yeah,
so
you
the.
B
B
We
try
to
keep
them
the
same
signature
if
possible.
Oh.
D
B
D
D
B
B
B
D
B
Yep,
this
is
something
that's
been
requested
a
few
times
and
I
guess
just
nobody
has
really
gotten
along
to
doing
it.
Obviously
there
was
a
PR
open,
but
the
CLA
never
got
signed,
which
is
obviously
a
blocker.
B
B
Okay,
I
wanted
to
bring
up
the
the
timing
issue
with
the
anchored
clock
one
more
time
last
week
and
I
believe
the
week
before
we
talked
about
this
and
moving
to
using
date
instead
of
using
the
performance
timer
for
the
last
couple
of
weeks,
I
did
link
the
the
discussion
issue
there
and
I
haven't
seen
any
objections,
so
I
just
wanted
people
to
know
that
I
did
start
working
on
this.
B
It's
a
little
bit
of
a
pain
to
do
it
in
a
backwards
compatible
way,
because
the
core
package
was
using
performance,
timer
and
everything
is
using
cord
for
for
timings,
so
keeping
core
backwards
compatible
with
what
it
already
was,
while
changing
the
underlying
stuff
to
use
date
is
taking
a
little
bit
more
time
than
I
expected
it
to,
but
look
for
that
PR,
probably
sometime
this
week.
B
Anyone
have
questions
on
that
before
I.
Move
on
I
can
link
the
discussion
with
you
in
case.
Anybody
wants
to
look
through
that.
A
B
I,
just
added
this
as
a
reminder
to
myself
is
Martin.
It
looks
like
Martin,
you
are
on
the
call.
Yes.
A
B
Last
week
I
asked
the
group
if
there
was
anyone
working
on
the
logs
and
events,
SDK
I
know
the
API
has
merged
and
Nev
mentioned
that
he
thought
you
might
be
working
on
something
or.
A
B
A
So
I
did
I
did
start
working
on
it,
but
I
was
I
was
out
now
for
a
couple
weeks.
So
I
put
it
on
this
on
side,
but
I'm
I'm,
getting
back
into
it
and
I
think
also
it's
something
she's
on
the
call
I
think
he
said
that
somebody
from
it
from
his
side
could
help
me
out.
B
Okay,
yeah,
mostly
I,
just
wanted
to
to
make
sure
that
this
was
that
there
was
ongoing
work
here.
I
didn't
know
if
it
was
up
for
grads
or
not
yeah.
You
know
if
they're
I,
don't
think,
there's
an
issue
opening
for
the
SDK
work
on
the
I.
A
B
A
B
I'm
not
sure
who
added
this
item.
That's.
A
Me
that's
the
too
much
requests
which
have
mag
welded
from
the
API
repository
as
you've
mentioned
before
flux
into
your
comments
already
I
wanted
to
get
some
eyes
on
this.
As
the
report
requests
on
the
API
has
been
long
open
for
a
while
now,
so,
if
you
move
forward
with
this
or
I
should
drop
it
basically,
okay,.
B
Yeah,
so
the
the
sugar
Tracer
I
know
you
opened
a
while
ago.
The
the
force,
flush
and
shutdown
I.
Think
generally
you
know
is,
is
something
that
we
probably
want
to
do
looks
like
all.
The
tests
are
failing
right
now,
I
think
that
this
one's
fairly
uncontroversial,
we
obviously
have
to
to
look
at
the
implementation
itself,
but
I.
B
Don't
think
that
there's
anybody,
that's
really
against
this,
so
I,
don't
I,
don't
expect
that
to
be
too
controversial,
the
sugared
Tracer
one
is
a
little
different,
not
because
I
think
it's
controversial
or
that
any
of
the
ideas
are
bad
in
here
or
anything
like
that,
but
maybe,
instead
of
a
separate
component,
some
of
these
methods
can
just
be
added
to
the
Tracer
that
we
already
have
also
I.
There's
some
things
like
I
think
with
span.
B
Maybe
is
a
method
name,
that's
already
being
used
somewhere,
we're
definitely
using
with
in
context
but
I.
Think
we're
using
with
span
on
the
Tracer
already
I
can't
remember,
but
I
want
to
make
sure
we
don't
have
like
naming
collisions.
A
There
has
been
some
different
discussions
around
this.
This
was
basically
the
consensus
for
for
my
first
draft.
Basically,
but
that's
the
reason
I
want
to
get
some
eyes
on
this
okay,
so
we
can
finalize
the
first
draft
and
can
move
on
to
because
I
really
like
to
see
this
intercourse
in
the
Upstream.
B
Adding
some
some
convenience
methods,
I'm
definitely
I,
definitely
think
could
be
a
big
help,
especially
for
newer
users.
A
A
B
A
Show
a
couple
feedback
so
that
sort
of
doesn't
hang
there
for
a
year
or
so.
Okay,.
B
Okay,
contrib
PR
is
looking
for
reviews.
I
assume
that
this
was
either
roundo
or
Amir
looks
like
Amir
is
not
here,
so
it
must
have
been
now.
A
C
A
B
Is
this
one
only
the
two
packages
we
were
talking
about:
yeah
data,
loader
and
mongoose.
A
Okay,
yeah
and
pick
them
up.
B
A
Yeah,
especially
the
one
with
the
type
definitions
being
removed,
the
one
the
last
one
you
had
open
I
think
that
that's
the
you
know
the
most
critical
and
the
others
they're
actually
like
fixing
some
bugs,
and
one
of
them
is
just
the
CI
fix
or
step
stability
fix.
So
so
nothing
particular
it's
they're
all
kind
of
the
same
area
priority.
C
I
can
just
give
a
quick
update
on
the
sandbox
I.
The
last
Auto
merge
that
I
did
from
concrete
broke.
My
script,
I
haven't
had
a
chance
to
get
back
to
it.
So
at
the
moment
the
the
auto
merge
doesn't
work
because
there
was
a
a
rename
from
a
package
to
a
file,
so
I
I
need
to
get
back
to
it.
It's
going
to
be
probably
next
week
or
the
week
after
before
that'll
be
picked
up
again.
C
I'm
just
trying
to
completely
wrap
up
my
internal
stuff,
so
I
can
get
on
to
more
full-time,
open
Telemetry
than
what
I
have
been.
C
So,
for
some
reason,
if
you
go
look
at
the
sandbox
what's
checked
in
is
there's
like
an
extra
sub
packages.
Folder
I'm
not
quite
sure
how
that
happened.
So
when
it
tries
to
merge
the
changes
in
again,
it's
trying
to
merge
a
folder
on
top
of
a
file
name.
I
have
a
look
at
the
detail.
I
just
I've
tried
to
run
the
script
twice
and
saw
that
and
went
okay
I
need
this
carve
out
some
time
to
fix
it.
B
B
A
B
Matt
exponential
histogram
progress.
Do
you
have
any
particular
question
here
or
are
you
just
wondering
looking
for
a
general
update,
yeah,
just
looking
for.
A
A
general
update,
I
know
you
mentioned
that
you're
gonna
work
on
a
prototype
and
yeah,
so.
B
It's
based
on
actually
implementations
by
lightstep
I,
believe
Josh
McDonald
wrote
them
particularly
the
go
implementation,
but
just
wrapping
my
head
around
the
the
the
underlying
maths
behind
it
has
been
sort
of
the
the
blocker
at
the
moment,
just
because
I
haven't
had
a
ton
of
time
to
to
really
go
through
it
and
understand.
B
What's
going
on
and
I,
don't
really
just
want
to
copy
it
line
for
line
without
understanding
what's
going
on
in
there,
so
yeah
I
haven't
really
worked
on
it
in
the
last
three
weeks
or
so
because
it
hasn't
really
been
a
high
priority.
Is
it
something
that
that
you
were
looking
to
to
pick
up
work
on
if
or
are
you
just
looking
for
an
update.
A
Mainly
looking
for
an
update
but
yeah
I
can
check
to
see
like
how
much
lights
up
cares
about
this
and
trying
to
figure
out
a
how
to
help
or
B
how
to
like.
Maybe
collaborate
with
Josh
to
figure
out
the
fancy
math
that
has
to
go
on
behind
this
or
or
something
to
kind
of
chip
in
one
way
or
another.
A
B
Okay,
I
think
you're
gonna
feel
well,
it's
it's
hard
to
tell
from
the
outside
whether
lightstep
really
cares
about
it
or
whether
Josh
really
cares
about
it,
but
certainly
at
least
one
of
the
two
of
those
things
is
true:
I
guess.
If
this
is
a
high
priority
for
for
you
or
for
light
stuff,
then
any
sort
of
contribution
would
be
welcome.
B
Otherwise,
I
do
plan
to
get
to
this
in
the
somewhat
near
future.
I'm
just
focused
on
on
other
things,
particularly
related
to
the
the
metric
GA
and,
and
things
like
that.
Right
now,.
A
Cool
yeah
sounds
good,
I.
Think
I.
Think
I
should
do
a
little
homework
on
on
the
implementations
and
read
through
a
few.
So
so
I
better
understand,
you
know,
what's
what's
behind
it,
and
at
least
that
will
help
help
out
with
reviews,
if
nothing
else,
but
if
I'm
feeling
confident
and
you
haven't
gotten
around
to
working
on
it.
Maybe
I
can
start
to
pick
up
some
of
that,
but
CBD.
A
B
B
On
bug,
triage
single
Express
at
printing,
odd
warning
open
three
days
ago,
it
looks
like
he
opened
this
without
the
triage
label.
Landing
instrumentation,
you
know,
register
a
module
of
the
instrument
of
it's
the
base
instrumentation
class
for
Warren.
Should
we
remove
the
warning.
B
I
think
I'm
gonna
put
this
as
G4,
because
it's
not
really
causing
problems.
It's
really
just
the
warning
announce.
A
B
Dmv
detector,
this
already
has
a
PR
yeah,
so
I
will
assign
Anthony
here,
remove
the
triage
label
and
add
we
have
to
view
I
guess,
because
this
is
not
causing
crashes,
but
it
is
causing
bad
instrumentation
for
those
that
aren't
aware
of
this
issue,
the
with
the
hotel
resource
attributes,
environment
variable
and
if
your
value
contained
a
space,
it
wasn't
being
parsed
properly.
B
So
he
opened
this
issue
a
week
ago
and
already
opened
a
PR
for
this.
So
right
now
we're
just
waiting
on
the
CLA
to
be
signed
for
that
PR.
B
Ad
response
body
of
I
think
I
know
the
pr
author
here
sometimes
joins
this
meeting,
but
it
looks
like
she's
not
here.
I,
don't
think
this
is
a
bug.
A
B
B
That's
it.
Anyone
have
anything
before
we
call
it
a
day.