►
From YouTube: 2023-01-25 meeting
Description
cncf-opentelemetry@cncf.io's Personal Meeting Room
B
B
A
I
think
we
can
probably
get
started
for
those
that
don't
already
know
or
don't
remember.
I
was
gone
last
week,
so
I'm
still
catching
up
a
little
bit
feel
free
to
add
anything
to
the
agenda
that
you
feel
like
I
missed
here,
because
it's
entirely
possible
that
I
did
miss
something.
A
A
There
was
a
fix
for
that
that
I,
already
merged
and
I'm
going
to
try
to
release
a
fix
for
that
this
week,
I
was
trying
to
create
the
the
release
PR
before
this
meeting
started,
but
just
didn't
quite
have
time.
So
look
for
that
this
afternoon,
it'll,
probably
just
be
a
minor
version
update
since
there
have
been
no
features
added
so
should
be
relatively
easy.
A
Another
issue
that
came
up
while
I
was
gone
is
that
the
the
clock
drift
fix
for
metrics
caused
a
flaky
test.
This
is
just
because
we're
using
date,
dot
now
and
the
tests
were
mocking
the
performance
timer.
Instead
of
mocking
the
date
timer.
A
The
tests
still
passed
most
of
the
time,
but
there
are
flaky,
so
I
will
create
an
issue
for
that
and
I'll
probably
just
mark
it
as
up
for
grabs
right
now,
since
I
don't
have
a
ton
of
time
to
work
on
it
today,
but
it
should
be
a
relatively
easy
fix
and
if
nobody
picks
it
up,
then
I'll
probably
get
to
it
tomorrow.
A
First,
real
item
is
the
exponential
histogram
thing
for
the
part.
One
PR
myself
and
Mark
have
already
reviewed
this.
It's
been
approved
by
the
two
of
us
for
a
while
and
open
for
a
while
before
that,
and
I
mostly
wanted
to
ask
Matt's
opinion
here.
Do
you
think
that
this
is
ready
to
merge,
or
would
you
prefer
to
wait
for
more
reviews?
C
Yeah
I
I
think
it's
ready
to
merge.
It
part
two
Builds
on
that
a
lot
so
I
will
have
to
rebase,
because
there
were
some
minor
changes,
but
if,
if
anybody
tags
along
in
part
two
and
sees
something
wrong
with
part,
one
feel
free
to
point
it
out
at
that
point
in
time.
So,
like
I
feel
like
this
is
by
no
means
last
call
and
part
two
is
where
you'll
see
the
stuff
working
together
a
little
bit
more.
C
So
if,
if
folks,
who
already
reviewed
it
sees,
you
know
something
in
part,
two
that
could
have
been
caught
in
part,
one
I'm
not
gonna,
be
upset.
If
you
point
it
out
later
on
so
I
feel,
like
everything
is
fair
game
up
until
part
three
is
merged,
and
even
after
that
yeah
just
come.
Hunt
me
down
for
for
any
improvements
or
issues.
A
Okay-
and
nothing
in
here
is,
is
public
or
usable
anyway,
even
when
it
gets
merged
until
part
at
least
two.
So
probably
the
risk
is
fairly
minimal
for
part.
One.
C
Yeah
I
think
all
these
things
end
up
being
just
internal
functions
anyways,
even
in
the
end
from
part,
one.
A
Okay,
okay,
so
I'll
probably
merge
that
today
then,
and
we
can
move.
A
C
Oh
I
was
just
bringing
back
up
I
think
we
looked
at
this
a
couple
weeks
ago.
Occasionally
I
get
asked
about
it,
because
people
would
really
like
this
environment
variable
that
are
kind
of
using
Lambda.
So
it's
just
going
to
to
mention
this
is
still
open.
A
Okay
looks
like
this
is
a
PRI
open,
so
I
will
make
sure
to
look
into
there's
some
review
comments
here
from
a
while
ago,
so
I'll
take
a
peek
at
them
and
try
to
bring
this
up
to
date
so
that
it
can
be
reviewed
and
merged.
It
looks
like
a
handful
of
people
who
are
already
happy
with
it
too.
So
as
long
as
nothing
too
big
changes,
I'll
probably
just
merge
it
so
that
we
can
include
it
in
the
release.
D
Yes,
it
is
correct,
so
thanks
Mark
for
the
review
and
I
was
just
wondering
if
it
could
be
merged
soon.
That
would
unblock
us
from
the
Oracle
unlock
SDK.
There's
another
PR,
that's
opened
by
someone
else,
but
that's
waiting
for
these
changes
to
be
to
be
done.
A
Okay,
I
haven't
really
had
time
to
look
into
this
PR.
It
was
open
while
I
was
gone,
but
I
will
take
a
quick
peek
at
it
and
and
merge
it
today.
If
I
can
it
does
look
like
there's
some
review
comments.
A
A
All
right,
well
I'll,
take
a
peek
at
it.
After
the
after
the
meeting,
then
I
assume,
there's,
there's
probably
not
much
functionally
different.
It's
just
splitting
it
into
two
separate
apis
right.
D
Yes,
it's
there
is
one
minor
change
with
the
domain.
The
domain
moved
from
to
the
provider
from
the
from
the
emitter
itself,
but
that's
minor,
in
my
opinion,.
A
A
E
E
But
it's
a
lot
lower
level
than
any
of
the
things
like
open
Telemetry,
so
like
doesn't
do
any
correlation
between
different
spans.
It's
just
like
a
a
single
span,
and
then
you
can
correlate
things
externally.
E
It
just
produces
like
pretty
basic
data,
but
the
ideas
like
nud
cars,
integrating
this
in
a
bunch
of
places
and
then
like
the
ecosystem,
can
integrate
this
as
well
and
rather
than
something
like.
Oh,
it's,
open
Telemetry,
which
has
like
opinions
of
like
this
represents
a
span
object.
This
is
just
a
whole
bunch
of
individual
events
that
you
can
use
to
express
things.
However,
you
want
yeah
yeah.
The
the
intent
is
like
eventually
for
things
like
open
Telemetry
to
build
their
instrumentation.
On
top
of
this.
A
Okay,
in
that
sound
promising
from
your
perspective,
what
are
you
asking
from
us
here
just
to
to
give
feedback
on
the
API
design
and
whether
it
would
fit
our
needs
or.
E
A
E
A
Okay,
thanks
for
bringing
into
our
attention.
F
Yep
so
I
have
an
initial
PR
out,
which
is
the
first
item
there
to
effectively
create
the
emerging
from
Stadium
to
Main
Branch,
the
pr
that's
there
are
only
merges
the
API
at
this
point
in
config.ts,
there's
a
table
of
items,
I
I've
left
some
comments
in
there
already
and
I
gave
a
talk
yesterday
in
the
Ramsey.
F
So
if
you
go
back
and
look
at
the
YouTube
recorded
video
you're
assuming
waffle
on
for
about
an
hour
or
so
in
terms
of
what
that
script
generates,
I've
got
an
example
of
the
generated
PR,
where
I
actually
enabled
more
than
just
the
API
to
go
through.
It's
like
this
one's,
the
initial
one,
so
it's
really
large,
which
is
why
the
one
that
I've
got
checked
in
the
pr
is
only
bringing
in
the
API
to
try
and
reduce
the
the
size
of
the
pr.
It's
all
pretty
much
automated.
F
What
the
script
does
do
is
effectively.
It
throws
out
learner,
replaces
it
with
Rush
ads
roll
up,
so
it
creates
bundles
for
absolutely
everything
that
that
is
bringing
over
the
entire
JS
and
concrete
repos
end
up
being
in
the
auto
merge
folder,
because
that's
what
happens
today
from
the
in
the
auto
merge
branch
and
then
the
new
structure
ends
up
being
in
the
packages
so
pkgs's.
F
If
you
go
down
to
config
TS,
there's
a
table
yeah,
so
this
table
here
the
whole
bunch
of
comment
of
that
stuff.
Just
we're
not
going
to
scream
at
all
yeah.
So
this
yeah
from
like
111
through
124.
This
is
where
it
brings
in
all
the
individual
packages.
Let's
say
it's
commented
out
scroll
up
a
little
bit
more.
F
You
saw
the
API
one,
so
it's
probably
a
bit
easier
to
read
uncommented,
so
you
can
see
on
110,
it
says:
grab
the
API
package
from
Auto
merge,
GS
API,
which
is
where
the
automotive
script
drops
it
into
put
it
into
the
pkgs
API
folder
create
a
bundle
called
Hotel,
sandbox
api.js,
and
when
it's
creating
a
UMD
format,
it
uses
a
namespace
of
open,
telemetry.t,
sandbox.web
API.
F
Let's
say
this:
does
all
the
magic
in
terms
of
doing
it,
you,
if
you
want
to
validate
it,
because
it's
like
a
thousand
line
with
a
typescript
could
be
the
main
one.
It's
easiest
to
probably
grab
this
and
run
it
yourself.
While
it
uses
rush,
I've
changed
all
the
npm,
your
other
scripts,
so
npm
run,
compile
converts
into
Rush,
rebuild
npm,
run
test
converts
into
Rush
test.
So
all
that
stuff
still
works.
F
I
was
hoping
to
have
this
out
last
week
and
checked
in
some
stuff
to
come
up.
So
it'll
probably
be
later
this
week
before
I
actually
get
this
in
one
of
the
reasons
for
only
dragging
in
the
API
first
is.
This
is
based
on.
What's
already
in
the
sandbox,
which
was
from
November
last
year,
I've
got
a
a
PR
which
I
pushed
out
yesterday
to
merge
the
current
country
in
Js
into
it,
which
I
want
to
run
after
I.
F
Do
this
so
that
I
can
then
check
to
make
sure
that
the
automatic
resolving
of
complex
works
properly
but
yeah
this
effective,
creates
a
structure,
so
you
can
see
like
a
111,
it's
grabbing
semantic
conventions
and
dropping
into
pkgs
semantic
conventions.
All
the
way
down.
I
think
I've
got
the
instrumentation
and
the
document
load
currently
defined,
where
it
Greg
Greg's
document
load
from
the
country
folder
and
puts
it
into
a
location.
F
I'm
not
quite
happy
with
the
instrumentation
structure.
I've
got
in
terms
of
pkgs
instrumentations
instrumentation
is
the
main
one
and
then
pkgs
instrumentations
web
document
load
for
I'm
thinking.
I
might
split
that
out,
because
the
contrabuto
calls
them
plugins,
so
I'm
open
to
suggestions
on
that.
A
F
F
Okay,
so
there's
a
lot
there.
If
you
have
questions
instead
of
the
pr
or
pygmy
either
one
works.
A
Yeah
makes
sense,
I
want
to
look
at
it
a
little
bit
more
before
I
come
after
all,
but
I'll
probably
bad
comments
on
the
pr.
If
I,
if
I
have
any
yep.
F
Yeah,
so
the
basic
structure
of
dropping
everything
into
pkgs
was
from
my
very
first
PR,
where
I
did
it
all
manually,
so
I,
don't
think
we
actually
got
to
the
conversation
of.
Is
that
the
structure
we
wanted,
but
it's
all
config
driven
now,
so
we're
going
to
change
the
location.
We
just
change
it
in
that
table
and
when
it
goes
through
and
rewrites
and
updates
all
the
configs.
It
changes
all
the
relative
paths.
So
that's
what
the
script.
B
A
All
right:
well,
let's
get
on
to
a
little
bit
of
bug.
Triage
then
I
did
see.
There
were
a
few
that
were
opened
while
I
was
gone,
I
guess
we'll
see,
I,
don't
think
there
were
yeah
still
information
requested.
There
can't
update.
A
Oh
I
remember
this:
one
I
had
to
sign
this
one
to
myself.
I
have
not
had
time
to
look
into
it
yet
so
we're
looking
at
it
today,
HTTP
instrumentation
modifies
header
casing
of
outgoing
requests.
This
already
has
an
open
PR,
which
is
already
linked.
A
So
this
definitely
is
a
bug
priority.
Two.
A
A
A
A
Get
node
Auto
instrumentations
mapping
to
my
exposed
endpoints
how
to
be
that
this
also
doesn't
sound
like
a
bug.
Okay,
you
know
it
is
from
the
code
in
order
to
automatically
fetch
the
counters
mapping
into
the
endpoints
application.
A
A
A
It
before
we
close
the
calls
does
anyone
have
anything
that
they
want
to
bring
up.
A
All
right:
well
then,
everybody
enjoy
the
rest
of
your
day.
Look
for
the
release,
PR
that
I
open
this
afternoon
and
I'll
talk
to
you
next
week.
Thank.