►
From YouTube: 2022-05-19 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
A
A
Hello,
hey
how's,
it
going
good.
How
are
you
doing
pretty
good
busy.
A
B
A
Just
8
a.m
to
5
p.m,
event
called
incident.
B
A
Okay,
cool,
I
think
we
could
probably
start.
I
don't
think
anthony's
coming
just
based
off
of
kubecon
yeah
he's
still
there
yeah,
okay.
A
Cool
well
yeah.
Okay,
welcome
everyone
thanks
for
joining,
please
be
sure
to
add
your
names
to
the
attendees
list.
If
you
want
to
talk
about
something,
please
add
it
to
the
agenda
right
now.
It's
very
me
heavy,
but
we
can
keep
going
on
that.
So
one
of
the
first
things
that
was
proposed
is
damian,
I
think,
is
his
name.
We
were
just
adding
to
the
approvers
list
and
just
added
them
as
an
approver
this
week.
A
So
I'm
really
excited
about
that,
always
great
to
see
the
community
growing
and
so
with
that
they
are
located
in
the
eu
time
zone.
I
think
I
can't
remember
what
timezone
it
was.
I
don't
know
if
aaron
remembers,
the
request
was
to
change
this
meeting
time
to
be
10
a.m.
Pacific
time-
and
you
know
it's
ironic-
because
that's
of
course,
I
think
where
the
meeting
originally
started
like
three
years
ago,
so
I
I
don't
really
have
any
opposition
to
that.
There's.
Definitely
other
meetings
in
the
open
semester.
A
Calendar
then
specifically
erlay
and
it
would
overlap
with
the
communications
sig
as
well.
So
those
are
the
only
thing
downsides.
I'd
see
to
changing
it
to
them,
but
I'm
wondering
if
anybody
else
has
any
strong
opinions
of
yay
or
nay
there.
A
B
A
Yeah
yeah
it
is.
We've
also
tried
to
do
like
a
time
that
would
work
for
australia,
but
that
was
short-lived,
but
yeah.
A
A
In
for
him,
that
would
be
ironic
because
I
think
he
was
the
original
supporter
of
changing
time,
I'm
not
sure,
though,
or
if
it's
movable,
but
we
may
want
to
ping
him,
okay,
yeah
good
to
know.
I
will
keep
that
in
mind.
A
Okay,
cool
with
that,
I
think
that
I
don't
see
any
major
opposition,
so
there
may
be
a
change
in
the
meeting
time
next
week,
depending
on
anthony's
schedule.
So
just
a
heads
up
we'll
be
meeting,
probably
at
10..
If,
if
we
do
change
I'll
try
to
post
something
in
slack
as
well,
okay,
so
moving
on
talking
about
the
metrics
sdk
and
the
progress
there,
so
here's
the
milestone-
I
guess
the
relevant
number
is
27.
So
moving
forward.
A
There's
still
a
lot
more
to
add
to
this,
so
I
think
that
you
know
take
it
with
a
grain
of
salt,
but
I
do
think
it's
nice
to
see
some
sort
of
progress.
Specifically,
we
can
look
a
little
bit
at
the
project
board.
There's
a
lot
in
progress
right
now,
there's
already
a
lot
done
as
well,
and
so
I
think
that
we're
definitely
making
some
progress.
A
A
Cool,
so
with
that,
there's
five
open,
prs
related
to
the
metrics,
sdk,
reshuffle
and
redo,
and
this
is
one
of
them
there's
an
with
rear
and
a
resource
option
pr.
This
essentially
is
adding
back
to
the
meter
provider
options
to
set
the
both
resource
and
reader.
This
is
just
like
filling
it
in
so
this
I
don't
know
if
this
is
too
controversial,
but
worth
taking
a
look.
If
you
have
some
time,
the
next
one
is
a
periodic
reader.
A
This
has
got
a
little
bit
more
discussion,
it's
a
little
bit
bigger
as
well.
It
looks
like
it
has
two
approvals
at
this
point,
so
it
does
look
ready
to
merge.
I
will
no.
A
Yeah:
okay,
good
point:
I
don't
know
how
specifically
we
need
to
be
on
the
restraint
or
the
criteria
here
on
this
branch,
but
I
agree
like
we
can
wait
on
this
one.
This
was
a
little
bit
more
of
an
involved
pr,
so
we
need
another
approval
here.
I
guess
is
the
answer
to
that.
If
you're
an
approver
on
the
call,
please
take
a
look
if
you're
not
also,
please
take
a
look.
This
is
one
of
the
more
useful
parts.
I
think
that
we're
going
in
the
next
few
pr's
aren't
as
useful.
A
C
I
would
also
just
reiterate:
we
need
people
looking
at
metrics.
We
don't
want
to
stall
just
because
you
have
just
so
few
people
with
eyeballs
on
it.
So
if
you
have
time
that
would
be
very
appreciated.
A
Yeah
agree,
definitely
yeah
if
you
want
to
get
that
number
of
20
on
the
milestone.
This
is
the
would
help
yeah.
So
please
this
is
a
good
one.
Take
a
look
at
next.
One
was
sinking
main.
This
is
a
pretty
big
numbers
one,
but
it
really
is
just
merging
in
a
bunch
of
commits
that
we've
already
made
last.
A
Actually,
I
have
no
idea
how
long
ago
some
of
these
were
from
from
probably
a
month
or
two
ago
before
we
branched
specifically
there's
some
good
ci
system,
things
that
I
wanted
to
get
merged
in.
So
I
think
doing
this
at
a
regular
case
is
probably
a
worthwhile
thing.
This
should
be
pretty
straightforward.
There's
some
small
changes
to
kind
of
like
keep
it
in
sync
and
also
workable,
but
otherwise
it
should
be
just
a
merge
from
me.
A
This
is
another
pr
that
is
adding
a
test.
It's
a
smaller
one,
44
lines
changed.
We
want
to
make
sure
that
we
follow
the
specification.
One
of
the
requirements
from
the
specification
is
that
readers
can
only
be
registered
once
so.
This
adds
a
test
for
that.
So
another
pretty
straightforward.
One
take
a
look
at.
It's
all
about
readers
by
the
way,
and
then
the
final
one
is
just
a
little
bit
more.
A
I've
been
playing
around
with
a
lot
of
the
testing
of
the
readers
working
with
the
periodic
readers
specifically
and
then
also
the
manual
reader.
So
this
switches
to
using
the
testify
testify
suite.
It
cleans
up
a
lot
of
duplicate
code
where
we're
just
shutting
down
or
worth
recreating
from
a
factory
and
uses
some
of
these
set
up
and
tear
down
methods.
It's
extensible,
I've
used
it.
I
can't
remember
it's
in
the
project
performance
in
the
other
places.
Usually
suites
are
not
really
that
useful.
A
I
I
think
that
if
you're
going
to
reuse
it
in
multiple
locations
like
in
this
case
where
we
have
a
manual
reader
and
we
have
a
a
periodic
reader,
I
think,
if
there's
value
in
it
and
then
especially
if
you
have
set
up
or
tear
down
that's
just
going
to
be
duplicate
code.
I
think
there's
value
in
it.
So
I'm
making
this
proposal
here
to
switch
to
doing
that
yeah.
So
a
lot
of
conflicts
there.
A
I
have
to
dissolve
in
a
little
bit,
but
that's
pretty
much
all
the
big
prs
I'll
pause
there
really
quick
and
see.
If
there's
any
questions.
A
Okay,
then
I
guess
the
only
other
thing
I
wanted
to
pass
by
people's
awareness
before
emerge.
I
think
this
is
an
emerging
state.
Is
this
lint
pr?
So
this
is
two
approvals.
It's
essentially
making
more
enforcements.
There's
file
changes.
We've
talked
about
this
in
the
past.
It
is
kind
of
setting
project
standards
so
like
if
there's
any
some
serious
opinions
about
it.
I
want
to
make
sure
that
I
vocalize
that
this
change
is
going
in
to
make
sure
that
happens.
Also,
you
know
just
aaron's
asked.
A
Also
we
don't
be
adding
things
all
the
time.
I
think
that,
like
one
of
the
other
things
that,
like
you
need
to
keep
in
context
here,
is
that
if
we
do
find
any
of
these
to
be
hindrances
more
than
helpful,
then
I'm
all
about
removing
them.
So
just
keep
that
in
mind
as
well,
but
yeah
just
a
heads
up.
I
don't
want
to
merge
this
too
quick,
but
that
said,
it
is
almost
20
days
old,
so
yeah.
I
just
wanted
to
keep
you
all
impressed
with
the
situation
there.
C
C
That
being
said,
I'm
willing
to
hit
the
button
up
for
you
if
you're
still
afraid
to
tyler.
That's
fine.
A
No,
I
was
thinking
of
merging
it
in
later
on
today.
If
you
want
to
merge
it
now.
That
sounds
good
too.
I
do
think
that
you're
right
there's
like
going
to
be
a
few
different
things.
I
do
think
that
it
does
help
structure
to
be
more
user
friendly.
Like
there's,
definitely
a
lot
of
like
auto
linking
for
comments.
It
also
helps
developers
because
one
of
the
things
I
hate
doing
is
just
repeating
comments
that
are
just
linting
style,
which
are
kind
of
not
really
adaptable
to
it.
A
So
yeah,
that's
really
where
the
benefit
comes
so
yeah.
A
And
cool
with
that,
we
can
jump
into
josh.
You
have
comments
here
hand
it
off
to
you
hi.
D
I
I
am
changing
the
topic
from
the
new
metrics
to
what
we
have
today.
I
know
that
really
nobody
wants
to
get
involved
in
that
code.
So
I'm
asking
a
question
of
the
group.
I
have
had
a
branch
on
pr2393
from
months
ago
with
a
prototype
for
the
exponential
histogram
that
was
used
to
qualify
the
design
phase.
I
believe
that
code
is
like
solid
and
ready
to
go.
D
It's
about
800
lines
of
data
structure
and
800
lines
of
test,
given
that
we
already
merged
the
mapping
functions
which,
which
were
was
one
of
the
things
we
got
done
already.
D
That
review
has
to
happen
eventually
by
someone
anyway,
so
I'm
planning
to
call
a
light
step
engineer
to
do
the
data
structure
review
there,
but
the
rest
of
this
change
would
be
rather
ugly
stuff
involving
the
otlp
exporter,
the
export
record,
the
reagent
interface.
Everything
needs
to
change
a
little
bit
to
allow
production
of
these
exponential
histograms,
and
so
it's
going
to
be
a
pr
that
no
one
really
wants
to
look
at.
D
My
question
is,
if
I
put
it
together,
will
you
look
at
it
and
will
we
possibly
be
able
to
get
that
merged
in
a
kind
of
given
the
the
stakes
are
not
very
high,
a
user
would
have
to
configure
exponential
histograms
to
get
them
and
we're
still
not
stable.
So
I'm
just
questioning
how
much
effort
is
worth
that
it's
worth
to
try
and
do
this
to
get
something
in
ahead
of
the
new
sdk.
D
D
A
D
I
think
it
would-
and
I
I
delicately
want
to
introduce
the
other
idea
that
we've
had
lightstep
basically
has
a
customer,
that's
bearing
down
saying
we
thought
we'd
get
exponential
histograms
a
while
ago
for
that
customer.
What
I'm,
what
I'm
proposing
that
we
do
right
now
and
I'm
going
to
say
it
now.
I
think
we
should
publish
a
light
step
sdk
just
to
get
our
customer
happy
right
now,
which
means
taking
the
implementation
that
I
drafted
and
publishing
it
in
the
open,
telemetry
launch
or
go
repository
of
light
step,
light
steps,
hotel
launcher!
D
Go
that
way.
I
could
also
prove
that
we
have
an
implementation
that
works
just
waiting
to
get
caught
up
with
the
community
sdk.
So
that's
probably
what
I'll
end
up
doing
I
just
wanted
to
make
it
look
like
I
wasn't
favoring.
You
know
progress
over
the
community
here.
So
if
people
are
willing,
I
will
backport
the
2393
histogram
code
to
the
current
sdk
so
that
it's
there.
A
A
I
I
can't
I
can
only
speak
personally
and
I
just
I
can't
commit
to
reviewing
something
like
this,
but
I
also
don't
want
to
speak
for
the
group,
so
I
would,
I
would
say,
like
if
you
can
get
the
community
support,
to
get
some
approvers
to
review
it.
I'm
no
problem
merging
things
but
yeah.
I
I
definitely
cannot
commit
to
or
doing
something
like
this
at
this
point,
I
definitely
think
that
we
want
this
in
the
the
new
sdk.
A
D
A
I
gotcha,
I
thought
dave
would
shake
his
head
for
wanting
this,
but
maybe
it
was
unrelated
when
everybody
thinks.
A
Okay,
cool
yeah-
I
I
think
josh,
I
think
that's
a
good
approach
then
just
maybe
like
including
this
in
the
the
go
launcher
for
light
step
and,
like
you
said,
then
you
have
a
working
example
and
it's
much
easier
to
to.
You
know,
motivate
the
inclusion
here
at
that
point.
So
yeah,
I
think
that's
probably
the
way
I
would
go.
But
again,
if
you
have
volunteers
that
want
to
review
this
okay
community.
D
A
You're
helping
in
that
charge
to
make
it
faster
thanks
for
all
the
reviews.
By
the
way,
I
will
continue.
Thank.
D
A
Okay,
we
have
made
it
through
the
agenda.
I
can
pause
here
if
there's
anything
else,
somebody
wanted
to
bring
up.
I
know
people
probably
want
to
jump
off
and
I
know
coupe
consoles.
E
I
got
I
got
one
thing
I
want
to
add:
tyler
just
froze,
I
don't
know.
If
I
just
cut
you
off,
I'm
gonna
keep
going
for
the.
E
What
was
the
benchmark
stuff.
I've
got
that
cnc
f
issue
out
there
asking
for
the
hardware,
and
I
got
an
approval
really
fast,
but
then
the
no
one
else
got
an
approval,
so
I
pushed
it
again
today.
E
I
also
asked
that
damien
matthew
be
given
access
as
well,
because
most
likely
he'll
be
the
one
to
run
with
this
going
forward.
I've
just
been
focusing
more
on
the
collector
than
I
have
been
in
the
go
space.
I
don't
want
to
that
that
work
to
be
dropped,
and
I
know
that
damian
was
interested
in
it.
So
I
did
push
cncf
still
today.
E
A
Awesome
yeah
thanks
to
the
update
progress
on
that.
Even
though
ball
is
not
work,
yeah.
C
E
Yeah
with
jurassic
being
gone,
it's
also
a
little
harder
because
he
has
done
all
this
before
so
maybe
when
he
gets
back
that'll
help
us
move
the
think
board
as
well,
but
we'll
see.
C
Tyler,
your
internet
has
been
breaking
up,
but
I'm
going
to
step
in
and
kind
of
take
over.
If
that's
okay,
we've
hit
the
end
of
the
agenda,
as
he's
mentioned,
is
there
any
other
updates
that
anyone
he
wants
to
bring
up.
C
So
that's
the
end
of
the
scheduled
meeting.
Is
there
anything
else
anybody
wants
to
talk
to
talk
about?
This
is
kind
of
open
forum,
time.
C
And
then,
with
that
I'll
give
you
a
little
over
half
an
hour
back.
I
got
the
easy
part
of
the
meeting.