►
From YouTube: 2023-01-18 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
A
C
A
C
B
A
B
Pretty
small
group,
okay,
kinda,
there's.
C
B
Sure
yeah
I'm
actually
refreshing
my
memory
about
what
we
talked
about
last
week.
I
mean
certainly
there's
the
one
for
conversations.
B
Food
cars
brought
up
the
pr
that
I've
been
reviewing
I
think
we've
got
a
good
path
on
it.
That's
from
the
contribute,
though,
868
I
think
it's
got
one
more
like
test
failing,
but
besides
that
I'm
pretty
much
ready
to
approve
it.
B
Anybody
else
wants
to
take
a
look
at
868.
They
can
it's
adding
enrich
to
the
SQL
instrumentation
out
there.
Oh.
B
And
I
talked
about
package
naming
I
actually
still
have
it
to
I
started
compiling
all
this,
but
I
haven't
gotten
in
a
form
to
to
plop
on
that
issue.
I
have
an
issue
open
regarding
package,
naming
it's
issue.
869.
C
B
Opened
this
a
while
back
for
us
to
consider
maybe
changing
the
name
of
some
of
our
packages,
namely
the
ones
that
we've
yet
to
stabilize,
but
obviously
rename
it
prior
to.
B
I
was
gonna
kind
of
reformat
this
issue,
because
I
don't
think
it's
super
easy
to
read.
I
was
going
to
make
a
table
out
of
this
and
I
kind
of
started
doing
that
work,
but
I
should
have
that
updated
here
soon.
B
This,
namely,
came
out
of
the
conversation
about
the
docker
resource
detector
package,
which
apparently
is
no
longer
just
like
docker,
so
yeah.
It
was
originally
called
open,
telemetry.extensions,
Dot,
Docker,
okay,
so
I've
kind
of
brought
the
term
extensions
into
question.
I
think
it's
kind
of
overused.
C
B
C
Okay,
there
is
already
a
PR
doing
this
same
thing:
okay,
yeah
paid
attention
to
all
the
PMS
so
how
to
take
care
of
one
by
one.
B
In
my
opinion
on
this,
PR
was
that
we
I
I
think
my
personal
opinion
is
that
I
think
this
is
a
good
name,
open,
telemetry.resource,
detector.container,
but
I
think
it
would
be
useful
to
just
kind
of
at
least
go
through
the
exercise
of
considering
the
other
packages
that
we
might
want
to
rename
as
well,
and
just
you
know,
get
a
sense
of
whether
going
through
that
process
influences
our
opinions
or
ideas
about
what
this
one
should.
A
B
C
B
C
Yeah,
the
process
is
a
little
bit
more
involved
because
we'll
have
to
go
to
the
new
kit
and
do
some
stuff,
so
that,
like
if
someone
comes
here,
they'll
see
that
this
is
duplicated
and
replaced
with
something,
and
so
we
need
to
do
like
some
manually
work
there.
So,
but
we'll
do
it
like
when
we
actually
find
this
yeah
I,
don't
see
anything
else
in
the
agenda,
so
maybe
I
just
want
to
get
a
sense
of
like.
Where
are
we
with
the
releases
I
did
see
their
Milestones
were
updated.
C
So
if
no
one
has
other
topics,
maybe
Eleanor
Blanchard,
who
whoever
asked
the
most
accurating
over
the
most
latest,
can
you
like
update,
and
what
is
the
general
plan
like
here
like?
Are
we
still
on
track
for
this
one?
There
are
like
few
new
packages
being
introduced,
I'm
still
going
through
it,
so
I
haven't
done
anything
but
I
reached
a
long
time
for
like
1.4.
And
what
did
we
decide
about
this
package?
Is
it
going
to
be
1.4,
so
it
will
be
released
along
with
the
rest
book.
B
D
I,
don't
know
if
that's
offense
or
release
it,
certainly
something
we
could
approach
after
one
of
the
internal
groups
at
Microsoft
just
trying
to
upgrade
to
1.4
and
you're
having
the
package
downgrade
problem,
so
we're
looking
at
what
we
could
do
there
self
as
a
blocker.
So
it's
I
think
it's
on
target.
But
there
are
some
balls
in
the
air.
B
B
And
then
also
that
version
downgrading
problem.
B
C
Okay,
so
like
is
it
decided
that
we'll
be
releasing
1.4
stable
of
extensions
dot
hosting
package
along
with
the
core,
or
this
is
still
undecided.
D
D
C
What
was
the
name
of
the
new
package?
It's
a
dependency
injection,
so,
okay,
it
will
be
the
same.
So
I'll
end
like
that
renaming
proposal.
Will
it
affect
this
one
like
here
we
have
used
the
extensions
dot
dependency
injection,
any
thoughts
on
like
whether
this
might
be
affected
by
the
renaming
thing.
B
You
know
they:
they
basically
mirror
the
microsoft.extensions
dot,
dependency,
injection
and
Microsoft
dot
extensions.hosting.
So
in
that
case,
in
the
end,
also,
in
that
sense,
it's
like
they're,
actually
extension
methods
that
these
dependencies
are
providing
yeah.
So
this
was
like
the
one.
This
was
the
one
like
Flavor
of
kind
of
packages
that
we
had
where
we
used
the
term
extensions
that
I
thought
was
like:
okay,.
A
C
Makes
sense?
Okay,
so
it
looks
like
the
the
milestones
we
have
sorry
milestones
in
this
place
used
to
like
great
I.
It
says
around
two
more
weeks
from
now,
so
we're
going
to
change
it
to
X
update,
but
otherwise
we
can
keep
it
and
we
do
not
have
the
date
for
log
exporter
as
of
now.
It
will
definitely
be
after
1.4.
C
This
one
I
think
Martin
had
some
suggestion
about
potentially
renaming
it
to
like
0.5
or
0.6,
and
call
it
stable
with
an
explicit
indication
that
it's
stable
from
an
AP
standpoint,
no
Breaking
Chains,
but
shape
of
the
Telemetry
would
change,
but
that's
something
which
we
will
discuss
in
a
week
from
now,
okay,
yeah
and
like
once,
1.4
is
done.
B
C
We
can
start
with
1.5,
and
this
is
okay-
we
don't
have
any
date
put,
but
we
can
only
do
it
once
we
done
with
one
also
okay,
any
like
anything
else
which
we
were
like
discussing
and
left
in
the
middle
from
previous
weeks
or.
C
So
Alan
I,
so
you
had
this
exponentials
yeah
like
like
any
any
comments
about
like.
Is
it
it's?
Definitely
not
1.4
right,
so
yeah.
A
C
Going
to
be
after
okay,
so
maybe
we
can
update
the
Milestones
to
indicate
like
1.5,
would
have
exponential
histogram
as
well.
If
you
are
like
relatively
confident
that
it's
it's
going
to
happen,.
B
B
That
was
interested
in
them,
and
so
that's
why
I
just
kind
of
like
pushed
up
that
PR
just
to
say
like
hey
this
is
this
is
roughly
what
the
work
entails
like
if,
if
the
spec
stabilizes
I
was
saying
at
that
point
in
time,
we
could
try
to
get
into
one
four,
but
I
don't
feel
strongly
about
that.
I
think
it's
at
this
point
in
time.
It
it
makes
sense
to
definitely
wait
until
one
five.
Okay,
that
said,
there's
another
PR
out
there
that
I
have
that's
still
just
kind
of
like
keeping
it
all.
B
Internal
and
I
would
actually
like
to
get
that
merged.
If,
if
other
folks
are
okay
with
that
yeah,
it's
that
aggregate
count
yeah
this
one.
This
doesn't
affect
anything
public
and
it's
still
like
exponential
histograms
can't
be
used,
but
it
does
just
start
to
like
Plumbing.
More
of
the
more
of
the
work
I
was
thinking
about.
B
I
think
I
got
a
maybe
another
PR
or
two
that
will
expand
on
this
a
little
bit
just
kind
of
get
more
of
the
more
of
the
functionality
plumbed
into
the
SDK
and
you
know,
add
some
more
test
coverage
and
so
on.
C
I,
don't
know
whether,
like
someone
did
explain
like
how
the
whole
exponential
histogram
works,
like
really
did
mention
few
months
back,
that
he
will
do
a
presentation
to
this
group
because
at
peers
we
basically
merged
without
like
very
detailed
review,
because
it's
anyway
not
part
of
the
actual
ship
product.
I
mean
it's
not
enabled
anyway,
so
that
has
never
happened
right,
like
I
assume
that
really
was
also
on
vacation.
So
it
may
not
have
happened
at
all
so
yeah
before
we
like
do
the
actual
shipping.
C
We
can
ask
if
I
need
to
do
a
quick
presentation
of
how
this
exponential
stuff
works
internally,.
C
Yeah
a
lot
of
heavy
Max
yeah.
It's.
B
A
C
Yeah
so
I,
just
updating,
like
my
own
status
I,
am
like
fully
back
to
work
as
of
last
week
mid
last
week,
so
you
can
expect
to
find
me
a
fully
active
this
week
onwards.
I'm
still
like
I
have
like
about
a
hundred
tabs
open,
I
I
can
see
like
so
many
of
them.
I
want
to
just
finish
at
least
half
of
them
before
I
start
actual
coding
so
but
like
workways,
I
I
would
be
working
on
XM,
plus
I.
Think
we
committed
it
for
1.5
I.
C
Think
I,
myself
updated
that
Excel
class
would
be
in
1.5,
so
I
don't
have
like
any
active
thing.
I
have
PPS,
but
I
am
still
going
to
work
on,
like
some
private
branches,
like
just
playing
with
Excellence
how
their
performance
looks
like
with
the
default,
because
there
is
also
some
discussions
in
the
specification
about
like
once
exemplary
study.
Is
it
going
to
be
on
by
default,
or
is
it
off
by
default?
I
am
still
doing
some
experiments,
so
I
think
by
mid
next
week,
I'd
have
some
PRS
which
are
ready
for
March.
C
B
C
I
think
the
problem
is
yeah.
We
have
to
help
implement
it
because
usually
they
ask
for
like
at
least
two
or
ideally
three
languages
to
implement
respect
before
calling
things
stable.
C
So
if
the
overall
matrix.net
was
one
of
the
languages
which
helped
with
the
prototyping
so
which
internally
or
which
in
turn
enable
the
spec
to
be
called
stable,
but
for
x
and
plus,
it
don't
think
any
other
language
other
than
Java
has
implemented
it.
C
C
plus
plus
has
some
implementation,
but
I
don't
know
whether,
like
the
security
to
support
declaring
respect
as
table
so
I
just
met
with
Riley
yesterday
and
told
that
I'll
try
implementing
it
in.net
and
give
my
opinion
on
whether
a
respect
is
Good
from
our
perspective
from.net
perspective
in
the
next
two
weeks.
So
so
it's
not
a
question
of
like.
When
will
the
spec
be
stable,
it's
more
like
when
will
be
implement
it
and
tell
that
we
are
ready
or
we
are
okay
for
respectively
table
assist.
C
So
that's
what
I
think
it's
going
to
happen
like
we
just
have
to
implement,
see
if
things
look
good
and
then
let
the
like
Josh
and
Riley
were
packing.
It
so
let
them
know
that
we
have
go
to
declarextable
from.net
standpoint,
and
they
can
actually
do
this.
C
B
Nope
I
yeah
I
just
haven't
paid
super
close
attention,
though
the
one
thing
that's
in
the
back
of
my
mind
is
regarding
exemplars
is:
if
there
is
any
sense
that
it
might
still
be
some
time
before
we
can
ship
a
stable
release
with
exemplars,
then
I
would
like
to
see
the
exponential
histogram
stuff
landed.
C
C
Yeah
I'll
see
like
it
might
be
like
very
easy
to
like
detach
it
like.
If
things
doesn't
work
the
way,
then
we
should
still
be
able
to
retain
all
the
classes
but
keep
it
internal.
So
we
should
be
able
to
figure
out
something
like
that
in
case
X
and
R
gets
delayed,
but
the
exponential
goes
faster.
Oh
user
exponential
is
already
stable,
so
there
is
no
question
of
Delete.
B
Think
as
soon
as
we
ship
one
for
then
you
know
I
I
think
we
could
be
ready
to
like
ship
another
beta
that
has
exponential
histograms
in
it.
C
A
C
Okay,
so
I'll
try
to
put
the
xmlr
code
along
with
the
betas,
and
if
there
is
a
significant
delay
in
the
spec
side,
aborted
sometimes
I'll
try
to
pull
it
out
and
keep
it
internal.
So
that
was
my
original
plan
so
like
when
I
submitted,
like
the
thing
I
did
like
really.
Basically
I
made
sure
that,
like
it's
not
going
to
affect
anything
like
it's
not
connected
to
anything,
you'll
have
filters
and
yeah.
C
All
of
them
would
be
there
as
classes,
but
it
won't
be
connected
to
the
Mothership,
so
it
will
just
remain
Standalone
for
now,
once
I
gain
like
sufficient
confidence,
then
I'll
start
actually
connecting
like
into
the
SMS
anyway.
We
have
like
at
least
two
more
weeks
before
we
can
do
1.5.
So
by
that
time,
I
would
get
a
good
three
down.
How
complex
is
the
whole
Exemplar
stuff.
C
Like
Josh
did
mention
that
it's
not
that
hard,
but
I
am
already
done
with
the
filter.
Part
I
haven't
looked
at
the
server
yet
so
maybe
once
I
do
that
I'll
be
able
to
give
a
more
complete
timeline
about
how
complex
it
is.
C
All
right,
nothing
else,
so
we
will
see
you
all
next
week.
C
I'll
ask
Martin
to
prepare
like
his
proposal
about
the
Discrimination
Library
questioning
before
the
next
one.