►
From YouTube: 2022-08-29 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
B
A
A
A
B
I
have
added
some
of
the
specs
update,
which
probably
we
just
need
to
have
and
look
into.
They
have
not
yet
merged,
but
something
I
do
probably
which
can
affect
our.
We
may
have
to
do
some
changes
if
they
get
merged
so
and
then
I
have
added
something
on
matrix.
I
think
probably
the
benchmarking
is
something
we
have
to
add.
We
have
to
do
it
or
to
make
the
issues
for
that.
It
would
be
important
to
have
some
numbers
for
us.
A
B
Okay,
I
think,
probably
for
the
specs
updates.
I
think
events
and
long
api
is
something
which
is
being
still
the
pr
is
there
and
I
think
it's
still
not
merged.
Still,
it's
open
so.
B
B
B
B
A
B
Okay
and
then
boundaries
there's
another
one.
The
default
boundaries
in
the
explicit
bucket
histogram
are
not
optimized
for
latency,
which
makes
sense.
I
mean
we
define
the
boundaries
to
start
with
minus
infinity
to
zero,
but
the
expression
is
in
histogram
only
supports
the
non-negative
values,
so
the
recommendation.
What
probably
the
suggestion
is
that
we
should
have
boundaries?
Something
like
this,
so
we
have
to
change
our
default
boundary,
so
this
gets
smudged.
B
B
So
if
you
declare
the
nm
can
be
center
string,
then
we
are
blocking
all
random
medicines
in
json
decoding
for
an
instance
or
not
present.
So
if
it
is
empty,
then
I
think
it
will
cause
error
for
any
new
values
for
any
new
nm
types
added.
If
those
values
are
empty
in
the
previous
implementation
that
that
will
cause
error.
A
A
B
B
B
A
A
B
B
B
Which
must
be
a
non-negative
integer
value,
so
we
do
I
mean
so
we
do
have
long
value
and
we
have
double
value.
So
if
it
is
a
long
value,
it
can
be
negative
and
not
it
can
be
non-negative
and
negative,
but
for
counter
operation.
If
we
have
to
do
a
check,
validation
check
that
if
it
is
a
counter,
we
are
creating
a
counter
type
of
measurements
for
the
instrument,
then
it
should.
The
record
should
not
take
non-negative
values.
It
should
only
take
negative.
It
should
only
take
positive
values.
Sorry
yeah.
A
B
A
B
Do
anything
in
that
or
if
we
create
a
yeah,
I
think
it
won't
be
possible
to
move
again
so
yeah,
let's
see
just
I'll
just
create
a
issue
for
that.
B
A
B
A
B
So
then
there
is
another
one
at
matrix
producer
to
allow
matrix
creator
to
collect
from
third-party
maps
this.
If
this
gets
sub
merged.
This
is
this
pr
get
merged.
Then
probably
we
need
to
add
a
different
another
entity
called
as
matrix
producer
again
it
is
not
properly
not
something
which
we
have
to
support
it
in
here,
but
I
think,
even
after
that,
we
can
do
it
so
that
vertex
producer
would
allow
us
to
embed
different
readers,
matrix
leaders
which
are
collecting.
B
Probably
this
one.
Definitely
we
have
to
stop
looking
the
validation
check,
because
this
has
been
something
which
is
existing,
so
this
definition
be
part
of
ga,
and
I
think
this.
B
B
Basically,
for
all
the
aggregations
and
all
that
offer
for
all
the
irrigation
for
all
the
kind
of
instruments
there
are
also
combinations,
will
come
into
the
picture,
different
aggregations
with
different
instruments
with
different
aggregations
and
then
also.
B
A
B
B
B
B
B
This
is
work
in
progress.
Etw
exporter
review
references.
There
were
some
conflicts.
Let
me
fix
it
for
everything
on
the
sheet.
Do
we
need
to
support
this
reference?
Soft
reset
world
of
function?
I
mean
the
reason
why
we
created
this
pr
because
of
some
issue
reported
internally,
but
I
think
there
is.