►
From YouTube: 2022-10-07 meeting
Description
cncf-opentelemetry@cncf.io's Personal Meeting Room
B
A
A
A
Right,
yeah
I
guess
we
told
them
don't
create
PRS
without
that
don't
have
issues
and
that
what
they
do
they
go,
create
an
issue
and
then
create
a
PR.
Well,
at
least
it's
something
at
least
I
mean
they
try
to
describe
what
the
problem
is.
You're
right,
I,
don't
know,
I,
don't
think
we
can
prevent
this
behavior
in
some
sense,
I.
B
A
B
A
Okay,
anyway,
the
for
this
specific
one,
we
already
have
I
think
one
or
two
ready,
specific
semantic
conventions
in
the
database
conventions.
There
is,
there
is
a
couple
we
have
some
Cassandras
and
I
think
we
have
a
couple
of
vertices.
So
what
they
want
is
I'm,
not
sure
what
the
TP
instance
for
red
is.
What
does
that
mean.
A
So
I
anyway
I
guess
we
don't
have
to
go
into
the
details
of
the
pr
the
point
I
like
I
guess
we
already
have
read
this
specific
semantic
conventions,
so
what
they
are
really
looking
is
that
enhancing
it
essentially
not
creating
a
new
kind
of
a
resource,
so
I
would
say,
seems
okay
to
me
right,
I,
don't
see
what
the
problem
is.
I
mean
the
specifics
are
important,
what
they
are
adding.
Maybe
it's
good,
maybe
it's
bad,
but
that
that's
that
we
can
discuss
in
the
pr
yeah.
B
A
B
Yeah,
so
I
just
want
to
count
one
thing:
if
you
look
at
the
pull
request
there,
there
are
many
semantic
conventions,
and
some
of
them
has
been
there
for
more
than
a
week,
and
now
they
reply
so
I,
I,
guess
and
some
of
them
people
have
strong
opinion
that
it
should
be
blocked.
They
just
went
ahead
and
blocked
it.
I
wonder
if
we
keep
doing
this.
This
guy,
who
filed
maybe
like
seven
PRS,
would
really
be
unhappy
with
open
television,
eventually
we'll
just
go
away
and
I
feel
yeah
that
might
be
even
worse
than
using.
A
Plenty
exactly
yes,
that's
that's
not
good
right.
We
don't
want
to
discourage
these
people.
They
they
want
to
this.
This
person
like
they
they
try
to
they
they're
engaged.
They
try
to
actively
move
problems
forward
like
make
progress
right,
so
I
I,
don't
want
to
just
tell
them
no
go
away.
That's
not
good
right!
So
yeah.
B
A
So
for
this
I
think
next
week
we're
scheduling
the
the
meeting
of
the
the
board
group,
that
is
about
semantic
conventional
stabilization
and
Joshua
and
I
were
discussing.
How
can
we
unblock
these
people?
It's
not
yet
clear
to
us,
but
we
want
to
try
to
find
a
way
right
and
I.
I
also
invited
that
guy,
who
created
all
of
those
issues
to
also
attend
the
meeting
so
I
guess
that
is
another
way
to
kind
of
show
them
that
we're
trying
right
we're
trying
to
find
out
well
find
a
way
right.
A
B
It's
not
saying
like
we,
we
absolutely
accept
everything
you
have
but
anyways
for
that
right,
as
I
I
feel
probably
we're
fine.
Let's
look
at
the
next
one
yeah.
A
B
B
Yeah,
while
I'm
opening
the
link
I
have
a
meta
question,
so
you
see
like
this
guy.
He
obviously
followed
the
process
we
have
and
if
we
allow
that
PR
that
that
radius,
one
and
we're
saying
no,
then
people
will
eventually
figure
out.
Oh,
if
I
just
go
ahead
and
create
a
PR
it'll
be
like
disapproved
yeah,
we
will
just
ignore
it.
So
like
it's
like
really
a
very
interesting
situation,.
A
Yeah,
so
for
for
this
is
about
phrases
right,
so
this
is
going
to
be
a
trace
systematic
convention,
particularly
for
traces
I,
think
our
only
concern
was
that
we
don't
add
semantic
conventions
with
very,
very
high
cardinality,
which
we
will
later
need
to
also
support
in
metrics,
somehow
and
end
up
in
a
bad
situation,
because
we
can't
right
so
if,
if
it's
a
low
cardinality
I,
don't
know
what
exactly
they
are
asking
for,
but
we
do
have
RPC
and
if
they
want
to
add
another
attribute
that
is
also
not
high.
B
A
B
A
B
A
A
Yeah,
so
is
this
is
equivalent
to
us
having
a
semantic
Convention
of
HTTP
headers.
We
have
one
right
if
I'm
not
wrong,
yeah.
Do
we
so
when,
when
you
want
to
record
the
span,
that
represents
an
HTTP
connection,
you
know
you
have
method,
you
have
a
URL,
you
have
headers
right,
so
this
I
think
is
equivalent
to
that.
If
I'm
not
wrong,
yeah.
B
A
C
B
C
B
A
A
I
did
create
it,
it's
kind
of
a
kind
of
more
of
an
idea
for
the
future
I.
Don't
think
we
want
to
do
it
right
now,
but
I
thought
it's
an
interesting
thing
to
capture
I,
don't
plan
to
work
on
this
myself
right
now
and
I,
don't
know
if
anybody
needs
to
work
on
this
right
now.
So
what
do
we
do
with
these
sort
of
things
like
something
that
I
think
probably
would
be
interesting
for
the
roadmap
of
logging?
But
it's
not
an
immediate
need.
Do
we
like?
B
I
think
the
current
definition
of
acceptance,
if
someone
sending
a
PR,
then
we're
obligated
to
review
and
how
to
merge
that
I
feel
this
one
is
probably
something
we're
saying:
this
is
a
good
idea.
We
need
more
discussion
and
maybe
it'll
end
up
with
an
old
Tab
and
and
based
on
the
whole
type.
We
might
have
multiple
PR's,
so
it
seems
like
something
to
explore.
We
need
more
information.
A
A
C
B
A
C
By
the
way,
these
things
that
we
are
like
interpretation
of
what
these
levels
mean
I
think
we
need
to
add
them
to
the
dogs.
You
know
just
it
tastes
like.
Let's
say
that
somebody
is
like
tragically
important
and
I
mean,
and
you
know
it's
not
clear
to
them,
but
yeah
I,
like
your
interpretation
really,
but
just
let's
document
that.
C
A
B
C
A
B
A
B
Single
one,
so
so
imagine
if
you
have
like
otltx
powder,
the
R2
cases,
one
is
you
have
two
otltx
partners
in
your
application
and
both
of
them
are
sending
data,
but
one
is
Foxy
on
trace
and
otherwise
on
metrics
and
you
only
want
to
compress
metrics
or
you
want
a
different
compression
for
trees.
The
second
scenario
is,
you
have
multiple
otlp
exporters
that
are
sharing
the
underlying
connection,
so
so
Precision
metrics
can
be
merged
together
on
a
single
transport,
and
then
it
seems
murky
here.
What
does
compression
mean.
C
A
B
Maybe
there's
no
reason
to
not
do
that
or
do
you
think
like?
Why
would
people
like,
if
you
ever
like
all
the
younger
exploiters,
are
doing
the
disabled
compression
by
default?
Why
would
people
want
to
turn
that
off?
Is
that
because
they
have
some
firewall,
they
want
to
do
some
level,
7,
filtering
or
or
it's
because
they're
using
something
very
old.
The
service
wouldn't
even
accept,
so
they
have
to
go
back
to
now.
A
B
So
maybe
I
also
just
when
we
let
people
respond
here,
telling
us
whether
we
see
the
need
yeah
and
why
and
if
there
is
I
agree.
A
C
C
B
C
B
A
C
C
B
A
B
A
A
It's
not
all
TLP
the
Java's
implementation
I
would
expect
it
to
be
all
TLP
Json,
and
it
is
also
wrong
what,
from
what
I
see?
That's,
not
all
TLP
Json
either.
That
is
not
the
right
format
for
OTL
PJs
or
the
attributes
are
not
specified
like
that.
They
should
be
done
in
a
different
way,
so
I
think
we
probably
need
to
do
two
things.
Make
it
clear
that
the
example
has
nothing
to
do
with
our
TLP
and
also
then
they
have
a
problem
in
Java,
so
I
see
so
yeah.
B
Can
make
sure
that,
for
the
clarification
are
you
thinking
we
should
either
statement
saying?
No,
the
following
equation
is
just
to
show
how
the
data
would
ideally
look
like,
but
it
has
no
mean
to
be
the
real
wire
protocol.
Yeah.
A
B
I
hate,
similar
problem,
Microsoft
and
and
a
lot
of
people
decided
they'll
just
take
because
a
cold
snap
head
it
can
catch
your
eyes.
A
lot
of
people
won't
read
like
with
this
message
at
all
and
we
ended
up
deciding
okay,
because
Json
is
part
of
the
wire
protocol.
Let's
just
put
yaml
here,
some
people
don't
be
confused.
Yeah.
Do
you
like
the
idea
of
communication
to
yaml
or
or
you
think,
I
need?
A
disclaimer
will
be
good
enough.
Yeah.
A
Yeah
yeah
I
I
can
try
to
fix
that
either
change
it
as
you're,
saying
to
some
other,
like
yaml
format,
or
maybe
maybe
actually
change
it
to
all
TLP
Json,
which
is
maybe
a
bit
less
readable,
but
in
that
case
there
will
no
longer
be
a
confusion.
I
can
take
that
part,
but
what
they
are
asking
about.
Java
is
Java's
problem.
Java
is
doing
it
wrong,
so
they
need
to
go
back
and
fix
it.
I'm
guessing
the
Java
should
be
all
TLP
Json.
It
shouldn't
be
arbitrary,
Json
format.
A
B
A
Yeah,
but
that
means
it's,
it
has
nothing
to
do
with
our
trp
Json.
So
what
is
it
that
they
are
expecting?
This
person
is
expecting
some
sort
of
compliance
right
and
compliance
of
an
implementation
with
with
an
example
which
is
unreasonable,
expectation
I
mean,
but
so
we
need
to
clarify
both
the
example
to
make
sure
people
understand
it.
It
is
not
a
standard
and
also
the
implementation
needs
to
comply
with
something,
but
what
is
it
complying
with
is
unclear
here
it's
not
complying
with
anything
right.
Did
they
invent
a
format
for
themselves.
B
B
A
Yeah,
do
you
do
you
want
to
keep
it
open
assigned
to
me?
I
will
make
the
clarification
of
the
example
as
a
result
of
that
I
can
do
that
right
so
that
we
don't
create
any
another
new
issue.
B
A
A
Like
yeah
accept
it
because
I'm
going
to
work
on
it,
thank
you
we're
good.
Thank
you
have
a
good
one.
Thank
you.