►
From YouTube: 2022-11-30 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
A
C
B
So
I
submitted
one
issue
in
this
pack:
repo
about
asking
for
ideas
to
represent
object,
schema.
C
Okay,
yeah
I
I
saw
that
last
night.
Also
thanks
for
doing
that,
Santosh
and
I
know
we
had
last
summer
discussion
with
with
other
folks
in
the
laws
big
right,
correct.
B
C
Okay,
I
was
you
know,
one
thousand,
maybe
I
didn't
you
know,
spend
enough
time
reading
through
that
I
don't
know
if
that
came
across
really
clearly,
in
that
you
know,
perhaps
we
could
reword
or
something
is
what
I
thought
initially
do
we,
oh
as
an
example,
you
have
pointed
to
that,
so
hopefully
people
will
be
able
to
see
them.
Okay,.
C
And
you
one
other
question
I
have
there
was
you
had
mentioned
some
machine
digital
format?
Json
schema
exist,
11
or
abroad,
or
something
like
that
right
does
that
matter
or
you
know
it.
A
B
Well,
I
think
there
are
two
reasons
right
to
you
know.
There
are
two
motivations
why
we
want
to
do
this
schema
one
is
you
know
we
on
the
agent
side?
We
want
to
make
sure
that
you
know
everybody
knows
what
we
are
publishing,
what
the
instrumentation
is
publishing
in
in
these
objects
and
secondly,
for
the
receivers
you
know,
given
that
we
are
most
of
the
ram
data
is,
is
unauthenticated,
you
know
we
we
want
to.
B
You
know,
be
more
cautious,
accepting
data
than
the
APM,
the
back
end
telemetry,
because
that
comes
authenticated.
So
you
can
trust
a
little
bit
because
it
comes
from
known
sources
while
for
rum.
While
you
cannot
entirely
eliminate
you
know,
data
from
untrusted
sources,
but
at
least
you
know
in
our
past
experience
we
noticed
you
know
a
lot
of
you
know.
We
receive
a
lot
of
garbage
now.
A
B
Messes
up
some
of
the
you
know
display
for
our
customers.
So
at
least
you
know,
we
want
to
ensure
that
you
know
we
validate
the
data
we
received
and
drop
it,
and
for
that
purpose
you
know
if,
if
things,
if
the
schema
is
represented,
is
in
one
of
these
standard
formats,
there
are
tools
already
available.
B
You
know
for
those
standards,
this
Json
schema,
Avro,
Nexus,
11.
and
those
three
terms
I
picked
up
from
cloud
events,
so
Cloud
events
schema.
You
know
they.
The
specification
says
this
schema
has
to
be
one
of
those
three.
C
I
see:
okay,
just
a
net
comment.
There
I
think
you
know.
Just
out
of
curiosity,
you
know.
Maybe
you
can
clarify
that
Avro,
you
know
is
also
Json
right.
So
I
don't
know
what
what
we
mean
by
Json
schema.
There
Json
schema.
B
In
my
opinion,
no
no
there
is
a.
There
is
a
another
standard
called
Json
schema.
C
B
Yeah
yeah
but
I
specified
using
Json,
but
the
specifications
are
separate.
Okay,.
C
Sounds
good,
the
other
thing
is
yeah
I
agree.
Essentially
what
we're
trying
to
do
is
you
know
we
may
want
to.
You
know
put
that
in
there
we're
trying
to
create
a
strongly
typed,
even
so
to
speak.
So
the
back
end
I
mean
when,
when
we
receive
the
event
the
event
receivers
Can
validate
when
especially
when
they
get
it
from
unauthenticated
to
avoid
garbage
data
and
things
I
think
that's
a
great
point.
C
I,
don't
know
if
it
came
across
there
in
that
issue,
when
folks
view
it
that
might
be
a
useful
thing
to
have
there.
The
reason
why
we
want
to
have
that
strongly
type
is
what
we're
going
for
I
think
right,
yeah
and
then
the
counter
argument
that
people
might
have
to
that
is
that
at
some
point
we'll
invariably
say
there
is
also
custom
section
of
that
schema.
The
schema
allows
end
users,
end
users,
many
developers
right
to
be
able
to
send
whatever
they
want
to
send
outside
of
the
typed
strictly
typed.
C
It's
only
typed
event
schema.
For
example,
a
customer
who's
using
open,
Telemetry
rum
instrumentations
sending
might
want
to
spend
a
bunch
of
custom
properties
along
with
the
page
view.
That's
going
out.
Our
click
that's
going
out
or
something
that's
been
a
regular
thing,
so
we
usually
have
we'll
end
up
having
Properties
or
some
sort
of
a
property
bag
or
something
where
they
can
put
whatever
they
want.
That's
where
not
a
garbage
might
end
up.
C
We
want
to
say
that
is
also
part
of
that
type
that
we
are
defining
right.
Counter
argument
might
be
that
there's
always
going
to
be
you're
allowing
job
to
be
there.
You
know
what
about
that.
So
you
want
to
say
yeah.
We
want
to
control
it
too,
that
that
one
particular
object
or
field
like
that,
which
is
a
collection
we
may
want
to
clarify
that
there.
B
C
I
I
think
that
might
help
you
know.
I,
don't
know
if
it's
because
holidays
or
anything
not
a
lot
of
people
have
looked
at.
It
commented
or
something
like
that.
So
correct.
B
I
think,
irrespective
of
the
motivation
I
think
my
my
main
purpose
of
that
issue
was
to
find
out
if
it's
even
possible,
with
the
current
romantic
convention
stooling
to
represent
schema.
As
you
know,
we
need,
or
rather,
as
you
know,
I
was
thinking
okay.
B
So
there
are
some
examples,
but
I
don't
know
if
it
is
by
accident
or
intentional,
so
yeah.
So,
let's
see
you
know
what
further
comments
we
get.
I'll
also
discuss
in
the
log
sick
today
about
it.
Okay,.
B
And
and
next
week,
When
Nev
is
back
I
think
we
should
discuss.
Irrespective
of
you
know
how
this
goes
into
the
semantic
conventions
in
the
spec
repo
we
should.
We
should
start
discussing
how
we
can
represent
it
somewhere
else
and
freeze
it
and
then
start
working
on
yeah.
C
Some
implementation,
so
I
was
thinking
the
text
to
bring
it
up.
I
was
thinking,
I
know.
Ideally
we
want
to
get
this
get
the
schema
defined
in
yaml
and
then
Auto
generate
the
MD
files
and
and
stuff,
but
perhaps
we
we
don't
do
that
and
just
go
with
the
whatever
the
current
PRS,
the
23
2933.
We
are
attempting
simply
just
Define
them
in
an
MD
file.
C
Knowing
that,
when
reason,
when
we
resolve
this
issue
of
how
to
represent
schemas,
we
can
take
that
data
and
convert
it
to
a
different
format
or
something
that
are
you
available
to
that.
C
Then
we
can
continue
down
that
path.
We
can
use
the
it's
this
documented,
that's
what
we
use
to
code
and
everything
the
instrumentations
and
and
stuff.
Then
we
can
continue
working
on
yeah.
B
Although
I
I'm
expecting
some
pushback,
you
know
I
think
they
couldn't
raised
it
earlier
that
this
is
browser
specific
it
put
it
in.
You
know
somewhere,
you
know
it
like.
It
doesn't
fit
in
the
specification.
Repo
yeah.
C
And
to
me
in
the
last
logs,
like
conversations
I
think
he
may
have
come
around
and
kind
of
understood
what
we
were
trying
to
do
and
that's
what
I
thought
he
was
going
to
go
back
and
look
at
that
PR
one
more
time,
yeah
that
first
one
he
asked
you
to
create
the
issue
correct,
so
we
can
put
some
more
context
around
it.
Hopefully
that
can
go
through
yeah,
okay
and
that's
what
you're
planning
on
bringing
up
in.
A
C
B
Not
I
think
yeah
I
forgot
to
Ping
Ted.
We
need
to
resume
our.
C
B
Yeah
yeah
I'll
ping
it
maybe
next
week
we
can.
A
B
If
there
are
no
General
topics,
T2
I
have
one
question
for
you
in
case
you
know
the
web
vitals
is.
Is
it
already
present
in
any
instrumentation
library
today
in
open
telemetry.
A
Not
in
open
telemet
itself,
but
for
Splunk
we
have
with
the
integration.
Okay,.
B
Got
it
I
see,
could
you
point
me
to
the
repo
if
it
is
open
source
yeah.