►
From YouTube: 2023-02-07 meeting
Description
Instrumentation: Messaging
A
Foreign
good
morning,
maybe
let's
wait
one
more
minute
for
nav
and
we
can
start
after
that.
B
C
Yeah
this
will
join
shortly.
He
is
still
in
another
meeting
and
yeah,
probably
in
like
several
minutes.
Okay,.
B
D
I'll
just
say:
hi
since
I'm
joining
I
have
not
attended
one
of
these
yet
and
should
be
joining
the
clients
like
a
little
more
frequently,
hopefully
getting
some
more
momentum
back
on
on
client
and
rum
based
instrumentation
work
that
Splunk
has
done
and
yeah
just
wanted
to
say.
Hi.
D
A
What
is
your
focus
area?
Is
it
web
itself
or
Android.
A
B
Foreign,
just
so
you
know
the
this,
this
Tuesday
meeting
is
usually
not
it's
not
it.
It's
usually
like,
like
the
core
members
or
people
like
who
like
work
on,
who
attend
like
work
on
the
projects.
The
client-side
related
projects
on
a
continuous
basis,
okay
and
the
Wednesday
meeting-
is
more
like.
Usually
more
people
join
on
Wednesday,
it's
cool
different
topics:
okay,
like
if
you're
gonna
choose
one
choose
the
Wednesday
one,
yes,
yeah,
okay
sounds.
B
A
Okay,
so
I
think
the
first
item
on
my
agenda
is
Ted
said:
tigran
is,
has
a
great
join
our
you
know,
Sig
meetings,
but
the
he
has
conflicts
with
the
with
the
times
yeah.
We
were
to
reschedule
one
of
the
two
which
one
do
you
suggest,
and
you
know
what
are
some
options
that
might
work
for
you
so
that
I
can
coordinate
with
tigrant.
B
Well,
I
think
it
would
make
sense
for
him
to
attend
the
to
Wednesday
meeting
because
that's
more
coverage.
You
know
this
more
broad
topics.
E
A
Okay,
but
how
late
can
it
be
I
think
you
are
Europe
right?
How
late
would
it
work
for
you.
A
11
pm:
okay,
okay,
so
maybe
even
if
we
have
this
at
like
one
hour
from
now,
is
it
okay.
A
No,
not
right
away
I
think
it
will
take
at
least
like
one
or
two
weeks
to
you
know.
I
have
to
still
check
with
you,
then
what
works
for
him
yeah.
A
B
We
could
also,
if
it
doesn't
conflict
with
other
you
could.
We
could
move
towards
team
meeting
to
this
time
slot
on
Tuesdays.
A
We
can
meet
like
Fargo
it'll
be
good
to
meet,
but
we
don't.
We
don't
need
a
full
hour,
so
we
can
meet
like
a
20
minutes
or
so.
A
Okay,
so
that's
on
this
topic
so
Martin,
can
you
tell
us
you
know
on
on
the
status
of
Vlogs
and
events,
SDK.
B
Yeah,
unfortunately,
there
hasn't
been
much
movements,
the
the
pr
to
to
separate
the
events
API.
It's
approved,
it's
it's
passing
tests,
but
it
hasn't
been
merged.
Yet
because.
B
A
Okay,
do
you
want
to
Ping
Daniel
to
yeah.
D
A
I
also
have
one
PR,
which
is
making
changes
to
the
spec
separating
the
logs
API
from
events
API.
But,
interestingly,
your
your
JavaScript
changes
are
coming
first,
which
is
fine.
F
Because
I
I
looking
at
it,
there's
nothing
blocking
it
apart
from
just
sorry
at
a
date
like
there's
enough,
approvers,
oh
and
congratulations,
I
saw
your
you're
being
added
too
excellent.
Thank.
B
F
Yeah
because,
like
Daniel,
signed
up
Virginia
because
Mark
myself,
that's
like
way
more
than
than
needed
so.
A
F
It's
that's
been
an
ongoing
issue
forever,
so
yeah
if
anyone's
got
ideas,
I'm
I'm,
all
open
as
well
so
perfect.
B
When
we
could,
we
could,
you
could
set
up
like
a
fork
Fork
that
all
of
us
have
access
to
and
where
we
could
just
work
on
everything
you
know
like
like
get
the
SDK
get
the
exporters
get
the
instrumentation
instrumentation
prototypes
done,
and
then
we
could
just
cherry
pick
things
into
PR.
So
the
main
branch.
F
I
I
thought
that
was
one
of
the
motivations
for
setting
up
the
sandbox
was
yeah
Becky,
it
was
a
complete
combination
of
I
was
going
to
be
the
API
of
the
JS
and
the
country
repos,
not
just
Json
country,
so
that
everything
is
in
one
place
and
we
just
create
branches
and
just
go
and
back
at
it.
So
I
guess
I
still
got
that
initial
PR
I
haven't
had
a
chance
to
work
on
my
last
week,
but
I
think
we're
really
really
close
to
be
able
to
start
creating
branches
on
that
now
yeah.
F
B
There
so
the
sandbox
could
be
used
as
like
implementing
neurotypes
and
then
then
bringing
them
back
into
the
main
repo
like
for
things
like
like
this,
like
logging
or
instrumentations
yep
yeah,
okay,.
B
You
know
that's
that's
more
kind
of
just
technical
part
of
it,
but
I
think
that
there's
a
Time
the
time
part
of
it
like
when
we
just
we
just
don't
have
that
much
time.
So
maybe
you
just
need
me
to
find
more
people
willing
to
contribute
yeah.
A
Well,
on
on
the
contribution
I
I
can
spend
some
time.
One
one
good
thing
is
I
think
the
SDK
components,
all
the
SDK
code
is,
is
mostly
going
to
be
plain
copy.
Paste
of
you
know
the
the
trace
SDK
right
or
we
we
have
trace
it
together,
ready
reference,
so
at
least
the
initial
PR
should
should
be.
You
know,
quick,
and
then
we
could
do
iterations
based
on
comments.
B
A
Okay,
can
you
okay,
I,
think
this
period
is
here.
A
Yeah
there
aren't
many
comments
on
that
PR,
except
for
one
from
you
saying
it
depends
on
your
PR
first
being
there.
So
if
we
could
get
more
comments
than
either
him
or
one
of
us,
you
know,
could
yeah.
A
He
has
a
PR
from
in
the
past,
which
he
has
mentioned
in
his
last
comment,
which
was
closed.
Okay,
so
he
has
the
code.
A
A
A
Okay,
so
so
it
will
be
great
if
you
could
get
it
merged,
maybe
by
tomorrow.
A
Okay
I
think
it's
there's
nothing
pending
in
terms
of
changes
right,
it's
only
like
Daniel
matching
it
correct
doing
a
maybe
a
final
review.
Okay,
so
maybe
you
could
ping
him
and
today
and
then
follow
up
in
the
meeting
tomorrow.
Okay,.
A
G
B
A
A
Yeah
I
think
so
we
only
need
to
make
progress
on
these
two
PRS
then,
and
and
then
add
the
exporter.
So
in
terms
of
exporter,
it
will
be
in
HTTP
exporter,
with
both
Json
and
protobuf
right.
B
Yeah
I
think
eventually
we
need
both,
at
least
at
least
the
Json
start,
with.
A
A
Okay,
all
right
anything
else
on
this
topic.
A
D
We
recently
yeah,
we
recently
hit
1.0
and
are
calling
it
stable.
Of
course,
it's
not
open,
Telemetry
stable,
it's
it's
spunk,
stable
yeah,
but
that
allows
us
to
then
begin
some
more
refactoring
that
allows
us
to
to
do
changes
required
to
donate
it,
which
primarily
is
around
just
separating
out
anything.
That's
like
Splunk,
specific
and
yeah.
We've
got
very
good.
Progress
made
on
that
and
I
would
love
to
see
us
start
getting
some
code
into
an
upstream
repo
soon
and
I'll
bring
this
up
tomorrow
as
well.
A
Okay,
so
couple
things
so
are:
are
those
things
specific
to
Splunk
are?
Are
they
like?
How
many
are
there
are
they
are
the
changes
already
made
to
separate
them
out?
Oh.
D
Yeah
we've
been,
we've
been
making
changes
for
a
couple
months
and
there
are
some
stragglers
for
sure.
So
it's
still
work
in
progress.
A
I
see
I
see
and
the
other
thing
is
like
in
terms
of
logistics,
to
create
a
repo
and
I
I.
Don't
know
what
all
is
required
in
open,
Telemetry
who's,
helping
you
on
that.
Did
you
start
that
conversation
with
anyone?
No.
A
Yeah
typically
Morgan
is
the
one
that
helps
us:
okay,
correct,
Martin,
I
think
in
the
past
it
was
Morgan
that
helped
us
yeah
anyone
so.
A
Well
so
far,
the
the
reports
have
been
language
specific.
So
so,
if
you
want
the
Android
specific
changes
to
be
in
the
let's
say,
the
Java
instrumentation.
D
D
F
D
A
I
think
the
the
only
argument
I
can
think
of
is
that
there
are
several
independent
instrumentation
libraries
there,
Auto
instrumentation
libraries
and
instrumenting
Android
would
be
one
of
them.
D
Okay,
I
can
also
bring
that
up
on
our
Thursday
javasig
and
see
what
people
think
about
that.
But,
okay,
so
that's
that's
either.
A
We
need
to
agree,
I
mean
whoever
are
the
stakeholders,
you
know
they
need
to
agree
yeah
on
the
approach
and
and
then
we
could
get
things
going.
So
so
maybe
why
don't
you
check
with
the
Java
sync
on
Thursday
yeah,
so.
A
A
G
A
A
You
know
Legacy
agents,
so
if,
if
there
is
anything
that
they
want
to
act
to
you
know
this
message,
you
know
they
can
add
that
and
then
we
we
agree
and
accept
and
then
you
know,
add
the
semantic
conventions,
but
there
are
also
some
messages
that
are
going
to
be
as
events,
not
necessarily
as
a
spam
like,
for
example,
there
is
a
network
change
or.
A
Or
a
cash
report,
these
three
would
come
as
event
and
there
is
an
event
spec.
We
are
building
so
so.
G
D
A
Yeah
so,
for
example,
for
the
crash
report,
there
is
also
you
know,
having
consistency
between
IOS
and
Android.
D
G
D
This
this
sheet
is
really
helpful.
I
think
it
would
be
good
to
have
an
analysis
of
like
where,
where
our
Android
offering
is
today
on
this
I
think
it's
it's
probably
pretty
good
I
mean,
of
course,
events
being
a
notable
exception,
but
I
think
a
lot
of
the
attributes
we
have
covered,
but
it'd
be
good
to
have
like
a
full
Matrix
filled.
A
So
yeah,
so
could
you
take
a
look
at
this
and
then
you
know
make
changes
further
as
you
see,
and
then
we
could
meet
next
time.
Yeah.
B
So
just
Jason
just
a
quick,
quick
question
on
the
spans.
So
do
you
capture
things
like
crashes
and
do
you
capture
them
as
zero
duration
spans
or
how
do
you
capture
those
yes,
okay,
yep.
D
B
G
B
That
was
I
put
it
in
there.
I
actually
was
going
to
talk
to
you
nav
about
the
pr
I've
I've
looked
at
it
one,
you
know
spent
a
little
more
time
on
it.
I
still
can't
get
it
to
and
to
work
locally.
I
think
it's
for
some
reason
like
it's
trying
to
push
something
to
GitHub,
but
nonetheless
I
I,
don't
want
to
block
this
PR
anymore,
so
I.
B
F
Yeah,
in
that
case,
I'll
I
have
two
minor
changes
to
do
to
that
before
we
push
it
in,
and
that
is
effectively
to
change
the
test
flag
to
be
false
by
default.
At
the
moment,
it's
true
and
then
there's
another
setting
there
that
I
had
to
if
it's
in
test
mode
it
blocks
creating
the
pr
so
I'll
just
undo
both
of
those
I'll
push
up
an
update,
then
yeah.
If
you
want
to
sign
it
off,
I'll
then
push
it
in
and
we
can
see
what
happens
sounds.
B
F
So,
looking
at
your
3550,
it's
sitting
neck
being
queued,
so
there
must
be
a
bunch
of
other
jobs
running
but
Daniel
said
he'll
merge
that
as
soon
as
it
goes,
green,
okay,
yeah.
So
that's
for
the
separating
events
and
logs
API
awesome.