►
From YouTube: 2022-09-26 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
A
B
B
B
So
probably,
whenever
you
have
time,
please
people
look
into
this
ads
and
books,
exporter,
post,
Plus
yeah.
We
have.
We
have
an
issue
race
for
this.
Probably
somebody
can
work
on
that.
B
A
A
B
B
Is
degree
disabled,
so
the
only
thing
is
that
this
is
something
which
is
an
optional
unless
until
we
feel
as
sick
community
that
we
really
there
are
coming
users
who
are
really
asking
for
this,
and
it's
definitely
required.
Probably
we
should
go
for
this
and
I
mean
and-
and
there
are
not
substantial
changes
required
in
here
in
SD
blueprint.
B
A
B
B
Yeah
this
is
this
is
again
more
on
the
semantic
convention,
but
changes
which
are
happening
and
I
think
Mark
I
was
going
through
one
of
the
issue
here,
I
think
there
there
is
a
plan
to
support
the
semantic
convention
ml
file
for
Matrix,
also,
okay,
so
I
think
going
forward.
They
are
going
to
support
it.
B
And
apart
from
that,
yeah
I
just
mentioned
that
I
do
plan
to
write
an
instrumentation
library
to
generate
the
CPU
members
taxes
measurements.
The
reason
why
I
was
planning
to
write
is
because
our
Benchmark
tests
for
Matrix
are
not.
We
are
not
getting
a
optimal.
We
won't
get
an
optimal
Benchmark
results
for
or
the
correct
Benchmark
results
for
releases
and
nothing
similar
thing
will
happen
for
Matrix,
also
because
GitHub
virtual
machines-
they
don't
they're,
not
reliable
enough
to
give
a
good
results
to
us.
B
Probably
we
get
allocated
different
configuration
material
machines
and
the
CI
is
running.
I
did
plan
to
I,
think
I
discussed
with
Assam,
also
and
Tom.
Also
I
did
plan
to
run
and
a
self-hosted
server
to
run
all
the
Benchmark
tests,
so
that
that
self-hosted
server
would
be
some
something
located
locally
in
one
of
probably
in
our
premises,
and
that
will
run
all
the
CI
test,
at
least
for
the
benchmarks.
B
It
will
run
the
Benchmark
CA
test
so
that
we
get
the
same
similar
results
on
the
same
similar
configuration
every
time,
but
the
only
problem
with
that
is
that
it's
it's
a
kind
of
security
loophole,
because
it's
possible
that
somebody
can
clone
the
repo
a
folk.
The
report
and
add
a
new
Theory
test
in
that
CI
test
would
again
start
running
and
new
CI
test
in
The
Benchmark,
and
it
will
start
running
in
that
cell
phones
dimension.
B
So
that's
that's.
Github
has
mentioned
that
that,
unless
until
it's
we
are
sure
what
we
are
doing,
we
should
not
be
using
the
self-motional
service,
so
I
dropped
the
plan
for
that.
So
right
now,
I'm
just
planning
to
write
an
instrumentation
diagnetic
for
this.
B
B
And
still
still
in
discussion,
yeah
nothing
more
to
do
when
that
is
a
problem
for
PRS,
probably
anything
else.
We
want
to
add
in
this,
so
you
can
quickly
go
to
the
Prius.
If
not.
A
B
A
B
B
Change
boundary
type
to
double
for
explicit
yeah.
This
was
a
bug
in
our
code,
so
basically
the
the
Bounty
types,
irrespective
of
whether
the
aggregation
type
is
for
long
or
double
the
body
types
at
all,
so
with
always
be
double
for
the
histogram
aggregation,
but
we
were
using
either
long
or
double,
based
on
the
aggregation
type.
So
I
pick
start
and
also
along
with
that
also
fixed
modified.
The
default
bucket
size
bucket
boundaries
probably
have
a
look
into
that.
Even
if
you
have
time
in
this
assumption
is
okay
guys.
B
B
B
A
B
A
B
B
I
think
that's
all
this.
This
I
need
to
fix
it,
but
this
Tom,
I
I,
think
it's
not
very
important
as
of
now,
because
no
one
is
blocked
on
this.
So
I'm
not
doing
any
change
for
this,
but
yeah
in
memory
supported
and
assigned
to
assign
it
to
myself
unless
until
somebody
want
to
demonically,
but
anyway,
we
have
to
raise
a
new
PR
for
this
and
any
issues
I
think
yeah.
This
already
you
have
created
for
Basel
I
have
to
look
into
this.
B
B
B
A
You
have
some
comments
from
the
redoc
team.
B
To
use
and
I
think
that's
totally
fine
I
think
probably
we
are.
We
should
be
good
with
them
if
all
the
sick
are
in
general,
having
their
documentation
and
open
television.io,
let's
not
Define
I
think
we
can
also
put
it
in
so
probably
let
them
raise
a
issues
and
mark
tag.
Tag
us
I
think
we
should
be
fine
with
this.
A
B
A
B
B
Yeah
there
are
some
some
more
issues
on
metrics
observable
gauge
is
not
reflected.
I
have
to
look
into
that,
and
now,
let's
assign
it
to
myself
partial
success
response.
We
have
to
work
on
this
I've
just
mentioned
it
as
I
have
prompted
along
with
specs
complex,
because
it's
something
new
inspects
right
now
how
to
work
on
documentation
for
Matrix
I
think
it
should
be.
It
should
not
be
a
bigger
task
for
the
API
documentation
just
need
to
ensure
that
we
are
creating
the
deoxy.
B
It's
it's
being
passed
through
the
oxygen
and
the
the
API
documentation
generated.
So
I
should
be
able
to
do
that
quickly,
move
more
into
the
text,
documentation
for
UPI
and
SDK.
You
need
to
write
some
getting
started
kind
of
thing
that
may
take
some
time,
but
I
think
I
can
split
it
into
two
parts:
the
API
documentation
and
more
of
a
getting
started,
and
then
I
think
at
least
we
can
close
the
API
documentation
for
this.
A
I
have
a
question
on
oxygen:
I
haven't
seen
any
CI
track
for
that.
So
do
we
have
the
oxygen
built
somewhere
and
how
do
we
know
if
it's
working
or
if
it
has
errors.
B
It's
being
sometimes
okay,
I
have
looked
at
that.
So
we
don't
have
this
part
of
yeah,
but
I
think
it's
pulled.
I
mean
it's
been.
There
is
a
trigger
from
read.talks.io,
okay
yeah.
Your
point
is
like
whether
it's
correct
correctly
generated
or
not.
We
don't
have
that
in
the
CI,
so
we
just
rely
on
the
trigger
from
Twitter
docs.io.
Wherever
we
have
a
new
release,
they
will
pick
it
pick
the
documentation
and
generate
it,
but
yeah.
That's
a
good
point,
I'm,
just
thinking
how.
B
B
A
B
Ask
to
do
it.
I
mean
I've,
seen
the
the
the
tool
service
they
use
to
generate
the
documentation
from
deoxygen
to
data
documentation
format.
So
probably
it
will
need
us
to
replicate
all
this
to
all
that
tool
set
and
generate
to
I
mean.
B
Yeah
some
basic
sanity
tests
in
generation
of
documentation.
We
can
do
that.
That's
a
good
point.
At
least
we
want.
We
won't
do
end-to-end
documentation,
the
generation
which
Rita
talks
expect,
but
at
least
from
the
oxygen.
If,
if
it's
being
generated
or
would
be,
oxygen,
is
able
to
pass
able
to
generate
it.
It
should
be
fine
from
from
our
format
to
MD
from
the
CPC
to
MDF
to
HTML
if
it
is
able
to
generate
yes.
B
Probably
we
have
to
but
I
even
that
won't
help.
Much
because
sometimes
sometimes
I
have
seen.
Xml
exporter
is
generated
correctly,
but
still
read
on
documentation
is
not
I
mean
the
rendering.
Indeed,
the
talk
is
not
proper,
so
even
I
saw
those
circles,
those
issues,
so
we
have
to
fix
our
internal
scripts
to
make
that
work.
So
basically
I
think
in
public
we
do
have
some
scripts
to
do
it.
B
Deoxify
between
transport
be
available
some
changes,
but
required.
Sometimes
in
this
script,
even
after
XML
generation
was
correct
to
ensure
that
read.ox
is
rendering
it
properly.
Okay,
but
yeah
at
least
to
some
sanity
test.
You
can
do
it.
B
A
B
A
I
can
I
can
raise
an
issue
I'm
not
sure
to
fix
it.
I
will
have
to
investigate
on
CIS.
B
B
B
B
B
B
I
mean
this
is
something
Matrix
Exemplar,
probably
you
should
be
able
to
close.
Hopefully
in
this
week
I'll
try
to
close
some
of
the
bugs
which
we
have
existing
right
now,
which
has
been
raised
by
the
customers
by
this
week.
The
segmentation
fault
I
know
there
are
three
issues
being
very
segmentation
fault,
and
this
observable
gauge
is
not
reflected.
The
values
is
not
reflected
and
I.
Think
one
more
was
there.
B
B
B
You
know
anything
else.
You
want
to
talk
platform,
I
I,
think
we
wanted
to
check
with
Mark
Mark.
We
don't
just
wanted
to
check
with
you
your
interest
going
forward
like
we
definitely
the
contribution
you
have
done
as
of
now
and
based
on
your
experience,
I
think
it
would
be
somebody
if
you
would
be
interested
to
be
as
one
of
the
approvers
for
open,
Telemetry,
C
plus
plus
just
wanted
to
check
your
interest
and
your
time
on
this.
A
Well,
I'm
interested
to
to
contribute
to
open
Telemetry
in
general
and
for
a
while,
because
I'm
I'm,
using
it
also
at
work.
So
yes,
I,
can
definitely
be
an
approval.
Yes,.
B
B
Thank
you,
I.
Don't
think
it
would
be
a
bigger
issue,
seeing
the
contribution
you
have
done
as
of
knowledge,
I,
don't
think
there
would
be
any
issue
coming
to
meet
you
as
an
approval.
So
thanks
for
agreeing
on
this
yeah
thanks
for
that.
A
Thank
you
thanks
just
so
you
know
so
I'm
currently
internally,
I'm,
writing
all
the
CI
and
the
make
files
to
include
open
Telemetry
in
a
bigger
project,
so
I'm
still
in
a
in
a
process
of
cleaning
up
a
lot
of
compiler
warnings.
Things
like
that
so
expect
that
I.
B
A
Follow
also
a
couple
of
PRS
to
do
more
cleanup
and
I'm
also
trying
to
use
open
Telemetry
in
in
cmake,
but
in
a
in
a
super
project.
So
I
don't
know
if
anyone
has
tried
that
yet
and
I
will
probably
probably
have
a
couple
of
PRS
on
the
make
file.
Also.
B
A
You
mean
Super
built
well
with
my
projects,
and
it
will
need
to
build
open
Telemetry
as
part
of
the
bigger
project.
So
I
have
to
I
have
to
figure
out
how
to
use
the
make
file
design
for
webc
make.
For
that.
B
B
B
B
Yeah
I'm,
just
hoping
in
next
couple
of
releases,
probably
will
have
a
matrix
G8,
definitely
looking
forward
for
that.
B
B
B
Thank
you,
I
think
I'm,
probably
I
also
have
a
look
into
that
and
yeah.
That's
all
in
IO
have
one.
This
is
what
Tom
basically
Tom.
This
is
for
our
internal
Matrix
exporter.
So
this
is
nothing
nothing,
something
which
is
proprietary.
We
use
Unix
domain
socket
over
Linux
to
send
the
Matrix
to
to
some
agent,
which
is
going
to
read
that
and
some
format
so
probably
I,
think
Tom.
Whenever
you
have
time
I
can
I
can
work
with
you
to
make.
B
B
Some
some
of
them
are
copied,
I
mean
and
modified
it,
because
the
previous
exporters
were
both
were
all
for
udptcp
and
Unix
domain,
but
this
is
specifically
for
Unix
domain
sockets,
so
I
just
copied
those
files
and
modified
modified
it
to
strip
it
down
a
bit
going
forward.
Probably
we
can
try
to
move
it
to
some
common
folder
here
and
try
to
use
both
of
them
from
there
yeah
yeah
but
yeah.
But
that's
that's
a
good
good
comment
and
probably
you
feel
free
to
have
look
into
that.
B
B
B
Yeah,
okay,
okay,
I!
Think
that's
something
from
the
something
to
discuss
in
my
in
the
qualitative
thanks
thanks
for
joining
I.
Think
if
not,
it's
I
think
we
have
good
to
close
the
meeting.
Let's
meet
next
week.