►
From YouTube: 2022-03-23 meeting
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
Yeah
hi,
so
I
want
to
discuss
about
about
the
open.
Telemetry
cpp
contribute
like
how
do
I
do
the
tagging
and
about
ci
introducing
ci
in
that
repo?
So
I
need
some
guidance
on
that.
B
A
A
Yes,
so
you
can
do
it
yeah,
it
should
be
in
the
you
can
create
it
in
the
in
your
fork
and
then
you
can
raise
a
pr
for
that.
Just
try
it
out.
That
should
work.
I
think
if
it
is
not
working,
just
probably
just
ping
me,
I
mean
we
can.
A
C
B
Okay,
so
this
gets
so
whenever
we
raise
a
pr-
and
it
runs
all
those
whatever
is
mentioned
in
the
ca
right-
that's
what
I
wanted
actually,
so
I
want.
B
A
A
B
Yeah
so
for
tagging
you
said
I
should
raise
appear
so
should
that
be
like
what
needs
to
be
which
file
needs
to
be
edited
or
something
I'm
not
clear.
Yet.
A
A
That's
the
same
way
you
should
be
doing
for
the
time
I
mean
tom.
You
have
any
experience
on
that
because
normally
I
do
it
directly
in
the
main
branch
and
I
just
push
it
in
the
main
branch.
Only.
C
B
A
E
B
Yeah,
we
need
to
do
some
kind
of
like
tagging
there,
so
that
we
can
do
some
kind
of
release
kind
of
like
yeah
the
way
they
have
done
for
http.
A
It
to
do
it
in
main
repo,
also
and
like
I
have
created
for
fluenty,
like
others,
have
created
for
httpd,
so
which
is,
which
is
the
right
way
of
right
way
to
do
it.
If
you
want
to
do
a
new
release
and
type
that
in
this
time,
but
just
try
it
on,
I
think,
with
branch
you
should
be
able
to
do
it
probably
just
read
the
documentation.
A
If
not,
then
just
let
me
know
probably,
I
can
create
a
tag
for
you,
but
you
I
mean
either
it
should
be.
It
should
be
possible
to
do
it
because
that's
the
right
way
of
doing
it,
creating
a
pr
for
that
I
mean
if,
if
that
works,
otherwise,
just
let
me
know
I'll
create
a
tab
for
you.
You
may
not
have
access,
because
only
maintainer
will
have
access
to
do
it
directly
in
the
main
branch.
So
you
can
do
it
through
a
pr.
A
E
Is
there
a
way
to
find
out
some
statistics
around
who
all
are
using
or
downloaded
a
particular
module,
or
how
many
people
have
used
this
kind
of
details?
We
wanted
to
gather
for
the
web
server
module,
which
we
have
put
it
into
the
github
right.
I
just
wanted
to
check
if
those
kind
of
provisioning
is
there.
A
A
C
E
By
repo,
can
you
just
suggest
our
demonstrate?
How
do
we
do
that,
and
only
maintainers
have
access
to
that.
A
A
E
That
will
be
maybe
more
advanced
aspiration.
Only
thing
I
wanted
to
just
see
specifically
if
for
the
web
server
thing
people
are
visiting,
they
have
very
questions
so
at
least
from
the
traffic
perspective.
We
wanted
to
gather
some
statistics,
but
at
least
when
we
were
looking
into
that
option,
we
would
not
find
anything
spacecraft
related
to
that.
A
A
E
A
E
A
B
Yeah,
so
owen
has
raised
one
more
pr
today
to
merge
it
to
async
changes,
so
that
has
some
of
his
changes
on
top
of
the
interface
change.
A
B
B
Yeah,
so
regarding
on
the
on
the
interface
size
side,
I
think
owen
has
made
some
more
changes
on
the
processor
span,
processor
and
the
log
processor,
okay
and,
and-
and
my
question
is
we
should
we
should
also
handle
like
as
discussed
in
the
discussion.
We
should
also
handle
those
callbacks
which
are
not
part,
of
course,
like
all
those
four
callbacks
which
we
have
not
handled
right
like
we
should
wait
for
those
callbacks
which
are
not,
which
didn't
come
back
right.
Okay,.
B
Yeah,
like
so
the
callbacks
which
are
which
will
be
coming
from
shut
down
or
force
plus
anyway,
we'll
be
doing
that,
but
you
also
mentioned
there
is
one
more
scenario
where
the
callbacks,
which
were
triggered
pre
shutdown
or
something,
and
that
we
have.
We
should
wait
for
that
as
well.
Right.
B
A
D
A
Yeah
I
mean
I
still
have
some
concern
whether
we
can.
We
should
have
parallel
callbacks
running
or
not
so
but
yeah,
but
definitely
irrespective
of
whether
we
have
concurrent
callbacks
ongoing.
At
the
same
time,
we
should
be
able
to
track
the
status
and
we
should
be
able
to
ensure
that
that
we
know
and
when
those
callbacks
are
going
to
end
or
when
they
are
ending.
We
should
be
able
to
do
so.
Some
status
should
be
maintained
at
the
processor
level
for
those
callbacks
which
are
running.
B
So
when
a
status
comes,
we
should
just
pray.
As
of
now
we
can
just
print
the
status
right
on
the
export
result.
B
And
the
export
result,
which
you
have
maintained
as
a
result
in
the
callback,
we
are
not
doing
anything
with
that
export
result.
As
of
now
yeah.
A
A
A
B
Yeah,
so
since
owen
owen
has
made
some
changes
on
top
of
that,
so
I
would
I
I
would
like
to
get
that
merged
so
that
I
can
make
those
remaining
changes
on
on
the
spam,
processor
and
log
process.
A
And
hey
tom,
I
think
you
had
some
concerns
right.
I
mean
what
why
we
are
doing
this.
I
just
saw
your
mess
comment
in
one
of
the
one
of
the
yeah.
B
Yes,
yes,
yes,
so
if
you
go
to
my
older
peer
like
the
one
which
was
for
the
main
branch,
I
think
that
will
have
the.
B
D
A
A
Okay,
so
this
this
basically
has
the
changes.
So
david
has
done
some
changes
in
the
exporter,
I
mean
in
the
export
function.
The
api
and
comment
is
doing
some
changes
in
concurrent
http
requests
and
also
he's
doing
some
changes
in
the
process.
The
processor
level,
I
need
to
check
more,
but
I
think
both
the
changes
as
of
now,
we
are
consolidating
on
this
branch
and
then,
once
once
it
is
ready,
then
then
a
formal
pr
would
be
raised
for
the
main
branch
which
will
be
reviewed.
A
A
A
A
The
subject
also,
I
mean
in
the
description
or
in
the
heading
in
the
title
of
this
pr,
but
it's
a
it's.
It's
a
given
feature
or
something.
B
Yeah,
so
is
there
any
like
planned
release
for
which
this
has
to
go
like
any
any
plan
like
this
async
changes
should
go
into
this
release
or
something
are
you
planning
anything.
A
E
A
A
Probably
just
a
request:
whenever
you
have
time,
please
go
through
this,
this
should
be.
This
should
be
good
to
add
it.
There
were
some
issues
which
I
fixed
yesterday,
and
I
think
I
really
I
have
put.
A
A
few
more
detailed
tests
also
to
test
it
both
cumulative
and
data,
so
it
should
be
working
fine.
There
may
be
performance
issues
as
we
could
anticipate
when
we
were
discussing
the
design
as
we
are
moving
and
tom
also
rightly
said
that
we
are
doing
few
copies
and
movement
of
data
points
from
one
one
structure
to
other
structures,
but
I
think
probably
I
mean
I'll
I'll
suggest
that
we
can
look
into
that
separately
once
we
have
an
end
to
end
working
functionality.
A
To
get
more
comments
on
things,
because
this
this
part
is
going
to
be
the
synchronous,
metric
collection
and
even
the
asynchronous
which
assan
is
working
with,
that.
That
would
have
most
of
the
logic
of
the
matrix,
how
we
are
going
to
create
the
aggregation
and
how
we
are
going
to
generate
the
merge
and
the
data
for
all
the
innovations.
A
A
A
A
All
the
constructed
explicitly,
even
though
they
are
default,
and
then
you
have,
we
have
to
do
initialization
of
all
the
variables
if
they
are,
if
that
class
is
part
of
a
vary
of
variant.
So
I
have
mentioned
that
that
remove
all
the
constructor
initializers
from
below
classes
when,
when
they
stopped
supporting
gcc
less
than
5.
A
A
Should
be
working
fine
I
have
as
of
now
it
only
is
doing
some
aggregation
or
of
the
counter
and
up
down
counter
the
other
one.
The
last
value
will
be
done
with
the
asynchronous
histogram.
I
think,
probably
you
can
take
it
separately,
I'll
raise
a
separate
fear
for
that,
so
that
it's
easy
to
review
it.
Otherwise
it
could
be.
It
will
become
a
very
big,
queer
I'll
raise
a
separate
issue
for
that
for
history,
so,
but
it
will
use
most
of
the
logic
as
which
is
there
here.
A
A
Do
you
think
we
should
we
are
ready
for
this
release?
I
don't
see
anything
new,
but
I
think
probably
you
can
take
a
call
while
you
click
on
this.
F
A
F
A
Probably
I
think
we
may
have
it's
two
weeks
we
may
have
async
functionality,
also
with
st
probably
you
can
you
can
see
if
we
can
have
that
also
complete
by
then
it
would
be
a
good
good
addition
and
tom.
You
had
some
comment.
I
remember
last
week
with
these
changes,
if
we
should
put
it
under
the
future
feature
flat
right,
yeah,
yeah,.
A
Is
it
possible,
like
the
async,
async
functionality,
functionality
which
we
are
implementing
if
we
can
define
it
under
a
feature
flag
or
basically,
we
we
can
use
a
macro
and
that
should
be
enabled
only
if
macro
is
and
if
macro
is
set,
then
only
it
should
be
enabled.
Otherwise
it
should
not
be
an
event.
Yes,
oh.
B
Okay,
so
that
should
be
under
like
this
async
functionality
should
be
under
macro.
A
B
A
A
C
C
A
And
I
think,
over
the
course
of
time
it
may
have
some.
We
may
have
some
better
understanding
how
it
looks
like
the
changes
which
are
happening.
C
A
C
D
A
A
C
A
Yeah,
but
I
think
this
average
also
will
not
will
not
have
more
accurate
result,
because
these
are
not
spikes
are
very
large
and
yeah,
I'm
not
sure
if
we
can
get
any
average.
I
mean
this
is
still
okay.
If
you
see
span
is
valid,
which
is
only
one
spike,
and
I
think,
but
probably
it's
a
low
value,
so
I
think
the
spikes
would
be
would
be
completely
less
variant.
A
A
Any
concerns
in
this
right.
We
are,
you
know
it
is
not
changing.
I
don't
see
a
j
here.
We
had
some
con
still
some
concerns
if
we
can
move
the
1
pm
meeting
to
9
30,
but
I
think
I
think
the
only
problem
here
is
that
the
recommendation
from
open
telemetry
org
in
general
is
to
have
two
different
time
times
of
the
meeting
spread
out
in
different
morning
and
afternoon
times,
so
that
it's
it's
even
it's
a
even
spread
across
all
the
time
zones
and.
A
A
So
that's
a
reason
to
have
two
different
meetings
waiting
times
to
have
try
to
cover
all
those
time
zones
and
also,
if
somebody
has
a
conflict
for
a
given
time,
they
can
join
the
government.
F
F
A
A
A
Okay,
I
forgot
that
you
have
to
join
this
meeting.
Also
yeah
yeah,
that's
a
valid
point
here.
So
let's
move
it
and
that's
totally
fine.
Let's
move
it
at
905,
assuming
that
the
primitives
meeting
is
not
going
to
extend
by
five
minutes.
Yeah.
F
A
So
that's
fine,
I
think,
probably
that's
a
valid,
and
that
would
be,
I
think,
good
for
the
I
think,
probably
and
practice.
A
Okay,
apart
from
that,
we
do
have
one
approval
event,
so
I
think
he
is
not
going
to
join
the
meeting,
but
I
think
he
has
been.
He
has
been
there
in
this
working
with
this
simple
sick
for
like
more
than
seven
or
eight
months.
I
think
he
has
done
lots
of
contributions
and
he
had
shown.
A
A
C
A
A
A
A
F
A
A
Your
links
between
spam
from
different
races,
I
think
this
we
already
support
it.
I'm
not
sure
why
he
has
raised
this
issue
saying
that
we
cannot
link
the
spans
from
different
races,
the
unrelated
spanx,
but
we
do
support
it.
So
in
this
we
do
have
so
you
can
create
a
span
when
you
can
specify
the
spam
context.
Key
value
table
and
it
could
be
any
spam
context-
can
have
any
trace
id
and
spin
id.
So
I'm
not
sure
why
probably
has
not
seen
that.
A
A
F
A
B
A
A
A
A
Yes,
this
is
our
ibm
and
wants
to
enable,
so
he
wanted
to
add
the
github
ci.
You
have
to
enable
it
in
ca,
which
we
can't
do
it
as
of
now,
because
we
have
to
use
some
emulations
or
some
virtualization
to
do
that,
and
it
will
be
very
slow
and
travis
ci
supports
it,
but
we
are
on
the
the
cncf
is
using
data
predictions
in
zero
pipeline.
A
A
A
Yeah,
so
this
was
the
issue
was
that
if
the
half
cpp
yesterday,
if
you
are
using
the
https
and
you
are
not
on
c
plus
17
or
20,
it
gives
a
very
weird
errors.
It
should
fail
saying
that
the
variant
variant
is
not
present.
I
mean
the
variant
is
not
present
when
you
are
trying
to
use
the
studio,
but
it
gives
some
weird
errors
and
then
it
fails.
So
it
should
fail
early
and
when
I
think
it's
somebody
want
to
pick
it,
I
think
we'll
just
kept
it
open.
A
A
Yeah,
that's
all
this
one
is,
I
remember
it's.
We
have
discussed
quite
a
lot
and
it's
okay
yeah
this
this.
This.
A
Yeah
again,
this
is
this
is
some
is
trying
to
do
a
cross
compilation
and
we
we
don't
have
I
mean
we
could
not
test
it,
and
I
think
it's
not
something
very
highly
prioritized
because
of
its
it's
using
some
particular
compiler,
which
we
don't
have
any.
We
don't
have
readily
available.
So
in
fact
this
guy
was
not
very
happy.
I
gave
up
the
library
because
my
first
attempt
was
using
it
properly
clash.
I
don't
care
anymore,
so
I
just
mentioned
that
it's
a
specific
tools
chain
and
not
widely
reported.
A
Yeah,
that's
all
we
have
and
nothing
in
the
pull
request.
So
we
just
had
three
minutes
and
in
contact
also,
this.
A
A
So
the
the
open
elementary
build
works
fine
with
alpine
3.14,
but
it
is
failing
with
3.15,
so
the
kubernetes
it's
being
used
in
kubernetes,
then
it's
they're
not
able
to
they
want
to
use
the
latest
latest
alpine
and
they
are
not
able
to
use
because
of
that
something
with
the
grpc
compilation.
The
grpc
not
jrpc
built,
but
vr
pc
is
giving
error.
A
For
some
missing
library
yeah,
I
thought
of
pulling
it
this
issue
in
our
main
report,
because
it's
not
really
related
with
nginx,
but
it's
actually
related
with
the
it's.
The
issue
is
with
the
open,
telemetry
coder
too,
but
it's
fine
just
just
just
wanted
to
call
out.
If
somebody
want
to
see
what
exactly
it
is.