►
From YouTube: 2021-12-01 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
B
A
B
B
B
B
B
B
B
Yeah
thanks
thanks
yeah,
that's.
I
think
we
are
getting
a
very
good
contribution
from
news,
nothing
really
it
has
helping
community
and
we
are
short
of
more
contributors,
and
I
have
been
asking
that
in
the
maintainers
meeting
for
open
telemetry,
that
we
need
more
people,
good
c,
plus
people,
I
mean
that
that's
something
in
c
plus,
plus
that
we
have
been
facing
specifically
once
the
traces
works
we
had
a
1.2
for
traces.
I
think
we
saw
a
slight
slight.
I
think
there
is
a
dip
in
the
contribution
from
the
community.
B
A
B
Definitely
hard
to
find,
I
mean
we
have
been
I've
been,
I
mean
in
every
weekly
maintenance
meeting.
I
have
been
talking
about
this,
that
we
need
people
from
other
companies
to
contribute
in
c
plus
plus,
and
I
mean
I
think,
all
all
the
companies
who
are
working
on
open
education.
They
have
been
trying
to
find
the
good
people,
but
they
are
not
able
to
do
that
so
so,
as
of
now,
I
think
we
are
struggling
with.
B
Oh
wow,
nice
so
did
this
software
development
just
as
a
hobby
right
I
mean.
B
B
B
So
I'm
not
sure
if
this
is
something
which
this
issue
keeps
on
cropping
every
now
and
then
windows,
dll
support
or
open
telemetry.
So
I
mean
we
we
do
support
so
by
default.
I
think
our
build
pipeline.
I
just
created
an
issue
also
for
that.
There
have
been
some
discussion
happening
in
the
community
on
this
on
on
the
slack
channel.
B
So
supporting
deal
building
dll
for
windows,
so
we
don't
so
by
default.
Our
build
pipeline
will
only
generate
the
static
libraries
unless
until
for
cmake,
we
specified
build
shared
libraries
as
on
and
we
specify
link
static
as
a
false
flag
in
the
bezel.
So
if,
unless
until
somebody
manually
updates
this-
or
I
mean
somebody
manually-
update
this
and
or
pass
it
this
in
command
line,
they
so
by
default,
it's
a
static
library.
B
So
we
so
if
we
enable
shared
library,
it
will
work
on
all
the
unix
like
systems
on
like
the
shared
libraries
on
linux,
they
works
pretty
well,
there
won't
be
any
issues,
even
if
you
generate
the
shared
libraries,
but
that
does
not
work
fine
for
the
windows,
so
the
windows
will
generate
dln
files
and
I
think
dll
provides
some
restrictions
again.
I'm
not
a
windows
expert,
but
it
provides
some
restrictions
on
the
way
you
can
access
the
heap.
B
The
memory
memory
from
the
dll
to
the
application
to
the
mem.
The
global
global
memory
is
not
shared
across
within
across
the
dll,
with
the
application
which
is
linking
the
device.
So
there
are
some
issues
and
it
needs
the
code
changes
by.
We
have
to
specify
some
specifiers
dll
import,
dll
export
in
front
of
each
of
those
functions,
so
all
these
specifiers
have
to
provided
which
we
really
want
to
import
both
function
and
global
data.
So
I
don't
want
to
do
this.
B
As
of
now,
I
just
created
the
issue
just
to
get
the
traction
of
the
user.
I
mean
how
many
people
are
really
interested
to
use
hotel
c
plus
as
a
dll,
unless
until
there
is
more
traction
or
more
more
interest
in
this,
probably
we
don't
want
to
to
touch
to
make
the
changes
in
the
code
base
to
support
this.
B
So
as
of
now,
I
see
two
requests
coming
in
the
slack
channel,
but
unless
until
they
are
more
more
of
the
interest,
I
think
we
probably
created
the
issue,
but
don't
want
to
do
anything
as
of
now.
A
B
Oh
okay-
and
apart
from
that,
I
think
I
I
I
know
I
think
the
the
for
matrix.
I
assigned
one
issue
to
you,
but
I
think
I
have
been
slow
this
and
last
week
on
the
rest
of
the
work
which,
basically,
this
matrix
reader
will
need,
will
be
needing
some
some
structures
and
classes
from
the
implementation
of
views
and
aggregation,
and
that
part
is
still
not
done
so,
probably
I'll
I'll.
B
I
mean
matrix,
that's
the
problem
with
matrix
sdk
I
mean
all
those
pipeline
is
totally
coupled,
so
it's
very
difficult
to
really
segregate
and
make
it
work
in
parallel,
yeah.
So
yeah
from
that,
I
think
I
we
can.
Probably
you
can
go
to
for
your
interest.
I
think
you
can
see
you
did
have
some
pull
requests
right.
B
And
apart
from
that,
I
think
there
was
yeah
the
race
in
this
one.
I
think
this
is
a
good
catch.
A
B
B
Yeah
yeah,
I
think
I
just
had
one
comment.
Probably
I
mean
because
that
the
test
case
was.
B
B
B
Think
this
is
this
is
for
comparing
that's
a
good
point.
Even
I
was
also
looking
why
we
have
really
created
this,
but
I
think
this
looks
to
be
comparing
the
different
mutex
implementation,
the
default
mutex
implementation
and
the
one
which
we
provide
more
of
a
performance
benchmark
benchmark
among
both
of
them
with
a
similar
scenario,
test
scenarios.
B
So
this
is
how
we
do
it.
I
mean,
if
you
see
internally
how
we
are
doing
the
the
the
mutex
implementation
which
we
have
in
our
in
our
code
base.
This
is
this
is
something
which
we
do
is
internally,
so
we
will
have
a
tight
loop
and
then
we
will
have
an
outer
loop
and
then
we
yield
in
between.
So
I
think
this
is
kind
of
mocking
the
same
behavior
and
then
doing
a
benchmarking
among
both
the
both
implementations.
B
B
I
think,
apart
from
that,
this
docker
was
something
which
I
have
to
do,
and
probably
I
can
I'm
not
sure
whether
we
should
change
it.
Tom
had
a
comment,
whether
I
I
purposely
did
not
use
any
any
internal
scripts
here
just
to
keep
it
clean,
but
I
think
tom
had
a
comment
that
better
if
we
use
the
internal
the
script
which
we
provide
for
the
installation
so
that
we
don't
need
to
repeat
the
same
same
thing
again
and
again,
internally.
B
Okay,
it
I
thought
of
doing
it,
but
just
just
just
heads
up
that
this
is
currently
I
mean
yesterday.
I
tried
this
and
I
think
it
was
failing
somewhere,
probably
the
version
of
protobuf.
I
can
merge
it
if
you
really
want
to
see
it
and
fix
it
fix
it.
A
B
A
A
So
I
want
this,
I
will
make
it.
I
will
add
some
options,
so
this.
B
B
B
Oh
okay,
probably
we
may
we
may
want
to
have
this-
is
currently
using
the
go
ubuntu
latest.
I
think
we
can
have
this
this
docker
file,
for
maybe
for
all
the
ubuntu
variant.
I
mean
versions
of
variant
ubuntu
like
starting,
maybe
from
I
mean
not
not
something
urgently.
We
should
do,
but
probably
over
the
course
of
time.
I
think
once
this
is
finished
we
can.
We
can
use
the
same
same
for
all
the
versions
like
16
18
20.
B
A
B
A
A
B
B
A
B
B
B
Okay,
that's
a
good
point
if
parents!
Finally,
okay,
I
bought
it
yeah.
That's
that's
a
valid
problem
here.
So
so
I
think
when,
whenever
the
first
span
is
there,
I
think
we
should
it.
This
should
set
the
parent
span
id.
Only
if
parents
pad
id
is
a
valid,
it's
a
valid
spam
id
or
then
only
it
will
say
it
should
set.
Otherwise
it
will
make
it
as
a
nun
so
better
not
to
send
the
parent
span
id
instead
of
sending
it
null
better.
B
B
Now
we
we
do,
we
do
check
in
other
places,
while
setting
the
parent
span
id
for
zipkin.
We
are
checking
it,
but
we
don't
check
it
for
yeah.
That's
that's!
That's!
That's!
That's
the
problem
in
otl
or
so
we
do
check
in
other
places.
I
think
we
only
copy
the
parent
span
id
if
it
is
a
valid
one
like
in
zipkin.
We
are
doing
in
the
right
way,
but
we
are
not
doing
that
in
otlp
yeah.
That's
the
right.
B
Okay,
you
should
be
able
to
assign
yourself
right
if
you
find
anything
which
is
not
assigned,
does
not
work.
No,
you
want.
You
want
me
to
assign
just
let
me
know
if
you
want
something
to
be
assigned
to
you,
I
don't
want
to
assign
unnecessarily
if
you
don't
have
a
bandwidth,
but
if
you
feel
something
you
can
take
it
you
just.
Let
me
know.
B
B
I
think
that
that's
all,
I
think,
we've
already
gone
to
the
pull
requests
and
the
issues
also
anything
you
want
to
discuss
this
one.