►
From YouTube: 2021-04-22 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).
A
A
Good,
just
waiting
for
yeah
rice.
A
B
B
B
C
Are
you
you're
in
ukraine
right.
C
C
B
B
A
B
Okay,
so
just
I
have
just
a
pair
of
topics
for
today,
one
is
reviewing
last
meeting
topics
and
another
talking
about
the
beta
release.
B
I
really
want
to
get
it
out,
so
let's
check
what
the
the
topics
for
last
meeting
to
see.
If
still
something
missing
that
pr,
I
think
it
was
already
merged
the
first
one,
the
ios
attributes.
B
So
that's
done
that
other
was
written
network
instrumentation.
I
already
created
test
for
it
and
created
appear.
I
also
modified,
I
don't
know
bryce
I
modified
the
did
you
create.
B
It
appear
and
I
closed
it
because
because
of
the
request
I
mean,
if
you
want
to
get
a
reference
to
the
real
request
that
it's
been
sent,
the
only
way
is
doing
that
after
the
spam
has
been
created,
but
I
also
have
seen
a
difficult
thing
for
changing
the
the
request
I
mean
the
span
after
the
so
I
have
faced
a
problem
that
I
would
like
to
change.
The
parent
of
a
span
of
a
network
request
yeah,
and
I
cannot,
with
the
current
situation
of
the.
C
Okay,
so
might
you
might
need
a
refactor
anyway
yeah
I
I
hold
down
hold
down
the
the
updated
network
instrumentation
and
I
was
able
to.
I
was
just
trying
to
add
some
additional,
so
I'm
messing
around
with
capturing
like
the
network
state.
You
know
wi-fi
cell
and
then
getting
like
carrier
name
and
adding
that
onto
the
network
request
as
it
gets
created,
and
so
it's
it's
working,
fine
that
way
putting
it
into
the
the
network
request
created
with
the
spam.
C
So
my
yeah,
I
didn't
realize
that
you
could
add
attributes
to
a
span
after
it
was
created.
I
thought
they
were
immutable,
so
yeah.
B
Yeah,
okay,
okay,
yeah!
I
I
would
like
to
change
the
parent
or
something
like
that
after
being
created
or
modifying
it,
but
I
just
was
fighting
with
it
before
this
meeting
but
yeah,
but
it's
like
having
two
ways
of
configuring
before
and
after
the
creation.
I
I
think
it's
too
much,
so
I
was
trying
to
find
another
way
to
do
that,
maybe
with
the
active
span
or
something
like
that
in
order
to
have
a
proper
parent
or
something
but
okay,
I
I
will
just
keep
pushing
there
so
that
the
pr
is
created.
B
I
yesterday
push
it,
but
I
still
need
some
review
from
from
sergey,
so
I
hope
he
will
do
it
soon,
because
I
also
created
another
pr
yesterday
that
he
reviewed
that
it
was.
I
did
it
very
fast
because
it
was
just
changing
the
configuration
of
the
tracer
provider,
so
you
can
change
the
clock
or
the
sampler
or
those
fee,
those
fields,
because
after
the
rework
on
the
classes
and
the
public
and
private
classes,
the
shared
state
of
the
tracer
provider
was
was
made
non-public.
B
So
it
was
not
possible
to
modify,
for
example,
the
sampler
of
the
tracer
provider
or
the
clock,
or
those
those
things
or
the
even
the
span,
processors
that
was
not
possible
before
so
so
now
I
change
that
api.
There
is
a
pr
already
there
that
should
work
with
that.
B
It
was
just
adding
an
interface
to
the
internal
shared
state,
so
it
would
could
be
usable
and
that's
from
my
side
and
waiting
for
the
network
instrumentation.
I
think
it's
so
if
it
worked
for
you.
C
B
Okay,
great
so
that's.
B
B
And
the
other
thing
is
yeah
the
type
safety
for
semantic
attributes
still
waiting
for
that.
C
B
There
is
no
changes
there,
we
will,
we
will
continue.
I
mean
it
shouldn't
break
changes
to
the
users
if
they
already
have
used
the
proper
methods
for
setting
the
attributes.
So
it's
not
so
important
but
yeah.
It
would
be
nice
not
to
have
a
breaking
change
or
that
first,
the
people
that
didn't
use
the
api
correctly,
but
maybe
we
should
yeah
about
the
treasures
porter
the
weight
I
was
going.
I
was
really
trying
to
fix
it,
but
we
we
we
we
are
facing
a
difficult
blender.
B
Is
that
it
it's
creating
it's
working
on
the
year
pc
calls.
It
already
has
a
queue,
but
the
problem
is
that
is
that
it
it
attacks
using
a
future.
There.
C
B
B
With
the
api
of
the
exporters,
we
cannot
we,
we
must
wait
there
to
have
a
result
in
the
way
we
have
the
api.
I
don't
know
if
we
should
try
to
change
our
results
to
be
a
future
like
value,
so
it
can
be
evaluated
in
the
future.
When
you
really
want
to
know
if
it
happened
or
not,
I
think
that
would
be
great,
but
then
we
have
the
problem
with
what
kind
of
future
we
want
to
evaluate?
B
C
Maybe
we
can
just
table
it
for
now,
especially
with
the
especially
with
just
recommending
using
like
the
batch
processor
instead.
So
that's
not
yeah.
C
B
Yeah,
the
only
problem
I
see
is
that
when
you
have
several
several
exporters
that
instead
of
working
in
parallel,
you
will
have
to
wait
for
each
of
them.
B
That
could
be
an
issue
if
they
are
long
to
export,
something
like
that.
That
won't
be
very,
very
efficient,
but
I
yeah
we
can
live
like
that,
like
that,
at
least
for
the
first
version
I
don't
know
yeah,
do
you
think
so.
B
Okay,
the
other
yeah.
The
other
point
is
the
prometheus
exporter.
A
Yeah,
I
looked
at
it.
You
know
for
some
time
last
week,
so
where
would
this
actually
come?
Is
it
going
to
be
because
I
was
looking
at
you
know
similar
ones,
another
language,
repos,
and
so
this
would
be
like
a
remote
right
right,
as
opposed
to
from
which
is
coming
in
and
pulling
it
from
an
endpoint.
This
is
this
sdk
sending
it
to
prometheus
yeah
right
so
so
that
would
be.
A
A
B
A
Yeah,
it
doesn't
seem
like
even
the
other
repos
seem
to
have.
I
mean
I
saw
some
issues
open,
but
I'm
kind
of
looking
for
some
example.
Guidance
to
you
know
kind
of
start
on
how
to
okay.
A
B
It
was
james
james
who
mentioned
it.
A
Yeah,
so
it
would
be,
let
me
do
some
more
research
and
I'll
get
back.
I
think.
B
A
B
Okay,
okay
and
that's
for
the
last
meeting,
things
missing
or
in
work
and
for
the
beta
release
plan.
I
I
went
to
the
maintenance
meeting
on
on
monday.
I
think
bryce
you
were
there
also.
B
So
I
will
work
on
it,
but
I
would
like
to
have
the
network
instrumentation
pushed
into
the
project.
B
And
I
think
that's
the
only
missing
point
for
the
library
and
whenever
that's
merged
I
will
I
will.
I
will
try
to
make
a
release
and
a
blog
post.
C
B
Yeah,
probably
probably,
but
really
yeah,
at
least
for
correctness
and
probably
expressiveness
also
but
yeah.
I
will
do
my
best
to
to
have
something
that
just
say
that
we
we
are
just
with
a
library
that
can
work
so
yeah.
I
have
to
review
also
the
specs
for
the.
B
Oh
will
it
was.
B
Yeah
I
yeah
I
have
to
review
also
on
the
spec.
There
is
a
there
is
a
document
that
says
I
don't
know
if
you.
B
B
I
are
not
done.
I
will
review
them
if
you
could,
please
take
also
a
review
and
see
if
we
are
covering
all
or
you
maybe
just
with
what
you
know
if
they
are
more
or
less
covered
or
not
covered.
B
Yeah
cool,
so
I
think
some
of
the
some
of
those
documents
have
that
has
been
updated,
but
I
use
that
for
reviewing
that
everything
was
done
more
or
less,
but
maybe
I
missed
some
of
them.
So
I
have
to
also
review
that
if
you
can
just
review
or
or
if
you
check
that
something
is
missing,
maybe
I
put
a
plus
and
it
shouldn't
be
there
or
something
like
that,
so
that
would
be
will
be
great
yeah.
C
I'll
take
a
look
at
it
and
add
what
I
find
just
underneath
that
review
bit.
B
Okay,
great
thanks,
so
I
can
create
a
also
appear
and
modify
that
yeah.
Some
of
those
things
have
changed.
New
items
have
appeared,
but
yeah
they
usually
change
it
more
or
less
weekly.
So
I
I
had
not
keep
up
to
all
of
them,
but
yeah
just
in
case-
and
I
don't
know
from
my
side.
C
I
had
just
like-
I
haven't
really
looked
into
this,
but
this
just
popped
into
my
head.
I'm
I'm
messing
around,
like
I
said
with
the
network
state,
and
I
was
wondering
if
it
would
be
interesting
to
have
an
event
fire
when
the
network
state
changed,
so
that,
like
active
network
requests,
might
get
flagged
with
those.
So
like,
let's
say
because
some
a
user
is
on
wi-fi
and
then
it
switches
over
to
like
3g
having
an
event
show
that
the
the
network
state
changed
to
3g
mid
network
transaction.
C
C
Yeah,
that's
what
I
was
wondering
is
their
way
to
apply
an
event
to
all
active
spans.
So
I
noticed
that
you
are
maintaining
a
list
of
active
spans
in
the
network
instrumentation.
C
So
there
is
a
we
do
have
a
a
record
of
it
and
it
could
be
as
easily
as
just
adding
allowing
a
you
know,
an
action
to
be
applied
to
all
those
spans
that
are
that
are
acting.
I.
B
So
if
you
have
a
method
that
just
checks
for
that,
you
could
access
those
spans
and
modify
their
attributes,
those
are
only
the
spans
that
are
updated
yeah.
When,
when
a
request
ends,
it
should
be
removed
from
that
list
yeah.
So
only
the
life
spans
are
there
so
yeah.
I
think
it
also
returns
a
copy
of
the
the
interface
returns,
a
copy
of
the
expanse,
so
you
can
safely
play
with
them.
B
I
think
that
that
could
work.
If
you
want
to
know
about
that,
there
yeah
yeah,
I
was
thinking
about
yeah,
okay,
yeah.
B
B
You
you
you,
okay,
there
is
a
way.
There
is
a
way.
Also
you
could
use
a
spam
processor.
You
can
create
an
spam
processor
that
I
mean
that
acts
on
all
the
spans.
On
the
end
you
can
at
the
start
at
the
end,
you
can.
You
can
have
a
spam
process
too.
B
Maybe
yeah,
if
you
mark
your
expanse
with
some
attribute
or
you,
so
you
can
just
act
on
those
you
can
you
could
at
least
yeah
modify
the
span
before
ending,
but
I
don't
know,
maybe
it
has
changed
twice
during
the
life
of
the
request
so
yeah,
I
don't
know
how
I
mean
you
can
know
that
it
changed
from
the
beginning
to
the
end,
oh
yeah,
I
think
you
you
could
just
look
for
some
of
the
network
tags
and
mark
when
they
are
created,
mark
them
with
a
without
the
same
spam
processor
I
mean
you
can
on
the
start
of
a
rig
of
any
spam
that
has
a
network
attribute.
B
C
Sorry
spam
spam
spam
yeah,
because
the
spam
has
a
ad
event
which
allows
you
to
tag
a
point
on
the
event
when
something
interesting
happens:
yeah
yeah.
So
I
was
thinking
of
having
her
applying
one
of
those
to.
C
Or
does
the
does
the
sdk
have
an
array
of
active
spans?
Are
you
referring
to
the
one
that's
stored
in
the
network?
Instrumentation,
no.
B
The
one
in
the
network
instrumentation
there
are,
there
are
no
other
arrays
of
estimates.
B
Yeah,
you
cannot
know
what
expanse
are
active
or
running
from
the
outside.
I
mean
you
just
create
the
expand
and
you
don't
have
access
to
them
until
they
end,
probably
when
they
start,
because
you
have
the
spanish
porter,
they
sorry
the
spam
processor
on
the
end,
also
the
spam
processor
and
then
directly
goes
to
the
exporter.
B
Yeah
that
array
of
the
network
is
mainly
done
if
to
know
which
network
request
didn't
finish
when,
for
example,
your
application
ends.
That
was
because
maybe
you
created
a
request
or
something,
and
you
just
want
to
know
that
that
through
request
never
returned,
or
so
you
can
monitor
that.
That
was
the
intent
of
that
having
data
reader,
that
I
think
it's
useful,
knowing
that
something
didn't
really
finished
before
you
exceeded
the
application
or
things
like
that.
B
Yeah
in
the
in
the
spec
for
the
network
expanse,
I
think
we
are
creating.
I
don't
let
me
check,
but
I
think
we
are.
B
B
B
Yeah,
I
think,
for
example,
it
has.
It
has
the.
B
But
yeah
the
thing
is
that,
as
we
talk
with
bryce,
sometimes
that
information
can
change
during
the
life
of
the
request.
You
can
start
in
a
wi-fi
and
editing
3d,
for
example,
so
yeah
and
it's
not
in
the
standard
for
the
network
instrumentation,
I
don't
know.
Maybe
we
could
just
have
a
network
status,
start
network
status
and
network
status
or
something
like
that,
but
it
will
be
another,
an
add-on
for
the
api
and
probably
can
be
done
on
the
on
on
the
client
side.
C
Yeah
I
have
I've
been
just
using
the
this
is
something
that
we're
interested
in
network
status,
but
I
just
use
the
network
configuration
to
apply
the
status
when
the
when
the
application
or
when
the
network
request
kicks
off.
C
That
in
that
callback,
so
yeah
it's
it's
it's
pretty.
It
was
a
pretty
easy
thing
to
do.
Thanks.
Thanks
to
that,
configuration
that
you
built
out.
B
Yeah
yeah
I
mean
you,
can
you
can
you
can
catch?
The
network
request
in?
I
think
all
the
use
use
useful
points.
Yeah.
I
am
doing
some.
I
am
also
just
taking
information
before
and
after
and
and
also
yeah,
all
those
I
I
I
wrote
them
because
I
I
found
them
useful
for
my
work.
I
mean
I
were
using
something
similar
inside
my
code
and
I
thought
that
having
them
as
a
callback
will
be
could
be
useful
for
anyone.
Yeah.
B
You
you
sent
something
sorry
I
I
didn't
saw
that
message
before
yeah
you
be
not.
Do
you.
A
Yeah,
I
I
was
kind
of
looking
at
you
know
either
implementation
in
other
languages-
and
this
is
seems,
like
you
know,
it's
kind
of
a
thing
that
has
been
discussed
and
not,
I
think
you
know
everyone
uses
promises
as
more
of
a
pull
model.
I
think
that's
why
it
is
oh
exporter
from
push
yeah.
B
Okay,
yeah,
I
sorry
yeah,
I
think
I
maybe
james
can
help
with
that.
Probably
will
be
some
json
format
or
something
like
that.
Yeah
yeah
yeah,
I
don't
yeah,
I
am
not
probably
metrics
are-
is
the
area
I
I
know
less
or
I
have
less
experience
with.
So
I
just
brought
the
the
current
promising
exporter
learning
a
bit
about
prometheus
on
the
while
right
in
the
code,
so
yeah,
not
an
expert
in
that
area.
Sorry,
okay,.
B
B
Yeah,
for
example,
I
have
been
trying
to,
as
I
have
said
before,
trying
to
change
the
parent
of
a
of
a
network
span
or
of
any
span,
and
I
was
just
trying
to
use
the
active
span,
but
I
don't
know
why
it
was
not
getting
the
active
span
when
the
network
request
was
created.
I
don't
know
why
I
was
just
debugging.
B
C
B
B
So
yeah,
it
was
something
really.
I
wanted
to
use
a
context,
con,
a
trace
id
and
a
spanidy
just
taking
that
text
and
making
that
the
parent
of
the
expanse
created
after
that
and
yeah
just
try
it
with
the
active
span.
But
I
don't
know
why
it
was
not
working
with
the
network
thing.
B
C
C
Yeah,
if,
if
you
keep
running
into
problems,
maybe
write
up
an
issue
and
then
we
can
I'll
take
a
look
at
it
and
see
if
we
can
come
up
with
a
solution.
C
You
know
I
specialize
in
figuring
out,
what's
going
on
behind
funky
behavior.
B
Yeah
at
trace
propagation
issue
environment
variables,
so
this
was
my
I
was
doing
it
manually,
but
not
in
the
project,
but
maybe
now
that
the
network
instrumentation
is
inside
the
open
telemetry
and
not
in
my
code.
I
cannot
do
that
anymore
easily,
but
the
idea
is
when
you
have
two
processes,
if
you
can
really
make
propagation
with
between
them,
maybe
not
so
useful
for
ios
app,
but
for
maybe
mac
apps
or
server
apps
can
be,
can
be
useful
to
just
to
trace
what
things
happen.
B
B
So
I
was
trying
to
do
this
before
I
was
doing
it
manually,
so
I
was
setting
the
parent
span
when
I
wanted,
and
now
that
I
am
using
the
open
telemetry.
I
cannot
change
the
parent
as
easily,
as
I
hoped
so
was
trying
to
find
another
way
to
do
that.
C
Is
it
because
the
telemetry,
the
framework
is
encapsulated
like
it?
Can't
it
can't
capture
things
from
from
like
the
app
space?
Is
that
why.
B
Yeah
I
mean
it's
because
yeah
it's
calling
the
spam
builder
creating
the
spam
with
the
name,
and
so
it
works
like
and
before
I
was
creating
the
spam.
But
I
had
the
spam
builder
myself.
So
I
could
change
the
parent
context
and
set
it,
and
now
with
the
network,
is
not
possible
because
it's
encapsulated
everything.
B
It
creates
the
span
with
the
current
with
the
active
span
if
there
is
any
or
with
the
yeah,
with
the
span
that
it's
there
that
that
so
now
I
have
to
figure
out
a
way
to
do
that,
or
maybe
we'll
add
another
yeah,
a
configuration
point
to
be
able
to
change
that
inside
the
network.
Instrumentation.
C
C
B
But
if
I
don't
find
any
other
way,
maybe
can
also
be
useful
for
more
people,
so
I
will
add
that,
but
they
are
first
trying
to
use
the
active
spam
thing
or
any
other
thing
that
I
can
think
of,
because
you
cannot
change
the
parent
of
a
after.
It
has
been
created
from
outside.
I
mean
you
can
only
change
that
from
the
spam
builder.
B
I
was
also
thinking
on
using
an
spam
processor
for
that
and
change
that
on
a
start
or
something
like
that,
I
yeah
and
replacing
the
context
with
another
or
something
like
that,
maybe
with
a
so
so
that's
why
I
knew
about
using
a
spam
processor
for
doing
things.
C
B
Yeah,
I
think
they
could
change,
because
you
can
have
I
mean
you
can
have
both
both
processes
running
on
the
system
and
you
can
have
a
connection
between
them,
so
you
can
do
be
doing
different
different
tasks,
so
you
can
want
one
task
being
in
a
span,
so
you
want
to
have
the
span
that
created
that
access.
B
Yeah
so
using
the
environment
variables
as
a
way
to
read
the
actual
span
that
created
that.
B
Yeah,
that's
right:
they
can
use
yes
system
pipes
or
whatever
you
want
in
your
local
computer
or
even
you
can
use.
I
mean.
B
C
B
So
this
one
doesn't
know
about
the
baggage
of
this
process
or
what
was
the
status
of
pen
telemetry
units
in
this
process?
So
the
only
so
I
I
thought
it
was
a
good
way
to
use
the
environment
variable.
So
you
can
launch
this
process
set
the
environment
variables.
This
process
do
its
work,
save
something
in
somewhere,
for
example,
and
you
create
another
task
from
here,
and
you
want
to
have
all
the
all
that
so.
A
B
You
can
trace
your
process
there.
That
was
my
idea
and
it
looks
like
there
is
a
respect
for
that.
So
nice
yeah.
I
am
doing
that
currently
with
with
custom
code,
but
I
wanted
to
add
that
to
to
the
project.
So
maybe
I
will
move
this
if
I
can
do
kind
of
do
that
correctly
with
the
network
and
see
if
we
can
yeah
having
something
better
than
a
trick
in
the
internet
for
instrumentation.