►
From YouTube: 2021-04-21 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
B
B
C
D
B
C
C
C
The
pr
is
actually,
I
think,
already
passed
and
everything,
but
I
just
don't
want
to
do
that
during
the
meeting
so
I'll
merge
it
and
release
it
after
the
meeting,
I
already
have
a
contrib
update
that
points
at
version
19,
which
is
unreleased
done
locally,
so
I'll
just
push
that
once
the
release
is
finished
and
after
that
we
should
be
able
to
release
contrib
today.
So
all
of
that
should
be
relatively
straightforward.
C
C
He
opened
a
handful
of
issues
that
I
would
like
to
be
resolved
before
the
release
candidate.
If
possible,
not
all
of
them
are
required.
Some
of
them
are
just
recommendations,
and
one
of
them
is
not
technically
required
for
spec
1.0,
but
is
spec
1.1,
but
all
of
them
are
relatively
simple,
but
I
did
want
to
just
go
through
them
one
by
one
and
talk
about
them
here
and
assign
them
to
people
so
that
we
can
make
sure
that
they
get
done
in
a
timely
manner.
C
If,
if
that
works
for
everybody,
so
I'm
going
to
start
just
from
the
top
here
and
just
go
through
them
all.
Having
the
the
context
method
on
the
span
is
apparently
confusing
to
some
people.
He
recommended
that
we
rename
it
to
spam
context.
C
Is
this
something
that
we
want
to
do?
This
is
just
a
recommendation,
but
other
clients
have
apparently
done
it.
So
do
we
do
we
like
this,
or
should
we
close
this
issue,
particularly
looking
for
part
and
balancing,
if
possible,.
B
For
me,
it's
mostly
most
likely
like,
and
we
need
to
to
have
your
question
answered
is
that
is
the
problem.
The
fact
that
the
the
method
is
not
clear
on
the
spawn
or
do
we
need
the
top
level
helper?
I
don't
really
mind
just
updating
the
name
of
the
method.
Actually
I
I
think
I
think
the
top
layer
function
is
quite
problematic
because
we
already
have
the
problem
with
all
the
width
span
and
stuff
like
this
issues,
so
so
yeah
it
depends
on
the
nature.
I
guess.
C
Okay,
so
we'll
we'll
wait
for
a
response
from
carlos,
I
think
I
agree.
I
don't
think
a
rename
would
be
a
problem,
but
I
don't
want
to
do
a
top-level
function.
So
we'll
wait
for
for
his
response.
A
I
mean
we
can
have
both
right,
because
so
many
people
are
already
using
be
a
confusion,
the
same
with
it,
with
removing
the
balance
from
creating
this
balance,
which
people
still
are
confusing
about.
C
A
F
G
F
It
comes
back
with
oh,
we
should
be
consistent
with
with
the
spec
so
that
anyone
jumping-
and
you
know
doing
the
full
stack,
isn't
getting
lost
because
of
names.
G
C
Essentially,
to
move
those
top
top-level
functions
to
be
name-spaced,
because
you
thought
they
were
confusing.
I
added
a
reply,
but
that
that
explained
why
we
did
it,
but
we'll
get
there
for
now.
We'll
just
wait
on
his
clarification
here
and
I
will
assign
it
to
myself
renaming
the
no
ops
fan
to
non-recording
span.
This
is
something
that
was
done
in
the
specification.
Apparently.
C
I
guess
because
it
doesn't
actually
it's
not
actually
a
no-op
because
it
does
propagate.
I
am
fine
with
this
change,
but
what
do
you
guys
think.
B
G
C
I
assigned
myself
the
last
one
so
who
wants
to
volunteer
for
this
one.
I'm
gonna
sign
me.
E
C
So
this
was
the
the
one
he
suggested
that
we
move
set
span
and
get
span
into
api.trace.gitzban.
An
api
trace.
Setspan,
he
says,
is
not
a
strong
feeling
and
you
we.
We
did
this
for
a
reason
before,
essentially
for
minification.
So
do
we
want
to
implement
this,
or
should
we
just
close?
It.
B
I
mean
we
are
already
at
the
debate
of
top
level
vs
namespace
function.
Actually
I
think
it
was
ted
that
suggests
that
the
the
global
one,
the
top-level
one
aren't
user-friendly
and
I
agree
with
him,
so
I
I
will
say
that
it
depends.
If
we
want
to,
I
mean
the
the
impact
of
minification
that
we
want
to
have.
Do
we
want
to
have
like
a
really
oh
sdk
to
be
really
minimizable?
B
I
don't
know
the
world
sorry
or
do
we
want
to
so
that
api
is
understandable.
I
mean
independent
of
this.
I
guess
I
guess
so
because
this
is
the
api.
Actually,
it
should.
A
C
B
Okay,
okay,
yeah.
I
see
yeah
okay
yeah.
I
think
it
makes
sense.
It's
it's
easier
to
understand
for
for
beginners.
I
think.
A
Let
me
just
understand
correctly
because
I'm
confused,
we
want
to
move
it
to
the
open
telemetry
like
to
the
main
natures
yeah,
not.
C
That,
currently
we
have
it
where
you
do.
You
know
imports
from
api
right.
You
have
set
spam.
What
he
wants
is
import,
trace,
api
and
then
do
trace
dot
set
spam.
A
C
C
Yeah,
the
top
one's
better
for
minification,
the
bottom
one
is
maybe
more
understandable
for
beginners.
I
think
you
know
there's
a
reason.
We
have
an
odd
number
of
maintainers.
We
should
just
take
a
vote
here.
I
I
would
vote
for
the
second
one
here.
I
understand
it's
a
minification
hit,
but
I
think
it's
small
and
it's
more
understandable
for
new
users.
A
D
H
F
F
C
C
A
C
A
H
C
C
B
But
about
the
the
with
method
which
actually
the
previous
execution
were
about
on
the
api,
this
with
meta,
it's
more
related
to
the
contacts
than
than
the
tracing,
so
I
don't
know
if
we,
if
it's
I
mean
if
it's
the
same,
if
we
need
to
add
it
on
the
tracer
or
on
the
source
api
or
do
we
do
it
just
enjoy
it
for
now
and
just
use
set
spawn
and
that's
it
I
mean
I'm
not
sure
we're
talking.
H
B
B
So
do
do
we
move
all
of
them
or
just
draw.
We
are
just.
We
are
just
agreeing
on
the
on
the
trace.
One.
C
C
B
I
was,
I
think,
a
week
I
think,
set
spell
and
gets
banned
is
more
understood
to
have
to
have
on
the
trace
level
yeah,
but
it's
more
complicated
for
the
span
context,
for
example,
because
it's
the
the
propagation
that
we're
talking
about
there,
and
so
I'm
not
sure
if
that
would
that
that
would
be
on
the
on
the
context
top
level
or
the
context
namespace
or
on
the
on
the
trace
one
and
same
for
packages,
for
example
yeah.
I
know
I
mean,
maybe
maybe
not
there
is
so
propagation.
B
So
do
we
have
the
set
context,
the
sentence
context
and
the
set
packages
on
the
propagation
api
or
the
context.
One.
B
But
there's
no
baggage
up
here.
I
think
it's
a
propagation
one.
C
B
Yeah,
but
it's
only
for
the
it's
actually
just
a
propagation
method,
so
I
I
think
it
makes
sense
to
have
it
on
the
propagation
space.
B
Yeah,
you
should
you
should
just
set
method
in
this
case
now,
because
I'm
not
sure
if
it's
mutable
like
there,
I
think
the
package,
the
baggage
is
invitable
like
the
like
the
context
one.
So
if
you
modify
it,
one
you'll
get
a
new
one
in
return,
I'm
not
sure
actually
but
yeah.
That
is
how
it
works.
If
you
modify
it,
you
get
a
new
one
yeah.
So
so,
if
we
modify
it,
we
need
to
to
reapply
it
like
the
context.
B
I
I
think
I
I
think
we
don't
necessarily
need
to
have
an
agreement.
I
mean
a
definitive
agreement.
I
think
we
could
still
open
api.
It's
just
small
change.
We
could
maybe
discuss
it
more
with
our
pr
open,
and
so
we
can
see
exactly
everything.
C
It's
vm
yeah,
okay,
moving
on
consider
moving
the
diag
section
to
core.
I
think
that
this
one
we
can
just
close
because
he's
suggesting
we
move
the
the
the
logging
to
the
to
the
sdk,
but
we
moved
it
into
the
api
for
a
reason
it's
needed
during
the
startup
process
yeah.
I
think
this
is
a
reasonable
explanation
and
I'm
just
gonna
close
this
issue,
you're.
C
Create
contact
key
should
use
the
string
for
debug
information
only
so
this
is
the
symbol.4
method.
You
when
you
call
it
with
the
same
string.
You
always
get
the
same
symbol
back
in
the
specification.
They
require
that
the
create
context
key,
if
you
call
it
twice
with
the
same
string,
you
should
get
different
context
keys,
but
in
our
case,
because
we
have
potentially
multiple
instances
of
the
api
in
the
node
modules
directory
nested
down
that
probably
can't
work
for
us.
C
C
I'll
comment
that
here
and
if
he
doesn't
reply
in
the
next
day
or
so
I'll,
just
close
it
so
I'll,
send
it
to
myself
for
now.
C
He
is
recommending
that
we
don't
export
these
for
the
most
part,
they're
not
used
by
anything
except,
I
think,
some
tests,
so
it
should
be
relatively
easy
to
remove
them,
and
I
think
I
agree
with
him
that
it's
confusing
to
have
both
this
was
left
over,
I
believe
from
open
census.
Originally
it
exported
all
these
constants
and
I
I
agree.
I
would
like
to
remove
them
unless
somebody
has
a
reason
that
we
shouldn't.
A
C
Yeah
we
so
we
have
the
not
the
type.
This
is
the
class
he
said
type
but
class
is
what
it
actually
is.
C
C
One
consider
using
span
context
for
lake,
so
in
our
link
type,
we
use
span
context
or
we
use
link
spam
context,
which
only
has
a
span
id
and
a
trace
id.
This
type
doesn't
exist
in
the
specification
and
he
was
suggesting
that
we
just
use
the
span
context.
C
E
G
C
That
this
definitely
makes
sense.
The
event
type
is
in
the
api,
but
the
span.ad
event
doesn't
actually
use
it.
So
if
you
look
at
the
span,
add
event.
C
Implementation,
it
actually
just
takes
a
time
attributes
and
a
start
time,
so
it
doesn't
take
that
type
at
all.
There's
actually
no
references
to
event
or
timed
event
within
the
api
at
all.
A
C
Yeah,
it
was
moved
over
when
we
created
the
api
package.
It
was
originally
part
of
the
core
package,
so
I
guess
that
was
all
part
of
the
sdk
at
the
time
and
I
think
it
was
just
moved
over
accidentally
essentially,
but
I
think
this
is
something
we
should
do.
I
think
I
heard
valentine
say
that
he
agrees
yep
yep.
C
That's
it
for
the
api.
These
next
ones
are
definitely
bigger
tasks
for
the
sdk.
C
The
jaeger
propagator
needs
to
support
baggage
valentina,
you,
okay,
if
I
assign
this
to
you
yeah.
G
C
Okay,
he
suggests
renaming
tracer
parameters
to
spam
limits.
It
was
renamed
in
the
specification.
Apparently
I
don't
see
a
reason
not
to
do
this.
C
C
E
C
I
think
your
connection
is
interrupted.
Yeah,
it's
definitely
a
weak
connection,
but
I
think
I
think
he
heard
me
so
right
now
we
have
get
active
spam
processor,
on
the
tracer
and
on
the
freezer
provider.
The
specification
only
has
it
on
the
tracer
provider
seems
like
a
straightforward
change
and
I
don't
see
a
reason
not
to
do
it.
A
C
I
think
they
both
kissing.
Let
me
take
a.
G
C
I
still
don't
see
you
on
here
I'll
figure,
that
out
later
so
this
would
be
in
tracing.
C
A
C
C
Yeah
yeah,
okay,
I
don't
remember
what
who
was
next.
Was
it.
C
Okay,
a
newly
created
resource
needs
to
have
a
service
dot
name,
so
it
should
provide
a
default
value
if
one
was
not
provided.
This
is
required
by
spec
and
I
think
we
need
to
do
it.
A
H
E
I
C
Apparently
the
other
cigs
have
also
done
this.
I
don't
see
a
reason
not
to
do
this.
B
C
C
Yeah,
so
I'll
I'll,
just
on
this,
oh,
where
did
I
go
here?
We
go
I'll.
Also
give
you
this
one.
You
can
just
do
it.
A
C
C
And
that
means
trace
context
should
exist
under
trace.
This
is,
I
think,
literally
just
moving
the
file.
It
has
nothing
to
do
with
any
of
the
api
or
how
it's
used
or
anything
like
that,
but
just
where
it
exists
seems
fine
to
me.
C
B
And
for
the
next
one,
it's
actually
required
for
the
1.2.
That
was
what
we
discussed
last
week.
When
we,
you
know
you
are
not
there
daniel.
So
it's
pretty
much
like
we
need
to
to
be
able
to
configure.
G
G
B
Sorry
from
the
other
method,
we
can
just
require
it
and
configure
it,
which
is
not
the
case
for
us,
because
we
have
like
different
packages
for
each
exporter,
and
so
I
ask
clarification
on
the
spec
that
we
don't
actually
require
them
for
the
the
actual
sdk
that
we
have,
and
actually
carlos
responded
on
the
issues
saying
that
it's
actually
not
required
on
the
sdk
itself,
like
the
sdk
that
we
have.
B
But
it
should
be
like
on
the
obviously
like
the
sdk
that
we
have
right
now,
if
it's
clear
or
not,
but
we
have
like
the
actual
tracing
sdk
and
we
have
something
called
the
node
sdk,
which
is
building
every
component
together.
C
B
Yeah
but
the
eyes
are
shows
the
issues
for
the
browser
side,
which.
B
Not
really
I
mean
I
don't
think
it
should
be
really
configurable
via
the
environment
which
doesn't
really
exist
on
the
on
the
browser,
but
one
when
I
just
check
with
other
sdk
and
especially
the
ruby
one.
Actually
just
try
to
do
the
sdk
under
this,
the
sdk,
like
not
the
sk,
that
we
have
the
actual
sdk
trusting
sdk
just
check
if
the.
B
If
the
exporter
is
present,
when
you
see
the
the
other
one
viable
and
if
it's
present
it's
configured,
but
if
it's
not
you
just
log
a
warning
saying
that
yeah,
the
the
exporter
is
not
installed,
please
install
it.
So
I
think
we
should
mimic
what
the
ruby
one
does
and
then
we've
do
not
the
sdk
package
that
we
have.
We
install
them
by
default.
C
To
me,
I
would
probably
prefer
to
just
move
all
configurations
in
the
node
sdk
package,
but
I
don't
feel
strongly
about
that.
I'm
fine
either
way.
B
Yeah
I
mean
the
the
the
the
fact
that
it's
on
the
actual
sdk
means
that
we
don't
actually
with
that
that
it
doesn't
actually
requires
people
to
to
to
have
the
sdk
you
installed.
They
can
use
the
actual
tracing
sdk
and
just
install
the
the
exporter
that
they
want
to
be
able
to
configure
it.
So
I
think
it's
just
better
for
for
for
the
majority
of
user
and
obviously
the
people
that
doesn't
need
to
go
like
deep
to
the
level
to
the
thrusting.
B
Use
the
node
sdk
package.
I
really
think
we
should
think
about
this,
naming
because
it's
really
confusing
by
the
way
the
the
wall,
like
we
have
the
sdk,
which
is
just
the
sdk
for
tracing,
but
with
more
component
than
the
actual
sdk.
B
I
think
it
should
be
on
issues
actually
before
we
got
rc,
because
it's
kinda
hard
to
understand,
which
sdk
are
we
talking
about?
Otherwise,
I
will
update
this
for
the
for
the
year.
B
E
A
Okay,
there
is
one
thing
with
having,
like
all
exporters
in
one
package
like
people
might
use
the
node
sdk
in
aws
environment,
and
you
have
limited
size
of
the
of
the
package
that
you
can
upload
and
having
group
with
all
three
exporters
for
the
collector
might
be
quite
a
lot
already.
I
mean
you
might
not
necessarily
want
to
have
like
three
exporters
being
installed.
At
the
same
time,.
B
Yeah,
but
if
if
if
he's
got
any
issues,
I
can
just
use
the
the
the
actual
sd
tracing
sdk,
which
just
we
doesn't
which
doesn't
have
the
dependencies,
and
so
the
sdk
actually
do
the
work
to
to
configure
the
exporter,
but
only
if
it's
present,
and
so
we
have
the
sdk
package
that
does
install
them.
I
don't
really
know,
but
I
mean,
since
the
sdk
is
pretty
much
a
helper,
that's
that
that
have
all
the
components
available.
We
can
just
change
it
later
on.
I
think
yeah.
C
I
so
I
think
that
node
sdk
package
that
was
originally
ted's
idea
and
I
think
it's
meant
to
be
like
batteries
included
basic
user.
It
includes
everything
we
can
maybe
just
rename
it
like
node
sdk,
all
or
something
like
that.
That
makes
it
clear
that,
like
this
installs
everything
and
then
we
can
put
right
on
the
readme,
if
you
don't
want
everything,
we
suggest
that
you
install
individual
components.
A
G
I
C
Talk
about
the
exact
name,
an
issue
or
a
pr,
but
that
was
that
was
essentially
what
I
was
saying
is
give
it
some
name
that
makes
clear
that,
like
this
is
the
everything
batteries
included
package
and
then
I
think
bart
was
saying:
that's
fine,
but
he
also
would
like
us
to
have
that
something
similar
to
what
we
already
have
now
with
that
package,
where
it's
only
the
basic
components.
I
C
That's
a
that's
a
big
project,
I
mean
if,
if
we
were
a
company,
you
know
like
dynatrace
or
lightstep
that
that
wanted
to
have.
You
know
a
support
team
and
things
like
that,
then
maybe
I
would
say
we
could
do
that
yeah.
But
as
far
as
the
open
source
project
goes,
I
think
that
is
a
much
bigger
project
than
it
sounds
like
in
initially.
C
I'm
not
saying
we'll,
never
do
it,
but
I
think
right
now
there
are
higher
priorities.
C
Okay,
is
that
so
is
that
I
think
we've
covered
basically
what
we
wanted
to
talk
about.
You're
gonna
update
your
pr
right
valentine
and
then
we
can.
Can
you
create
an
issue
to
to
discuss
yeah.
B
B
Perfect
and
just
for
the
for
the
other
issues,
I
took
the
feedback
that
we
made
that
it
was
really
complicated.
I
just
made
a
new.
I
made
a
new
pr
that
is
much
much
simpler
for
the
for
configuring,
the
collector,
which
is
the
next
topic,
so
you
just
just
preserve
you
and
that's
it.
Okay,.
J
Yeah
me
on
behalf
of
honorable,
he
was
just
looking
for
some
reviewers
on
that
to
continue
the
lambda
spec
yeah.
I
know
that.
C
I
said
that
I
would
share
what
we
had
internally
and
I
got
the
go
ahead
to
do
that
and
then
I
went
on
vacation,
so
I
have
not
forgotten
I
I
will.
You
know,
try
to
try
to
work
on
that
this
week,
if
possible.
If
nothing
else,
I
will
at
least
review
this
vr.
J
C
Cool
is
there
anything
else
that
anyone
else
wants
to
talk
about
today.
C
Okay,
then
everybody
have
a
good
good
week
and
I
will
talk
to
you
next
wednesday.