►
From YouTube: 2021-03-04 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
A
Hey
hello
right,
I
guess
we'll
give
you
a
few
minutes
for
our
folks
to
join.
In
the
meantime,
we
shall
please
add
your
names
to
the
candies
list
and
if
you
have
any
issues
with
pr's
on
them,
also
add
them.
A
A
D
D
D
E
All
right:
well,
the
agenda's
looking
pretty
light,
although
it
feels
like
there's
a
lot
of
stuff
going
on
but
nobody's
put
any
items
in
the
agenda.
Yet
I
heard
from
carlos
that
he
might
be
joining
us
to
talk
through
some
other
review
process.
I
know
laden's
gonna
be
out
today.
He's
got
some
other
call
that
he's
on
so
all
right
here.
Let
me
show
my
screen
and
we
can
get
started
here.
E
Yep
yep
sweet
all
right,
so
those
who
don't
know
we're
working
our
way
through
the
issues
in
rc2,
thanks
for
everybody,
who's
been
putting
npr's
and
for
everybody,
who's
been
doing
reviews.
E
So
I
think
if
we
can
get
confirmation
on
what's
left
after
we're
done
these
rc2
issues,
I
think
it
would
be
great
if
we
could
kind
of
settle
in
on
maybe
a
target
date
for
the
next
rc.
I
think
the
next
rc
we're
hoping
will
be
the
tracing
ga
release.
E
So
if
you're,
if
you're
looking
for
something
to
work
on-
and
you
have
some
bandwidth,
any
of
these
issues
are
really
good
to
jump
on.
I
just
went
through
with
laden
and
we
assigned
some
issues
to
him
and
my
myself.
I
know
that
aaron
you've
got
a
few
shoots
in
here
and
always
got
a
couple
as
well
so
yeah.
If
you're
looking
for
somebody,
we
can
help.
Please
feel
free
to
pick
some
of
these
issues
up.
E
All
right
so
talking
about
the
first
issue
here.
So
there
is
this
issue
that
I
wanted
to
kind
of
talk
through
and
maybe
oh,
I
think,
you're
on
the
call.
Maybe
you
have
some
thoughts
here
as
well,
so
this
was
part
of
the
review
that
carlos
put
in
and
oh
great
carlos,
is
on
the
call
as
well.
E
So
we
have
everybody
here
so
the
the
question
here
was
whether
we
should
move
the
use
fan
out
of
the
tracer,
because
we
have
some
of
the
other
functions
that
exist
in
other
places
like
for
set
spanning
contacts
and
gets
current
spam,
for
example,
live
in
the
trace
module,
and
I
know
you
started
looking
at
moving
use
spam
out
of
the
tracer
and
into
the
trace.
E
I
guess
the
the
question
that
I
wanted
to
discuss
here
is
whether
it
makes
sense
to
do
this
yeah.
I
guess
I'll
just
start
start
with.
E
C
As
an
external
observer,
I
could,
I
think
that
actually
something
more
important
here
is
what
what
he
found
regarding
the
fact
that
this
logic
depends
on
the
on
the
sdk
implementation,
so
I
would
say
that
that's
actually
more
important,
the
other
part
that
you
are
highlighting
there.
Yes,
probably
this
is
much
more
important
and
I
guess
fixing
that
moving
things
around
is
just
cherry
on
top.
F
F
But
I
think
that
can
be
counterintuitive
in
some
examples.
One
I
shared
above
like
if
someone
reads
this
code,
I
think
the
expectation
would
be
for
the
context
manager
to
still
certain
exception
and
set
of
status
if
something
goes
wrong.
Irrespective
of
whether
we
said
I
think,
and
then
I
went
over
to
the
specification
and
the
specs
say
that
it's
expected
not
expected,
but
it's
ins
and
someone's
libraries
and
implementations
are
allowed
to
overwrite
the
status
and
the
last
data
set.
F
F
F
That's
that's
one
existence,
so
they
can
set
it
to
false
and
we
won't
satisfy
this
one.
One
actually
real
example
of
this
is
grpc
instrumentation.
A
grpc
sets
custom
status
based
on
its
own
custom
errors,
and
in
that
case
we
pass
in
the
instrumentation.
We
pass
set
status
on
exception,
equal
false,
so
it
works
as
it.
F
So,
if
you
look
at
the
current
state
of
mpr,
I
think
that
makes
sense
to
me,
even
if
we
decided
not
move
it
out.
I
think
we
should
still
update
you
spam
to
to
the
new
behavior,
so
it
complies.
F
F
Yeah,
so
the
so
the
gist
is
that
in
the
latest
version
the
implicit
dependency
has
been
removed.
So
that's
no
longer
a
blocker.
So
whether
we
want
to
move
this
or
not,
is
a
separate
discussion.
E
Sorry,
I
just
realized
that
I've
been
talking
muted,
okay,
so
I
guess
the
the
the
dependency
on
the
sdk
has
been
resolved,
but
the
question
is
still
whether
or
not
we
shouldn't
be
moving
this
into
the
trace
at
all.
I
guess.
F
Right
yeah,
I
I
don't
have
a
strong
preference
either
way,
but
I
know
node
sdk
removed
similar
functions.
They
just
removed
them
completely,
just
delete
them
and
they're
having
some
issues
that
they've
been
asking
them
for
for
those
functions
back
and
they're,
adding
convenience
methods
to
some
other
place.
E
Yeah
I've
been
I've
been
following
some
users
on
internally
that
are
also
kind
of
struggling
to
figure
out
the
right
way
to
create
to
to
do
this
to
like
use
a
span
without
doing
all
of
the
instantiation,
and
I
guess
there
isn't
one
because
right
now,
they've
removed
those
methods,
so
yeah,
okay,
so
I
guess
I
guess
now
that
that
dependency
is
resolved.
Maybe
I
don't
know:
do
other
people
feel
strongly
about
whether
this
u-span
should
be
like
a
module
level
method
versus?
E
I
guess,
like
the
difference
between
this
and
the
start
span.
Is
that
you?
Don't
you
don't
need
a
tracer
anymore
at
this
point,
you've
already
used
your
tracer,
I
guess,
as
a
user,
it
seems
strange
to
like
call
tracer.createspan
and
then
call
trace.uspan
or
something
like
that.
It
just
seems
like
it's.
It's
not
super
ergonomic,
but
I
don't
know
I
could
be
convinced
otherwise,.
E
I
could
be
wrong,
but
doesn't
that
do
the
same
thing
as
start
and
then
use.
E
Yeah,
that's
right
so,
like
ergonomically,
could
that
be?
Okay,
like
I
know,
I
know
we're
talking
about
convenience
api
as
well.
What
seems
to
be
what
what
most
people
would
probably
want,
just
like
create,
start
and
use
at
the
same
time,
yeah
the
only
the
only
difference
is
for
people
that
do
want
to
create
a
span
and
then
use
that
later,
like.
E
I
think
that
was
a
use
case
that
was
highlighted
early
on
during
the
project,
so
so
having
having
the
two
separate
ways
of
going
about.
This
makes
sense.
It's
just
a
matter
of
like.
Does
it
make
more
sense
to
like
have
it
as
a
as
a
tracer
method
or
having
it
as
a
module
method?
E
And
I
don't
know
I
could
go
either
ways
it
since
it's
not
using
the
tracer
there's
no
real,
like
requirement
for
use,
span
to
be
a
tracer
method,
but
then
I'm
not
I'm
not
convinced
that
it
makes
things
any
easier
to
be
able
to
call
it
on
the
module.
I
guess
yeah
like
I
guess
you
could
imagine
a
use
case
where
someone
passes
a
span
around
and
they
don't
pass
the
tracer.
E
C
I
guess
that
as
a
as
a
reviewer,
I
am
fine
either
way
and
actually
what
was
found
regarding
the
sdk
implementation
part,
that's
the
most
important
thing,
whatever
you
guys
decide
after
that,
whether
it
stays
in
trace
or
not
I'm
fine
with
it.
F
Yeah
perhaps
another
way
to
look
at
it
is
we
provide
a
tracer
sort
span
and
then
to
activate
this
pen.
You
have
to
do
all
the
plumbing
around
the
context.
Api,
so
context,
start
activate,
get
active
or
context
are
set,
that's
quite
verbose.
So
we
could
look
at
this
as
just
a
convenience
method
over
the
context
methods
where
you
have
to
call
34
methods
yeah
at
the
same
time
to
activate
the
span.
E
E
Thank
you
all
right.
Does
anybody
else
have
any
other
topics.
E
Carlos,
I
know
I
I
reached
out
to
you
earlier,
but
I
just
wanted
to
bring
this
up,
because
I
know
laden
was
hoping
to
get
some
clarity
on
this
as
well
is
whether
the
the
issues
that
you've
identified
like
are
you
still
going
through
the?
Are
you
still
going
through
the
implementation,
or
is
that
is
your
review
complete
and
are
there
any
other
issues
that
you
haven't?
C
Yeah,
actually
there
was
one
last
thing
I
didn't
do
and
I
totally
forgot
about
it
now.
I
realized
that
and
it's
that
the
back
implementation,
it's
different
to
the
one
in
other
languages,
and
I
think
I
will
probably
have
to
review
that
yeah,
but
I
mean
tracing
basically
the
tracing
part,
both
api
and
sdk.
C
E
Cool,
so
I
guess
once
so
once
you've
done
the
review
of
the
baggage
and
you're
either
you
foundation,
filed
issues
or
you're
happy
with
it.
We
can.
C
C
E
All
right
cool
yeah,
so
I
guess
assuming
there
isn't
too
many
additional
issues
with
the
baggage.
E
I
guess
maybe
we
can
talk
about
this
next
week,
but
it
would
be,
it
would
be
cool
if
we
could
get
kind
of
a
rough
idea
of
when
to
release
candidate.
We
can.
We
can
create
those
and
based
on
the
issue
that
we
have
today.
I
would
say
that
you
know.
Probably
next
week
is
a
little
aggressive,
but
maybe
not
depending
on
how
much
time
people
have,
but
maybe
the
following
week.
So
I
don't
know
mid-march
or
something
okay.
A
Okay,
yes,
so
these
pr's
they
have
been,
they
have
all
the
comments
addressed
and
everything
that
was
requested.
A
So
just
please
take
a
look
and
hopefully
approve
because
I've
been
updating
the
show
once
and
over
and
over
and
over
again.
A
So
hope
it's
friendly
enough
yeah,
I
am,
I
cannot
make
this
shorter.
I
know
you,
you
want
this
message
to
be
shorter.
The
mostly.
E
A
On
this
one
yeah
that'll
be
those
are
two
pr's
right,
one
in
the
corner
and
one
in
the
control
reverb.
E
All
right,
I
can,
I
can
repeat
the
other
one.
Let
me
just.
A
E
Right,
okay,
I'll
ping
ping,
please
get
directly
in
there,
okay
and
then
the
other
one
I'll
take
a
look
at
it.
This
one!
Oh
hey!
Do
you
want
to
talk
about
it.
F
Yeah,
this
is
the
same
one
from
last
week
that
injects
the
tracing
contracts
into
log
messages.
I
have
one
approval
meet
another
one.
I
think
we
discussed
it.
F
E
E
All
right
I'll
I'll,
merge
it
after
this
call
this
one
we
already
talked
about,
but
I
guess
we
we
just
need
people
to
review
it
now.
F
E
Nice,
it
looks
like
you've
already
got
aaron
and
diego
assigned
as
reviewers.
Can
you
guys
take
a
look
at
that
pr,
oh
yeah,.
B
E
E
I
guess,
if
you,
if
aaron
and
diego,
if
you
can
take
a
look
at
the
companion
pr,
that
would
be
helpful.
All
right.
B
This
this
one
is
about
the
examples
on
the
country
documentation,
so
they
are
not
a
fully
functional
example.
So
we're
discussing
you
know
how
do
we?
How
do
we
address.
G
B
Give
anything
so
like
as
an
user.
You
know,
I
would
expect
it
to.
You
know,
be
be
fully
functional.
So
do
you
how
so?
How
do
you
want
to
do?
We
want
to
keep
the
usage
darker?
Do
you
want
to
provide
the
fully
functional
examples
in
the
core
repo
examples,
directory
and
and
just
point
point
them
to
that
link
yeah?
So,
just
just
look
at
it
and
say
what
you
think
about
it.
E
Yeah,
I
guess
the
yeah
documentation
is
always
really
difficult,
but
okay
I'll
take
a
look
at
this
see
what
we
can
do
with
it.
E
All
right
any
other
pr's
or
issues
people
want
to
talk
about.
I
thought
we
can
take
a
quick
glance
through
any
of
the
issues
here.
F
So
no
pr's
or
issues,
but
in
my
vr
I
was
getting
so
on
python.
3.5
b3
propagator
was
failing
randomly
because
of
what
looked
like
some
flaky
usage
with
context
wires.
I
don't
know
if
that's
expected
or
not,
I
re-ran
the
test
and
they
passed
and
luckily
they
always
pass.
F
E
E
I
have
not
seen
that
before,
but
is
this
mentioned
at
all
in
any
of
the
the
testing
and
tooling
discussion.
F
Sorry,
what
do
you
mean
by
this
question?
Oh.
F
I'll
try
to
find
them.
I
still
don't
know
how
to
how
find
the
all
all
runs
on
github
actions.
It
was
so
simple
on
circles
here
I'll
try
to
find
it
and
link
it
on
the
discussion.
I
guess.
G
His
voice
is
really
quiet.
It's
his
new,
his
new
microphone
man.