►
From YouTube: 2021-03-24 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).
C
Hey
sorry,
I
didn't
communicate
as
clearly
last
week
that
I
was
gone
for
the
whole
week.
I
didn't
really
even
think
too
much
about
it.
You
know
I
sent
the
email
to
my
to
my
company
internally
and
I
will
do
a.
D
D
C
Our
company,
our
our
fiscal
year,
ends
at
the
end
of
march
our
vacation
resets
at
the
end
of
march
instead
of
the
new
year,
so
I
was
getting
close
to
the
end
of
the
year
and
I
hadn't
taken
any
any
vacation
this
year,
like
almost
at
all,
so
I
just
took
a
week
to
do
nothing.
It
was
fantastic.
E
C
Oh,
I
thought
valentina
that
you
weren't
going
to
be
here
this
week.
Were
you
able
to
shuffle
things
around?
It
looks
like
yeah,
I'm
doing
glad
to
see
you
made
it.
C
Okay,
thanks
for
running
the
meeting
last
week
bart-
I
appreciate
it.
Let's
see,
I
guess
we'll
start
at
the
top
here.
I
didn't
really
put
these
in
any
particular
order.
Today.
Somebody
created
a
pr
on
the
api
to
add
an
esm,
build
option.
C
There's
no
functionality
changes
here,
but
it
builds.
He
added
a
second
ts,
config
file
that
builds
an
esm
version
and
adds
it
in
a
different
folder
in
the
the
build
directory.
C
This
is
better
for
bundlers,
I
guess,
and
I'm
not
particularly
familiar
with
with
how
it
affects
bundlers,
specifically
but
bart.
You
may
be
more
aware
of
that
and
nev.
You
may
be
more
aware
of
that,
but
one
thing
that
does
sort
of
concern
me
about
this
is
that
it
immediately
doubles
the
size
of
the
build
directory
for
your
average
npm
install,
because
all
of
the
javascript
files
will
now
be
built
twice
in
his
summary.
C
C
I
don't
know
if
anyone
else
on
this
call
is
more
familiar
with
these
types
of
issues
than
I
am.
That
would
like
to
speak
up.
F
Yeah,
I
guess
for
app
insights
we
deploy.
I
think
it's
either
two
or
three
versions,
and
it
is
just
yeah
the
basic
esm
and
a
couple
other
formats
we
use
roll
up
to
to
generate
them
for
us,
but
yeah
it
does
mean
your
deployment
directory
is
doubled.
F
I
know
for
dynamic
proto,
I
drop
out
every
single
type
of
module,
so
it's
like
six
different
versions.
There
is
one
format
that
actually
has
issues
if
you're
running
in
a
require.js.
So
if
require.js
is
loaded
before
your
module
but
you're
not
using
require
require.js,
the
module
actually
never
gets
executed.
So
that's
another
issue.
You
got
to
watch
out
for.
C
With
what
particular
do
you
know,
which
type
of
loader
that.
F
Yeah
umd
is
the
one
which
is
our
default
module,
but
we
have
to
actually
go
to
a
common
js
for
our
main
one
as
well
as
well
as
the
esm
files.
C
Okay
and
common
js
is
the
default
I
think
for
typescript
yeah.
It
looks
like
this,
so
one.
C
I
think
particularly
for
the
web,
it
makes
sense
he
did
it
in
the
api
first.
I
think,
because
this
is
the
only
package
that's
separate
from
all
of
the
others,
so
it's
easier
to
make
build
changes,
but
definitely
for
the
web.
You
know
anything
we
can
do
to
improve
minification
on
the
web.
I
think,
is
a
win,
but
for
the
api
also,
I
think,
because
we
are
going
to
ask
a
lot
more
modules
to
build
in
api
support.
C
Any
small
wins
we
can
get
with
the
api
will
have
an
outsized
impact
on
the
global
install
base.
I
think
that
being
said,
I
don't
really.
I
think
we
should
probably
do
it
everywhere.
I
don't
see
any
reason
to
limit
it
to
the
web
as
long
as
we
can
do
it
in
a
way
that
you
know
with
good
compatibility
in
a
way
that
makes
sense.
C
Never
you
familiar
with
with
this
option
that
he
mentions
to
do
both
using
a
single
export
and
then
actually
adding
dot,
js
files
to
our
dot
js
to
the
end
of
the
imports.
F
I
haven't
actually
read
the
the
issue.
Let
me
is
it
a
pr?
Is
it
okay,
yeah
not
off
hand,
is
the
simple
answer.
I
think
I
read
it.
Yeah.
C
Okay,
so
I
guess
for
I
would
appreciate,
for
those
that
are
more
familiar
with,
you
know,
webpack
and
roll
up
and
various
bundlers
to
take
a
look
at
this
pr.
I
feel
a
little
bit
like.
I
don't
have
enough
experience
to
effectively
evaluate
this,
so
I
would
appreciate
it
if
others
would.
C
The
second
one
that
I
wanted
to
bring
up
was
this
context
activation
for
outgoing
requests.
There's
been
some
ongoing
conversation
here,
but
I
think
this
right
here
sort
of
sums
it
up
nicely.
C
C
He
changes
it,
so
you
can
see
it
before
here
we
have,
if
you
make
an
http
request
in
the
data
callback,
you
keep
your
you
know,
original,
active
parent,
that's
the
way
it
works
now.
He
would
like
it
so
that
in
the
data
callback,
the
parent
is
actually
the
outgoing
request.
C
This
seems
wrong
to
me
and
I
actually
think
before
we
had
it
actually
this
way,
and
we
just
we
changed
it
to
be
the
current
behavior,
because
I
think
the
data
and
the
end
handlers
were
previously
getting
the
parent
from
the
from
the
outgoing
request,
but
before
I
close
this
pr,
I
just
wanted
to
or
before
you
know
before
I
comment
or
close
or
whatever
I
wanted
to
get
opinions
from
anyone
else
here
is
this:
is
this
a
change
that
other
people
would
like
to
see
happen,
or
do
you
think
I'm
thinking
about
this?
G
On
my
end,
I
think
we
discussed
this
issues
last
time
when
I
met
the
dpr
that
changed
this
behavior
and
that's
exactly
what
we
didn't
want,
so
yeah,
I'm
really
fine
to
to
close
it.
I
mean
it's
in
its
current
home
because
I
don't
see
I
I
agree
with
you
that
we
don't
want
what's
suggested.
C
Yeah
so
I'm
going
to
yeah.
F
C
C
Let's
see
the
sdk,
I
merged
a
pr
earlier
this
week
to
use
the
release
candidate
from
the
api.
I
actually
thought
we
had
already
done
that,
but
I
guess
not
so
I
I
did
that
this
week
and
I,
after
that
we
have
to
cut
our
release
so
I'll
try
to
cut
a
release.
Probably
today,
assuming
nothing
is
blocking
it
is
there
any.
I
I
looked
through
the
various
open
pr's,
but
does
anyone
know
of
any
open
prs
that
should
block
a
release
of
the
core
repo.
C
F
C
If
so,
just
comment
on
the
release,
pr
I'll
make
the
release
pr
anyways.
C
Okay,
so
this
I
copied
from
last
week
it
looked
like
a
good
list
to
me.
I
modified
it
to
show
the
current
state
of
everything
that's
done,
but
all
the
plugins
are
done
now,
which
means
we
can
begin
the
process
of
deprecating
the
plug-in
loader
system
and
deprecating
the
plug-ins
in
npm,
and
everything
like
that.
C
I
noticed
that
some
of
these
prs
are
assigned
to
like
multiple
people
for
now
mayor's,
not
even
here
anymore,
so
I
should
probably
remove
them,
but
I
figured
we
should
go
through
these
and
make
sure
we
know
who
is
working
on
what?
A
Not
yet
I
mean
I
was
waiting
for
the
final,
the
last
the
last
plugin,
so
I
can
basically
remove
it.
Okay,.
C
Removing
the
base
plug-in
and
the
plug-in
loader
actually
we'll
do
this,
we'll
yeah
decorate
a
and
I'll
create.
A
Cannot
be
instrumentation
because
the
reactor
the
way
it
is
done,
it's
basically
extends
the
base
class
from
the
react
itself,
and
if
you
want
to
use
it,
you
have
to
use
this
class.
So
there
is,
there
isn't
any
connection
to
instrumentation
and
so
on.
G
Okay,
so
should
we
close
the
issues
that,
because
there
is
an
issues
to
immigrate
to
instrumentation
in
the
country,
ripple
yeah
that
should
be
close,
yeah.
C
C
A
A
Yeah
but
it
hasn't
been
released
yet
because
otherwise
people
will
lose
the
possibility
of
creating
like
auto
instrumentation
from
the
meta
package
hcv.
You
have
to
figure
out
all
the
configurations
and
so
on.
A
So
I
would
say,
after
the
after
we
released
the
code,
repo
and
country
people
with
rc,
and
then
it
means
we
have
the
meta
package.
We
can
start
doing
this.
C
Okay,
so
let's,
let's
just
in
order,
let's
see
what
are
we
gonna
do
here?
C
C
Here
I
noticed
from
the
the
schedule
last
week
there
is
an
item
update
compatibility
matrix.
Did
anybody
open
up
vr
for
this
or
is
it
just
yeah?
I
did.
B
C
There's
one
pr
here
that
I
just
wanted
to
quickly
mention:
it's
just
a
documentation
pr,
it
changes
almost
nothing,
but
it's
been
open
for
a
few
days
and
just
needs
a
couple
of
reviews.
C
So
appreciate
reviews
on
that
and
jacob
are
you
on
the
call
today
looks
like
no
I'm
worried
about
this
jaeger
pr
becoming
stale.
There
were
comments
on
it
from
nine
days
ago,
from
both
florina
and
bart,
and
there
hasn't
been
an
update
since
then,.
C
Sounds
like
no,
so
I
I
made
a
comment
here
just
to
to
find
out
what's
going
on.
Hopefully
you
know
he's
just
on
vacation
or
something
simple
like
that.
I
just
want
to
make
sure
that
it
doesn't
fall
through
the
cracks
because
it
will
need
to
be
done.
C
That
was
all
I
had
on
the
agenda
for
today.
Does
anybody
have
anything
that
they
would
like
to
bring
up.
H
Hey
this
is
aaron.
I
just
have
one
quick
question
about
the
the
rc.
Have
we
started
the
the
technical
committee
review
and
is
there
like
a
timeline
for
the
actual
real
1.0.
C
H
I'm
not
sure,
but
in
the
pythons
that
we
did
but
go
ahead
and
and
have
somebody
look
through
our
code.
So
I'm
not
sure
if
this
one
merges
a
good
question.
H
Just
before
the
1.0
release,
so
we
did
it
like
with
our
rc,
so
I
think
it
was
carlos
went
through
and
he
looked
at
our
rc
tag
and
then
he
opened
up
issues
for
anything
that
didn't
look
quite
right
and
then
we're
gonna
have
an
rc2
got
it.
Okay,.
F
H
Okay,
yeah,
it's
still
not
merged.
If
anybody's
curious,
I'll
put
the
link
in
here.
C
Often
I
mean
I'd
like
to
if
possible,
I
don't
see
any
reason
not
to
do
this
so
I'll
reach
out
to
the
tc
and
try
to
find
someone
to
do
this.
Let
me
grab
this
link
and
throw
it
into
the.
C
C
Thanks
for
bringing
that
up
by
something
that
I
was
sort
of
aware
of,
but
haven't
really
been
following
that
closely.
H
Yeah
no
worries
glad
to
see
the
rc's.
I
All
right,
it's
right.
I
had
the
question,
yeah
go
for
it,
so
we
got
this
api
release
candidate.
Is
it
and
it's
not
compatible
with
the
version
18?
I
So
it's
at
least
for
me,
things
just
don't
work.
If
I
try
to
mix
those
two.
I
C
So
I
think
you
should
be
able
to
use
any
version
of
the
sdk
with
the
api
once
it
stabilizes.
I
need
to
look
into
why
exactly
it
is
a.
C
I
think
the
issue
stems
from
something
we've
already
talked
about,
which
is
that
the
sdk
depends
directly
on
the
api
when
it
should
actually
be
a
peer
dependency.
C
But
that's
not
a
change
that
we've
made
yet
but
yeah.
It's
something
that
we're
aware
of,
but
in
the
future
we
will
want
to.
You
know,
have
the
api
able
to
use
any
compatible
version
of
the
sdk.
C
I
Okay,
yeah,
I
was
thinking.
Maybe
it
would
be
useful
to
have
some
warning
messages
regarding
that
when
you
mixing
these
versions,
because
it's
my
understanding
is,
if
you
mixing
it
like
this
at
the
moment,
you
end
up
with
noob,
spawns
and
nook
meters,
and
things
like
that.
I
C
Actually,
do
you
mind
the
governance
committee
right
now,
as
part
of
applying
for
the
cncf
sandbox
is
looking
for
a
list
of
customers
that
are
using
it
in
production?
Do
you
is
that
public
information?
Do
you
mind
if
we
use
you
as
a
reference
or
whatever,
for
that.
I
I
Add
one
more
question
and
it's
morbid,
maybe
so
in
the
current
version
of
the
export
parameters.
I
know
it's
beta,
but
I
was
wondering
so
you
you
can
create
an
instrument
with
dots
in
it
and
gets
automatically.
It
will
replace
those
not
allowed
characters
with
underscores.
I
think,
but
if
you
use
dots
in
the
label
naming
you
pass
a
label
using
the
bind
method,
it
doesn't
do
the
same
kind
of
stripping
and
then,
if
you
use
prometheus
it
it
starts
complaining.
C
Are
documented
but
not
actually
checked
in
our
current
metrics
sdk,
I
don't
think
we're
actually
doing
any
validation
there,
but
the
exporter
prometheus
will
have
a
different
set
of
allowed
characters
than
what
we
have.
So
the
exporter
should
be
doing
its
own
validation
and
normalization.
Also,
okay,
so
I
would
say
that's
two
issues.
C
One
is
that
the
the
sdk
should
be
validating
that
we're
using
you
know
valid,
open,
telemetry
characters,
and
then
the
exporter
should
be
validating,
that
the
characters
are
allowed
by
prometheus
and
doing
whatever
transformations
need
to
be
done
to
make
that
happen.
I
Yeah,
okay
yeah
I
can.
I
can
make
a
pr
properly
for
the
promiscuous
exporter
and
I
will
at
least
I
will
raise
some
issues
and
then
I
will
see
if
I
can
pick
it
up
myself
to
make
these
changes.
But
yeah,
okay,
cool.
C
C
C
E
Else,
okay,
then,
I
will
talk
to
you
all
next
week.
G
E
C
Hey
well
just
so
you
know
this
meeting
is
recorded.
So
if,
if
there's
any,
if
your
question
is
yeah,
okay,
yeah
yeah
just
wanted.
J
To
be
here,
yeah,
no,
it's
it's
no
biggie!
So
I
I
was
just
wondering
about.
I
think
I
emailed
you
earlier
a
week
or
two
ago
about
the
membership
in
the
open,
telemetry
community,
and
I'm
just
wondering
if
you
would
be
able
to
be
one
of
my
two
sponsors
since
I
need
one
from
aws
someone
from
another
company
yeah.
Of
course
I
think
I
didn't
I
already
you.
C
Contact
or
whatever
they
call
it.
Yes,
not.
C
Yeah
no
problem,
just
when
you
create
the
issue,
obviously
mention
me
on
it
and
I
will
I'll
comment
all
right.
It
may
help
to
send
it
to
me
on
slack
also
just
to
make
you
know.
I
have
right
now
438
open,
github
notifications,
so
sometimes
they
fall
through.
C
G
C
J
All
right
well
yeah,
so
I'm
just
saying
I
can
just
send
you
on
slack,
then
yeah
channel
or
rear
economist
channel.
J
Yeah,
that
would
be
perfect,
I'll
shoot
you
a
message
when
I
make
the
issue.
Yep
sounds
good,
all
right,
yeah!
Thank
you
daniel.
I
really
appreciate
it.
Do
you
prefer.