►
From YouTube: 2022-04-19 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
Okay,
share
screen
number
two:
can
you
acknowledge
you
are
seeing
the
right
meeting
notes:
yep?
Okay,
thank
you.
Let's
give
it
another
minute
or
two
for
others
to
join.
A
I
think
we
can
start
now.
Allen
had
some
agenda
items
so
he
might
be
joining,
but
let's
quickly
see
if
anyone
has
any
open
questions
to
be
discussed
so
go
through
that
and
wait
for
a
line
to
come.
The
next
two
minutes
just
to
discuss
the
naming
plan,
any
questions
or
issues
to
be
discussed
before
we
go
to
the
next
topic.
A
Okay,
I
do
have
like
some
topics,
but
I
haven't
done
the
homework
on
like
formulating
it,
maybe
I'll
sing
with
folks
to
like
offline
and
come
back
and
comment
in
the
vrs
itself.
A
Okay,
so
going
back
to
the
update
yeah,
we
did
it
actually
release
it.
I
I'm
just
checking
yeah
okay.
We
did
release
it
last
friday
and
I'm
about
to
do
the
1.3
series
of
release.
In
fact,
I
called
it
like
not
alpha
and
there
is
a
pr
where
the
namings
are
being
discussed,
so
it
would
be
here.
A
It
looks
like
we
have
some
agreement
so
I'll
be
proceeding
with
this
later
today.
Just
to
do
the
release
so
we'll
start
with
1.3
series
for
the
core
packages
for
in-memory
and
exporter.
They'll,
naturally
follow
the
same
as
the
core
package,
so
they'll
also
be
doing
1.3
beta
one
and
we
obviously
skipped
prometheus
and
in
memory
from
one
point
two
stable,
so
we're
still
like
in
the
place
where
we
can
still
afford
to
break
things.
A
A
So
it
should
be
like
always
the
same
questioning,
but
there
is
also
issue
open
about,
should
it
be
all
same
or
different,
we'll
discuss
it
like
when
we
actually
reach
the
point
where
we
can
do
the
stable
thing.
A
Next
thing
is
I
mean,
since
we
just
covered
this
one.
I
wanted
to
quickly
ask
or
quickly
share
the
milestones
which
I
have
updated
on
friday,
so
just
to
make
sure
like
everyone
can
see
it,
so
we
just
released
1.2
so
that
milestone
is
closed
and
we
need
to
do
1.3.
A
I
mean
not
not
like
we
have
to
do,
but
it
would
be
better
if
you
do
it
really
soon,
because
we
did
not
release
the
in
memory
and
we
also
don't
have
a
stable
version
of
from
this
exporter.
So
this
is
what
we
intend
to
do
for
1.3,
like
removing
461
support.
A
A
The
second
thing
is,
we
already
agreed
to
split
the
package
into
multiple,
and
we
already
have
issues
open
for
that.
I
think
eunteen
is
assigned
that
issue
about
like
splitting
the
prometheus
exporter
into
multiple
one
for
sp
net
core
and
listener
http
listener.
A
So
if
the
stack
is
not
stable,
we'll
repeat
what
we
did
with
1.2,
it
will
do
everything,
but
except
we
will
not
release
permit
this
exporter
in
the
actual
1.3,
so
it
will
be
like
1.3
rc,
then
we'll
bump
it
to
1.4
and
so
on,
but
there
is
a
good
chance
that
spec
would
be
stable
soon,
because
there
are
a
lot
of
activity
in
the
spec
world
about
stabilizing
prometheus.
A
So
since
the
sdk
is
already
mark
stable,
I
think
we
have
a
good
conference
that
this
will
be
stable
in
the
next
few
days
or
a
few
weeks.
So
I
put
like
may
end
as
the
my
best
estimate.
Of
course
it
could
change
and
there
are
like
few
other
things
which
is
about
logging,
so
we
have
corresponding
issues
or
peers
about
these
things,
so
that
would
be
the
immediate
like
next
release
of
1.31.
A
There
are
no
questions.
The
next
one
is
the
1.4.
So
again,
this
is
a
rough
estimate
because
we
want
to
do
a
release
when
dot
net
7
comes
and
that's
when
the
up
down
counters
would
be
available.
A
But
that
doesn't
mean
that
we'll
be
only
touching
it
with
the
intention
I
mean
we
can
only
release
it
on
number
end,
when
dot
net
servant
comes
as
they
stable,
but
we'll
quite
likely
have
a
release
candidate
much
earlier.
A
So
we
need
to
like
decide
on
how
the
how
exactly
we
orchestrate
the
plan,
because
we
don't
want
to
update
the
entire
report
to
use
the
diagnostic
source
version
7.
So
you
we
may
like
do
it
in
a
separate
metrics
brain,
so
we'll
figure
out
the
logistics
and
update
in
the
next
sig
or
updating
the
issue
itself,
how
we
are
going
about
updating
to
dot
net
server?
A
The
reason
is,
we
don't
mean
if
we
start
bombing
a
diagnostic
source,
let's
say
at
may
31st,
then
we
will
not
be
able
to
do
any
stable
until
number
10
from
that
main
branch.
So
if
we
need
to
do
any
anything
stable,
we
have
to
like
do
the
go
back
in
time
to
1.3,
and
do
it
maybe
like
little
bit
more
involved.
A
So
again,
I'm
not
sure
like
whether
there
will
be
a
need
to
do
a
release
before
number,
because
the
critical
features
like
histogram
experimental,
histogram,
more
advanced
features
from
views
like
baggage
examples.
We
don't
really
know
how
much
time
we'll
have
to
work
on
that
so,
depending
on
I'll
also
sing
with
other
maintainers
to
see
who
has
time
to
like
address
those
features
and
based
on
that
will
give
you
a
better
update
in
the
next
few
days.
A
I
think
allen
did
mention
that
you
want
to
touch
upon
the
exponential
histogram
side
and
he
has
lower
priority
for
the
assemblers
from
microsoft.
Also,
that
would
be
likely
the
same,
because
there
is
no
microsoft
backend
for
examples,
yet
so
quite
likely,
we'll
like
stick
with
this
plan,
like
do
1.3.
A
That
would
that
way
we
can
try
to
stabilize
the
prometheus
and
after
that
we
can
see
whether
to
take
exponential,
histograms
or
examples
or,
like
the
remaining
view,
features
or
depending
on
the
feedbacks
from
people
like
whether
they
want
to
see
it
earlier
or
not,
and
also
depending
on
our
actual
availabilities,
and
I
also
updated
these
two
milestones.
This
is
hosting
package.
This
is
instrumentation,
so
a
lot
of
people
are
asking.
A
A
Any
any
questions
on
this
one.
Otherwise,
we'll
go
to
the
remaining
topics.
C
I,
like
it,
I'm
just
having
a
conversation
today
with
my
team
about
what
status
as
open
telemetry
is
for
net
we're
trying
to
figure
out
a
suitable
implementation
for
our
communities
cluster,
so
this
is
starting
to
look
more
real
now.
A
Yeah
think
like
we
are
like
pretty
solid
on
the
traces
and
metrics
load.
Part
is
like
part
of
the
sdk,
so
it's
like
really
stable
from
like
first
version,
except
that
there
is
no
otlp
exporter,
so
I'll,
be
like
spending
some
more
time
on
like
tuning
this
further,
and
that
mike
branch
has
already
opened
an
issue
with
some
suggestion,
because
we
have
a
fundamental
issue
in
the
way
we
are
logging.
We
we're
not
we're
like
buffering
where
we
are
not
supposed
to
buffer
and
vice
versa.
A
So
we
have
like
some
fundamental
issues
and
you
did
have
some
pr
to
improve
the
credentials
leaking
accidentally,
do
logging
and
protecting
that.
So
those
are
the
things
which
we'll
be
improving
and
hopefully
like
by
end
of
this
one.
We
would
have
a
good
story
for
all
three
signals
like
including
lobes,
except
that
the
log
exporter
itself
would
be
marked
an
experimental
thing
or
really
scandal,
but
it
should
be
like
very
close
to
or
what
people
can
use
in
production,
yeah.
A
D
Oh,
maybe
you're,
referring
to
the
the
thing
I
slapped
you
about
earlier
this
morning,
yeah.
A
The
something
about
the
I
didn't
read
it
because
I
was
on
a
different
meeting
earlier
yeah.
D
I
meant
I
meant
to
try
to
get
the
the
draft
pr
up
before
this
meeting,
but
I
haven't
been
able
to
get
to
it
but
soon.
D
Pr
that
I'm
gonna
submit
that
limits
attribute
count
and
attribute
value
length
on
specifically
just
spans,
but
I
think
it
may
apply
to
log
data
as
well.
So
the
specification
has
some
guidance
on
attribute
limits
and
we
don't
implement
any
of
that
today.
Of
course,
it
should
be
implemented
in
the
api,
which
would
require
a
change.
You
know
that
runtime
kind
of
change.
D
Most
of
the
most
of
the
work
can
be
done
here
within
this
class.
So
it's
basically
the
work
is
done
here
and
then
then,
I'm
exposing
some
configuration.
A
So
it's
not
just
about
the
like
value.
Well,
it's
also
about
like
the
number
of
attributes
right.
It's
like
bit
more
involved
than
just
the
value
being
correct
or
not
it's
about
like
the
number
of
attributes,
and
if
you
exceed
that
size,
then
we
start
dropping
like
it's
a
bit
bit
more
involved
than
just
violating
the
value
right
right
right.
D
Okay,
so
yeah
going
going
through
and
doing
this
with
the
shim
was
was
useful.
I
think
the
specification
you
know
technically
it
says
that
it's
these
limits
should
be
configured
at
the
level
of
the
tracer
provider.
So
when
you're
configuring,
your
tracer
provider,
you
declare
your
your
limits
so.
D
A
B
To
quickly
ask
did
we
run
this
by
the
runtime
team
to
see
if
it
would
make
sense
to
support
this
in
the
runtime?
No.
D
A
A
I
don't
know
whether
dotnet
seven,
we
have
any
hopes
it's
april
only
so
maybe
there
is,
but
unless
it's
like
a
very
important
thing,
it's
very
difficult
to
get
dotnet
to
change
their
schedule
because
they
already
have
like
everything
lined
up
for
dotnet
seven,
so
it's
a
good
chance
that
it
won't
make
it
2.7.
So
if
that's
the
case,
our
nearest
next
release
is
the
which
should
be
2023
number
right,
yeah
so
like
like.
We
can
try
out
like
to
see.
A
If
there
is
like
enough,
I
mean
whether
we
can
convince
the
dot-net
team
to
incorporate
that
into
dot-net
seven
itself.
It's
a
spec
like
stable
racket.
It
is
stable
right.
D
D
This
is
unlimited,
there's
a
more
general
if
you
go
back.
D
D
A
A
Like
we
can
say,
like
activity,
dot
set
even
count
is
maximum
one
to
date
and
they
can
do
some
checks
to
validate
yeah.
I
mean,
let's
not
worry,
about
the
actual
implementation.
Let's
create
an
issue
link
to
the
reload
section
of
the
spec
and
meanwhile
we
can
unblock
users
by
doing
that
in
the
shim,
but
it
won't
work
for
the
logging
part
right
for
logging
like
we
need
to
do
it
in
the
like
logging
sdk
itself,
like
somewhere
as
an
option
to
the
lower
options.
A
So,
even
if
you
use
a
call
like
I
logger
dot
log
with
a
like
thousand
attributes,
the
sdk
can
choose
to
take
the
first
thousand
or
whatever
is
the
limit.
Okay,
okay,
at
least
let's
start
with
the
traces
one,
because
logs
might
be
a
little
bit
more
tricky
because
we
might
have
a
bigger
refactor
coming
here.
A
A
A
D
Go
ahead,
I
have
one
other
thing
to
discuss:
it's
yeah,
I'm
just
going
to
put
a
link.
It's
a
slack
link.
I
don't
know
if
you
have
slack
open,
but
I'm
noticing
now
there's
been
a
few
conversations
where
people
are
not
in
the
know.
That.Net
supports
logs
and
metrics,
and
I'm
wondering
if
there's
anything
that
we
could
be
doing
yeah.
I
think
it's
that.
A
D
I
mean,
if
complete
means,
you
know,
still
a
release,
but
I
think
I
saw
it
here
and
then
I
think.
A
D
D
Yeah,
I
think
yeah.
That
was
my
impression
too.
So
first
first,
the
two
two
comments.
So
the
spec
compliance
matrix,
you
have
to
scroll
the
metrics
chart
to
c
dot
the
dot
net
column.
A
D
D
I'm
inclined
to
submit
pr
and
move.net
ahead
of
all
the
blank
languages
yeah.
There
is
no
that
link.
A
A
Make
a
case
to
do
it
alphabetically,
in
that
case,
dot
net
would
definitely
come
because
of
the
dot,
even
not
like,
we'll
still
be
like
way
ahead
of
ruby
and
by
then
that
would
impact
php,
so
that
would
also
put
us
like
somewhere
in
the
middle.
So
that
way,
it's
still
neutral
alphabetically
is
considered
neutral.
D
C
Yeah
for
me,
looking
at
it
it's
already
kind
of
hard,
I
kind
of
yeah
I'm
working
with
some
sres
on
my
team
to
try
to
figure
out
how
we're
moving
forward
with
vlogging,
and
I
was
like
yeah-
I
knew
open
tomotry
has
it.
I
wasn't
really
sure
what
the
state
of
it
was.
That's
actually
going
to
be
one
of
my
questions.
A
Yeah,
so
logging
is
a
little
bit
tricky
to
explain.
Let's
see
what
the
spec
complaints
metrics
talks
about,
so
we
updated
like
this
last
week.
Only
so
we
technically
had
logging
sdk
from
the
first
release,
because
we
had
incorporated
the
open,
telemetry
logging
provider
for
I
logger
into
the
sdk
itself,
but
that
doesn't
mean,
like
logging
is
stable,
because
no
one
can
export
it
because
there
was
no
exporters.
A
There
was
no
dlp
exporter,
so
it's
I
mean.
If
you
write
your
own
exporter,
then
yes,
you
can
use.
I
logger
to
log
and
use
open
elementary
to
capture
it
decorate
it
with
trace
id
span,
id
trace
flags
and
export
it
to
your
own
place.
So
that
worked
and
it
was
like
really
easy,
because
we
all
these
things,
are
like
generics
and
dot
net.
So
if
it
worked
for
activity,
then
it
worked
for
long.
A
So
it
was
like
a
fairly
easy
thing,
but,
like
that's,
why
I
don't
know
like
what
would
I
say
about
the
status
of
logan
like
it's
definitely
not
alpha,
but
it's
not
correct
to
say
it
as
stable
either,
because
without
any
exporter
I
don't
know
what
would
be
the
like
use
case.
Unless
you
are
writing
your
own
export,
which
is.
A
I
don't
know
of
many
people
who
have
returned
their
own
exporters
within
my
company,
like
microsoft,
has
exporters
for
all,
but
that's
just
microsoft,
so
I
learned
like
maybe
another
thing
which
we
could
potentially
do
is
like
in
this
repo
like
somewhere
in
here.
We
could
include
a
table
which
indicates
the
status
some
languages
have
done
that
maybe
the
collector,
I
think,
collector
or
some
place.
I
have
seen
like
very
easy
like
when
you
open
the
the
repo
the
root
page.
A
Should
be
good,
but
I
mean
yeah,
I
mean
we
could
like
it's
right
along
with
here.
We
because
we
previously
had
a
hint
to
the
matrix
saying
that
it
was
experimental
which
was
removed
recently,
but
maybe
it's
good
to
like
document
right
here,
the
status
of
each
one
logs
requiring
a
special
asterisk
saying
that
the
logo
provider
is
stable,
but
the
otp
exporter
is
not.
A
I
don't
know
if
there
is
any-
and
there
is
this.
This
was
an
issue
when
I
talked
to
someone
else
also
for
the
sp
net
core
part.
So
here
it's
not
clear
that
it's
not
stable.
I
mean
it's
not
clear
that
this
never
published
anything
stable.
It
just
points
to
the
latest
one.
So
maybe
we
can
put
a
big
banner
saying
that
this
is
based
on
experimental
and
as
such,
it
can
change
and
also
say
that
it
is
more
stable
for
the
tracing
side
of
things
than
the
metric
side.
C
A
A
D
Yeah,
I
think,
that's
a
good
idea.
I
might
actually
open
up
an
issue
for
the
specification
repo
too,
because
I
mean
like
it
or
not.
I
think
I
think
end
users
do
end
up
there
sometimes
and
it
might
be
nice
to
have
a
condensed
view.
You
know
the
spec
compliance
matrix
is
just
it's
noisy
and
if,
if
there
was
just
a
really
condensed
view
that
hit
the
top
level
features
for
each
language
in
a
really
simple
table,.
A
But
somewhere
here
there
was
a
table
with
languages,
but
now
I
don't
see
it
anymore.
Maybe
they
changed
it.
D
So
this
this
site
was
actually
my
next
point,
which
is
that
if
you
go
to
the
net
docs
here,
unless
I
missed
it,
I
think
it
only
covers
traces
at
this
point
in
time,
which
may
also
be
part.
E
D
A
But
one
thing
is
I:
I
am
well
aware
of
the
fact
that
this
I
mean
not
just
the
folks
who
maintain
this
page.
They
are
aware
of
the
fact
that
it's
not
always
update,
but
I
don't
know
specifically
what
action
they
are
doing
to
make
it
ready
so
for
the
dotnet.
I
think
most
of
this
documentation
were
recently
submitted
by,
I
think
philippe
carter
from
honeycomb.
A
He
did
like
all
those
these
things
like
it
was
not
even
usable,
but
unfortunately
it's
not
kept
up
to
date
like
there
isn't,
I
mean
he's
just
one
person
like
no
one
is
actively
updating
it
compared
to
the
one
from
our
repo,
the
actual
item,
repo,
where
we
host
the
source
code.
So
I
did
race
I
asked
him
like
who
is
updating
it?
He
said
like
he
do
it
something,
but
there
is
no
well-defined
process.
D
D
Too,
oh
you
mean
like
examples
yeah
on
this
website.
Yeah,
if
you
go
to
getting
started,
I
think
is
where
I
was
at.
No,
no,
no
sorry
I
mean
I
mean
getting
started
yeah
from
this.
C
D
D
I
guess
it's
not
this
either
manual.
E
D
D
A
Like
who
can
keep
this
up
to
date,
because
the
general
guideline
is
to
use
this
as
the
source
of
truth
once
it
is
ready,
but
as
of
today,
if
someone
asks
me
like
where
should
I
look?
I
simply
ask
them.
Look
at
the
report
cause.
That's
the
thing
which
at
least
the
maintainers
of
the
repo
are
updating
constantly,
so
we
always
ensure
that
these
examples
run
because
it's
part
of
the
ca.
So
if
you
break
something
and
then
ta
will
catch
it,
but
that
guarantee
is
not
there,
like.
A
A
C
A
One
thing
which
we
did
attempt
was
from
the
repo.
We
could
publish
the
talks
to
the
read
the
docs
page
and
then
retire
this
one
and
point
it
to
that
locally.
That's
a
version
one
like
I
think
python
does
something
similar.
Maybe
let's
quickly
check.
If
it
is
here,
oh
no,
they
actually
copied
it
right
here.
So
so,
every
time
you
do
the
copy,
then
you
run
the
risk
of
not
being
seen
with
the
latest
truth.
A
A
Yeah,
so
it
looks
like
python
house
also
has
a
reasonable
just
like
dot
net,
but
except
that
in
dotted
case.
I
know
for
sure
we
haven't
reviewed
this
for
accuracy
or
correctness,
so
it
could
be
like
way
off,
and
I
I
mean
I
had
like
some
concerns
about
the
title
automatic
instrumentation,
because
this
could
so
the
docs
page,
the
open
elementary
io,
slash
docs
uses
the
word
automatic,
instrumentation
yeah
instrumentation.
I
did
open
an
issue.
Maybe
you'll
find
it
here
that
issue
still
open.
A
I
I
didn't
pay
close
attention,
but
I
did
open
an
issue
saying
that
this
is
incorrect,
because
we
have
a
auto
instrumentation,
which
is
separate.
Maybe
I
did
not
open
it.
Someone
else
opted
and
updated
yeah,
but
I
stopped
paying
close
attention
because
I
didn't
really
have
time
to
like
update
this
one
either.
A
So
one
possible
way
I
can
see
is
like
once
we
are
done
with
the
logging,
which
I
expect
the
logging
specification
would
be
stable
like
later
this
year.
Once
they
are
done,
then
we
can
make
an
effort
to
like
keep
this
updated
for
one
time
based
on
the
status
at
that
time,
like
let's
assume
that
it's
1.4
so
1.4
is
the
first
version
of
open
elementary.net,
which
has
traces
metrics
and
logs,
including
the
stable
exporters
and
probably
the
instrumentations
are
stable.
A
I
I
don't
know
about
that
part,
but
at
that
stage
we
can
do
a
one-shot
one-time
snapshot,
update,
saying
making
sure
this
instructions
are
correct.
That
might
help
users,
but
it
still
has
the
risk
of
like
not
being
kept.
A
How
it
goes,
I
think,
like
folks
who
own
this
one
are
not
from
our
team.
I
mean
no
one
in
this
call
is
owning
this
document,
so
we
cannot
really
make
a
official
statement
that
this
will
be
fixed
or
not.
A
So,
let's
at
least
agree
to
do
two
things
which
we
discussed
so
we'll
modify
the
repo
to
include
the
better
status
here
and
I'll
modify
the
individual
instrumentation
libraries
to
talk
about
its
status
like
whether
it
is
really
nrc
or
because
it's
an
rc4
tracing
for
sure,
but
for
matrix
it's
really
more
like
an
alpha
or
beta,
so
that
status
could
be
like
indicated
in
the
home
page
of
every
every
thing
like
same
with
prometheus.
A
A
C
D
C
Least,
a
little
bit
more
to
where
it's
accurate,
that'd
be
great.
C
E
C
D
C
Where
this
has
kind
of
come
up
because
we're
like
well,
what
do
we
use
now
that
we're
moving
to
a
different
hosting
platform
and
so
we're
going
to
get
an
erase
source
code
and
upgrade
them
to
dotnet?
Six,
probably
so.
This
was
the
conversation.
It
was
basically
gonna,
be
I'm
having
with
sres
right
now
they're
trying
we're
like
a
customer
of
new
relic.
C
A
A
There
is
a
change,
talk
yeah,
it
says
the
package
name
will
be
new.
We
point
to
the
right
one,
so
that
effort
is
done
and
aws
confirmed
that
they
are
not
renaming
it.
So
their
ws
packages
here
would
remain
with
the
name
contribute
because
they
already
went
ga
with
that.
So
they
cannot
change
it
unless
they
are
bumping
the
major
version.
So
the
issue
is
closed
and
second
update
is
we
from
microsoft?
We
contributed
something
which
it's
like.
A
The
name
is
not
very
useful
for
external
folks,
but
we
and
again
this
is
also
not
updated.
So
we
just
donated
this
to
open
source
like
last
week.
I
expect
us
to
update
this
document
here.
It's
it's.
A
It's
a
microsoft,
specific
exporter,
but
truly
it's
just
writing
to
etw,
so
it
can
be
like
used
by
anyone,
not
just
microsoft,
but
that
information
is
not
yet
mentioned
here.
So,
like
you
can
expect
like
some
pr's,
which
will
update
the
dog
initially
at
least
I
don't
expect
like
anyone
outside
microsoft,
to
be
of
finding
this
extremely
useful
because
it
tries
to
etw
and
you
don't
really
need
a
you.
A
Don't
really
have
a
way
to
like
parse
it
out,
but
the
main
reason
I'm
sharing
it
is
right
now
is
we
are
modifying
the
underlying
serialization.
So
once
that
is
done,
it
will
happen
right
right
here.
A
So
you
can
use
tools
like
perfume
or
log
man
just
listen
to
I
mean
it
could
be
listening
to
the
exporter
telemetry
and
it
can
display
the
span
or
metric
or
logs
right
in
like
from
your
command
line,
and
if
you
create
an
agent
which
can
like
subscribe
to
this
and
export,
that
would
work,
but
that's
really
a
little
bit
farther
out.
So
in
case
you
are
seeing
a
lot
of
peers
here.
It's
just
like
we
are
migrating
this
from
microsoft,
internal
to
open
source
yeah.
A
So
today,
like
it's,
not
many
useful
but
yeah,
we
expect
it
to
be
extremely
useful
to
the
whole
community
once
the
like.
Full
migration
is
done
and
we
have
a
good
documentation
updated
here.
So
just
fi
for
folks
from
microsoft,.
A
All
right,
we
don't
have
any
other
topic,
so
I
will
I
mean
alan.
You
probably
just
missed
the
very
first
topic,
so
we
I'll
be
doing
this
release.
You
upload
the
pr
so
we'll
figure
out
the
like
oddness
with
beta
or
alpha
naming
for
prometheus
and
instrumentations
later
before
we
go
stable.
So
I'll
just
do
this
actual
thing,
which
is
to
release
the
1.3
thing
I
called
beta1.