►
From YouTube: 2022-09-06 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
B
C
Okay,
I
hear
no
question
so
I'll
quickly
proceed
to
the
relationship
that
we
we
did
the
last
release
about
two
weeks
back,
but
instead
of
doing
the
regular
beta
now,
we
have
also
had
to
deal
with
a
what
we
call
catch
release,
but
potentially
it
can
be.
I
mean
not
potentially
treated
as
a
actual
patch.
So
let
me
find
the
actual
issue
which
prompts
us.
C
Not
this,
I
think
this
would
be
it
yeah
yeah,
so
I
think
blind's
already
investigated.
This
act
unloaded
that,
like
we
need
to
do
some
fix
to
not
crash
the
user
app.
We
did
discuss
with
the
dot
net
team,
also
about
potentially
fixing
it
in
the
diagnostic
source,
but
it
eventually
concluded
that
the
issue
is
not
really
in
the
diagnostic.
So
if
it's
just
said,
open
telemetry
had
independency
on
some
internals
which
got
changed
in
the
latest
version
of
diagnostic
source.
So
now
we
have
to
patch
it
ourselves.
C
So
we'll
use
this
opportunity
to
define
like
how
do
we
deal
with
like
hot
fixes
in
general,
like
how
do
we
communicate
to
users
that
if
you
are
using
1.3.0,
you
should
immediately
get
this
hot
fix
else?
Bad
things
could
happen,
so
we
are
working
on
the
messaging
and
how
do
we
surface
this
to
the
user?
Is
it
just
a
like
announcement
here
or
some
thin
tissue
here,
or
is
it
more
like?
C
Do
we
have
to
do
like
duplicate
the
nuget
itself,
so
that
those
are
the
things
which
we
try
to
learn
and
document
as
part
of
this
release?
So
the
release
the
pr's
are
already
the
changes
required
for
addressing
this
are
already
in.
So
one
of
us
will
do
the
release
today.
C
So
that's
the
first
update
the
patch
and
then
the
regular
beta,
which
I
believe
is.
I
think
we
did
one
for
alpha.
No
two,
so
we'll
do
either
alpha
dot
three
or
beta
dot
one,
depending
on
what
has
changed.
C
There
is
a
good
chance
that
we,
we
might
have
enough
payload
to
call
it
a
beta,
because
we
added
like
new
metric
instruments
like
up
down
counter
and
most
likely
we
would
have
up
min
max
for
histogram
and
higher
performance
histogram
for
large
number
of
histogram
brackets,
like
we
do
have
some
payload,
which
might
make
it
and
also
like
we
already
done
with
two
alpha.
So
might
we
might
just
call
it
beta
yeah?
C
So
I
think,
martin,
you
have
an
issue
to
be
discussed
or
okay
make
senate
conventions
public.
Oh,
can
you
describe
what?
What
is
the
issue
with
this?
Oh
yeah,
I
don't
know
like
we
had
a
dr
to
be
merged
right.
That's.
A
C
Yeah
yeah
this
I
mean
the
pr
was
having
some
issues,
so
I
think
it's
resolved
now
I'll
see
what
the
api
combat
issue
is,
probably
of
like
somewhat
unstable
test
course
in
this
one,
but
trust
looks
good
yeah.
I
spoke
to
like
joe
like
four
months
back
and
he
said
he'll
come
back
to
me
like
later,
but
yeah.
I
totally
forget
like
he
did
come
back
on
too
yeah.
So
no
problem.
C
Yeah,
I
haven't
looked
at
the
most
recent
change,
but
this
doesn't
modify
the
existing
projects
right
yeah.
So
this
is
only
adding
in
project.
We
are
not
yet
changing
the
existing
project
to
depend
on
this
one,
because
that
might
have
issues
about
the
machining,
because
this
this
new
thing
itself
will
not
be
a
stable
thing.
But
if
you
are
modifying
the
instrumentation
like
this
to
depend
on
this,
that
would
be
fine,
but
the
sdk
itself
has
some
dependency
on
resource
simatic
convention
and
exception.
A
I'll
have
a
chat
with
phillip
about
what
he
wants.
Yeah.
I
think,
basically,
what
he
was
asking
was:
can
we
can
we
do
it
and
is
there
any
problems
with
us
doing
it
obviously
he's
put
some
stuff
on
the
that
pr
now
about
it
as
well,
but
that's
fine.
I
just
said
I'd,
raise
it
and
we'll
put
some
effort
behind
it.
So
your
concern
is
that
we
don't
want
to
change
stuff
to
depend
on
it,
but
there's
no
reason
why
we
can't
produce
a
library
that
does
for.
C
C
Has
two
places:
one
is
exception.
Recording
I
mean
the
api.
It
has
a
extension
method
for
record
exception.
I
think
yeah.
So
this
one
depends
on
the
semantic
convention.
So
if
you
modify
the
apa
to
take
a
dependency
on
a
non-stable
package
that
would
prevent
us
from
declaring
ap
itself
stable
when
the
time.
A
C
Yeah,
I
haven't
had
a
chance
to
ask
other
maintenance.
What
do
we
think
of
the
questioning
like?
What
do
we
call
it
like?
Do
we
call
it
like
one
dot
followed
by
alpha
or
some
pretty
release
identifier,
or
should
we
buddle
it
as
part
of
the
core
sdk
and
follow
the
same
conventions?
C
Think
we
want
to
have
it
separate
because
we
know
for
sure
we
want
to
release
the
core
sdk
and
other
components
like
stable
by
now,
but
this
one
is
really
depending
on
that
convention
itself
being
stable,
so
it
would
be
best
to
like
have
a
separate
tag
and
release
it
separately.
So
we'll
take
care
of
it
today,
because
we
have
other
packages
which
have
which
have
need
to
be
released.
C
We
integrated
with
thing
event
source,
and
I
think
we
integrated
with
siri
rover,
so
yeah
two
things
which
we
haven't
released
in
the
last
alpha
and
we
would
be
doing
a
release
this
time,
so
we
have
to
solve
the
versioning
problem
anyway
for
them,
as
also
most
likely
thing
allen
suggested.
We
follow
what
we
follow
in
the
contributor,
where
we
give
it
a
specific
tag
and
it
versions
independently,
because
ideas,
this
would
eventually
move
to
contra
paper
for
semantic
conventions.
C
B
C
C
All
right,
no
other
comments.
We
can
move
to
anything
else.
Left.
Okay,
know
about
what
we
discussed
for
the
semantic
conventions
top
as
well
yeah,
so
yeah,
nothing
else.
We
will
work
on
releases,
including
the
patch
one
and
we'll
figure
out.
What
is
the
best
way
to
announce?