►
From YouTube: 2020-11-25 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
A
A
B
A
We
can
start,
we
don't
have
a
lot
of
agenda,
but
I
can
quickly
summarize
what
I
learned
over
the
last
two
days
about
rally.
So
since
the
spec
has
not
officially
announced
what
does
it
mean
to
be
one
point:
dot
one
dot
outdoor
dot,
we
are
asked
to
refrain
from
releasing
the
actual
one
point
of
package,
so
I
do
not
know
yet
when
we
will
be
able
to
officially
release
the
ga
version
or
lts
version,
because
the
definition
of
ga
slash,
lts,
slash
1.0,
is
still
undecided
from
the
specification
committee.
A
So
what
we'll
do
is
we'll
continue
to
release
rcs,
which
is
still
okay
to
be
released,
but
the
actual
date
for
1.0.0
is
still
unknown.
I'll
have
some
update
as
soon
as
I
hear
from
the
technical
committee,
slash
governance
committee.
A
So
that's
the
only
major
update
I
want
to
make
about
release
plans,
so
we
would
still
be
doing
like
the
next
rc's
first
client
and
for
number
30
instead
of
1.0.
We
may
just
do
another
c
or
if
there
is
a
clarification
already
made
in
the
specification,
we'll
actually
do
1.0
on
the
specified.
A
Okay,
couple
of
updates
related
to
the
release
as
well,
so
there
are
two
packages
which
I
am
initially
proposing
to
hold
off
from
the
1.0
release.
I
mean
this
is
assuming
that
1.0
is
approved
from
the
spec,
but
we'll
still
be
holding
off
these
two
packages.
The
primary
reason
is,
we
don't
really
think
they
are
of
ga
quality
for
z
pages.
It's
not
just
about
the
quality,
it's
just
to
make
sure
that
the
tech,
the
respect
for
z,
pages,
it's
not
yet
officially
in
the
specification
report,
so
we
just
need
to
hold
off.
A
A
So
we
probably
did
not
have
time
to
finish
this
in
the
next
couple
of
days,
so
my
backup
plan
is
to
don't
release
it,
so
we'll
have
like
enough
time
to
design
it
right.
There
are
many
suggestions
for
the
package
how
to
like
how
to
name
the
methods,
how
to
configure
things,
since
I
did
not
have
the
time
yet
to
incorporate
all
of
them.
A
If
I
ever
finish
this
before
1.0
time,
then
we'll
do
it,
but
my
backup
plan
is
to
hold
it
off,
there's
an
open
pr,
but
it
still
has
questions
which
are
unresolved,
but
that
I
said
like
every
other
package
would
use
the
same
version
of
whatever
we
choose,
whether
it
is
rc
or
one
parameter.
It
will
be
the
same.
C
There
has
been
some
churn,
so
in
terms
of
what
do
we
actually
want
to
call
the
release,
so
I
think
instead
of
ga
the
new
the
new
I
guess
verb
picture
is
long-term
supportability,
so
I've
tried
to
incorporate
lts
and
general
availability
as
the
way
to
announce
this,
but
I
definitely
need
feedback
comments
from
everyone
in
the
community
to
make
sure
we're
hitting
the
right
set
of
a
release
items
and
be
calling
out
things
that
people
should
keep
in
mind
while
using
this
sdk.
C
So
please
go
through
this
send
in
your
comments.
I
think
I
want
to
definitely
wrap
it
up
before
our
meeting
on
monday.
So
before
before
the
release
on
monday
and
before
you
know
everyone
before
the
community
meeting
on
monday
itself
as
well.
A
C
So
I'm
not
sure
I
think
I
reached
out
to
morgan
and
I
think
every
individual
language
will
be
making
their
own
announcements.
That's
the
sense
that
I
got
from
morgan
and
sergey,
but
I
think
that's
why
I
drafted
this
out.
Hopefully
by
monday
we
have
a
little
bit
more
clarity
from
in
the
maintainers
meeting,
because
I
think
that's
when
the
announcement
actually
goes
out
right.
So
I
feel
like.
C
A
One
more
question,
so
we
will
be
like
aligning
this
with
the
date
when
we
actually
are
able
to
ship
1.0
right.
So
if
1.0
is
delayed
by
like
two
weeks,
then
we'll
delay
this
yes.
C
But
please
go
through
this.
I
think
I've
I've
taken
a
rough
hack
at
it
being
a
pm,
there's
certain
things
that
I
want
more
clarity
on
especially
hey.
Are
we
going
to
be
doing
any
sort
of
breaking
changes
to
metrics
api
and
should
that
be
called
out
explicitly?
C
So
I've
definitely
tried
my
best,
but
look
at
it
make
sure
you
add
links
make
sure
we're
calling
breaking
changes
out
explicitly
because
this
is
a
little
bit
confusing
saying
that
this
is
a
long-term
supportability
path
and
generally
available,
but
we're
still
introducing
breaking
changes
later
on
right,
so
hey
exception
for
matrix
and
it
should
be
called
out.
Yeah.
A
C
A
1.0
is
cleared
by
the
yeah
on
monday,
right,
which
hopefully
they'll
clear
it,
but
if
not
we'll
just
have
to
delay
it
until,
like
governor's
committee,
gives
the
green
light.
D
So
on
this
I
know
we
don't
know,
you
know
the
dates
for
that
1.0,
but
my
impression
was
that
we
weren't
going
to
be
able
to
you,
know,
announce
ga
or
like
it
was
going
to
include
metric
support
and
that
the
metrics
you
know
spec
freeze,
isn't
till
like
the
january
time
frame.
So
is
that
still
the
case
here
for
for
net,
or
are
we
going
to
be
able
to
say
ga
before
that
happens?
No.
A
I
don't
think
we'll
ever
wait.
We
don't
plan
to
wait
for
matrix
as
soon
as
the
tracing
is
clear.
We
should
be
able
to
call,
because
for
us
the
metrics
are
already
marked
as
absolutely
we
already
marked
all
the
matrix
as
absolute,
which
means
we
should
be
able
to
break
it
in
a
without
bumping
the
major
version.
So,
if
matrix,
if
you
want
to
ever
add
it
back,
we
can
still
break
the
api
without
bumping
to
2.0,
because,
like
even
before
1.0
it
is
marked
absolute.
A
So
in
any
version
of
like
1.5
or
1.6,
we
should
still
be
able
to
replace
the
api
with
a
new
one,
okay,
yeah
and
like
the
there
is
still.
I
don't
think
it's
agreed
by
everyone
that
matrix
will
freeze
on
january.
It
is
from
what
I
hear
it
is
still
still
debating
in
the
spec
committee.
It
could
be
very
well
into
february
or
march.
A
I
do
not
have
any
like
exact
answer
to
that.
So,
but
at
least
one
thing
is
clear:
if
we
continue
with
the
original
plan
of
building
on
top
of
dot
net
api,
then
we
will
not
be
getting
before
november
2021.
Like,
even
if
the
spec
itself
is
ga
by
january,.
A
It's
very
unlikely,
but
this
is
my
guesstimate
based
on
the
information
which
I
have
so
if
we
are
going
to
recommend
opentelemetry.net
for
any
customer
who
is
serious
about
metrics,
they
should
wait
until
november
2021,
but
we
still
have
a
like
alpha
beta
quality
matrix
api,
which.
A
A
Yeah,
so
we
should
definitely
include
this
link
to
this
issue
in
the
doc.
We
already
linked
it
in
the
absolute
message
for
metrics,
so
anyone
who
is
installing
the
package
and
using
metrics
they
should
see
a
warning
which
will
take
them
right
into
this
issue.
A
A
Okay,
yeah,
I
mean
it
should
be
called
out
here
as
well
about
maybe
it's
like
something
which
we
can
avoid
like.
We
can
totally
avoid
mentioning
metrics
in
the
overall
messaging
and
at
the
bottom
say
that
we
do
have
a
like
alpha
version
of
metrics.
Please
try
it
out
and
share
feedback.
Something
of
that
so,
but
those
feedbacks
we
can
handle
in
the.
C
A
A
Okay
got
it
so,
let's
hope
that,
like
we'll
have
something
by
next
monday,
so
we
can,
like
I
mean
just
to
make
sure
like
we
are
not
missing
any
key
things
which
other
languages
are
yeah
for
sure,
yeah,
okay,
yeah
and
I
can
share
an
update
as
soon
as
I
learn
from
the
aspect
committee
about
the
1.0
time.
A
A
I
don't
know
what
this
really
means,
like
github
actions
could
not
connect
to
download
the
actual.net
call
runtime,
but
I
haven't
seen
this
occurring
in
like
last
day,
it's
like
very
rare.
Now
it
used
to
be
like
very
common.
We
initially
assumed
that
it
was
because
like.net
release
something
new
and
everyone
is
like
overloading
the
download
servers
and
that's
why
we
are
having
this
issue
it.
It
is
still
possible
that
that
is
the
case,
but
we
don't
know.
A
Since
we
are
not
seeing
this
lately,
I
think
we
can
assume
that
this
is
already
solved,
but
that
will
bring
us
to
the
actual
ca
issue
which
is
in
our
control
to
solve
by
the
way,
if
anyone
sees
like
more
of
this
just
being
in
the
jitter,
and
we
can
try
to
prioritize
that
so
far,
we
were
hoping
that
those
dotnet
download
issues
would
just
go
away
on
itself,
but
that
I
said
we
still
have.
B
E
A
I
think
I
did
retry
one
of
this.
There
was
a
failure,
but
I
retread
it
and
now
we
don't
have
it
so
yeah
this
one.
It
did
not
actually
fail
it
just
aborted.
We
have
seen
this
also
occurring,
but
not
very
common.
It's
usually
there
is
a
one.
There
is
like
one
or
two
tests
which
is
failing,
but
this
is
not
a
good
example
either,
because
this
is
also
very
like
a
boating
due
to
some
infra
issues.
B
Was
a
grpc
one
that
was
failing,
but
I
think
for
now
they're
all
passing,
there's
no
good
example,
but
I
think
there's
a
grpc
one.
That's
failing
sometimes.
A
These
three
are
the
ones
which
I
have
seen
failing
frequently
and
it
should
be
like
the
batch
processor
involves
some
timer
and
sleeps,
so
it
may
be
a
little
bit
more
tricky
to
fix,
but
we
should
be
able
to
get
to
event
source
very
quickly,
so
the
next
task
is
like,
if
you
see
a
actual
test,
failure
just
create
an
issue
paste
the
actual
link,
preferably
the
screenshot
or
or
the
actual
failure.
So
we
can.
A
I
mean
it
shouldn't
be
hard
like
we
should
be
seeing
it
very
frequently.
It's
just
that
we
don't
have
anything
right
now:
okay,
yeah
sure
yeah.
I
don't
know
about
grpc
recently.
It
was
there
like
a
few
weeks
back,
but
I
think
I
learned
like
it
did
fix
the
grpc
flickiness
about
week
and
a
half
back.
I
do
not
recollect
seeing
anything
specific
to
grpc,
but
if
you
do
see,
one
just
keep
an
issue
like
create
an
issue,
so
we
can
prioritize
it.
F
A
Even
source
one
was
what
I
saw
like
at
least
three
times
in
the
last
two
days,
and
I
just
rerun
it
so
it
passed
eventually,
but
this
has
been
a
point
of
frequent
ca
failure,
so
I
I
think
I
can
create
an
issue
for
this,
because
I
have
seen
it
like
so
many
times,
so
I
can
create
an
issue
for
this
and
start
looking
at
it
everything
else.
A
If
you
find
just
open
an
issue
link
to
it,
I
don't
think
anyone
except
approval
for
the
supposed
to
like
rerun
the
ca,
so
the
repro
should
still
be
remaining
there.
When
we
get
issues
so
yeah,
it
should
help
me
help
us
with
the
investigation,
but
sometimes
it's
not
possible
to
initiate
just
based
on
the
logs,
but
we'll
see
how
it
can
be
done
all
right,
yeah.
I
do
not
have
any
agenda,
but
just
a
quick
update.
There
are
like
plenty
of
issues
which
were
open
over
the
last
two
three
days.
A
Thanks
everyone
for
opening
issues.
I
did
not
go
through
full
list.
If,
like
anyone
has
free
cycles,
please
take
a
look
and
like
if
it
is
worthy
of
fixing
before
ga.
Please
bring
it
to
my
attention
or
any
one
of
the
programs
so
that
we
can
tag
it
accordingly,
but
any
other
thing
which
does
not
require
any
change
to
api.
A
We
can
like
let
it
be
like
that
for
at
least
another
week
and
then
fix
it
after
after
the
day
yeah
I
mean,
let
me
quickly
open
the
issues
which
are
still
required
as
ga,
so
the
there
are
three
issues
which
are
open
about
the
extension
stored
hosting
package.
That's
the
one
which
I
was
talking
initially
about
potentially
delaying
from
ga.
A
A
Okay,
I
do
not
have
any
other
topics
to
discuss.
So
if
you
have
questions,
we
can
take
it
other
ways.
We
can
get
back
40
minutes.
B
One
tiny
question:
cj:
can
you
open
the
readme
for,
like
the
sql
client
instrumentation,
for
example,
yeah,
and
on
this
example
here
that
doesn't
compile
right
there
there's
no
lag
version
that
allows
this.
It
has
to
be,
has
to
be
using
var
or
has
to
be
something
yeah,
that's
right,
yeah!
This
is
bug.
Okay,
there
are
a
few
of
those
throughout
different
readmes
I'll
I'll.
Do
this.
A
Into
apr,
okay,
yeah
it.
This
is
right,
like
the
the
way
we
handled
that
initially
was
in
the
docs.
We
tried
to
link
to
the
actual
code,
so,
for
instance,
if
it
is
getting
started,
we
don't
write
like
a
lot
of
code
here.
We
just
ask
people
to
copy
it
from
the
actual
program
so
that,
if
it
is
a
compilation
error,
we
would
catch
it
right
here,
because
this
is
actual
code.
That's
one
way
of
handling
it.
A
I
do
not
know
if
there
is
any,
because
this
way,
if
any
apa
changes
will
break
the
build,
so
whoever
is
breaking
it.
They'll
just
go
ahead
and
fix
it.
But
that's
a
good
call
like
this
is
a
general
question
which
I
haven't
really
thought
about.
How
do
we
publish
dogs
because,
right
now
we
are
duplicating
a
lot
of
things
in
the
readme,
for
instance,
let's
take
the
secret
itself,
so
we
have
all
this.
A
I
mean
some
of
them
require
explanation
like
and
usage,
but
most
of
them
should
be
just
documented
in
the
like
code
command
itself
without
requiring
an
explicit
entry
in
the
document.
So
it's
not
yet
clear
to
me
like
what
is
the
best
way
to
handle
it,
because
we
will
eventually
want
to
use
docs
as
the
official
place
for
every
construct
and
then
link
to
like
if,
if
you're
talking
about
like
some
specific
concept,
which
is
only
explained,
I
said,
document
sorry
as
a
code
command,
we'll
just
link
to
the
corresponding
one.
A
But
it
is
not
because
I
think
python
is
only
language
which
is
formally
doing
the
actual
publishing.
We
don't
yet
do
it.
But
this
is
something
which
we
are
trying
to
figure
out
like
what
is
the
best
way
to
publish
the
version,
though
so
that
it
should
work
whenever
we
release
yeah.
So
for
now.
The
only
action
item
is
please
update
like
if
you
see
any
like
typos
or
bugs,
which
doesn't
compile
it
and
we'll
eventually
move
everything
into
actual,
for
example.
A
So
that
will
catch
compilation
issues
all
the
time,
and
this
was
raised
back
in
another
case
also
before
I
think,
because
for
some
of
the
exporters
we
have
too
many
options
listed
here
like
each
and
everything
we
don't
need
to
list,
at
least
because
this
should
be
like
difficult
to
maintain
as
we
modify
things.
So
it
is
best
documented
as
a
code
command,
and
we
can
have
like
automatically
generated
those
from
that
and
that's
what
python
is
doing.
A
F
F
One
thing:
actually,
I
see
that
he's
on
kevin
mckinley
is
new
kevin.
You
used
to
be.
F
Yeah
I'll,
let
him
chime
in,
but
he
was
asking
a
question
basically
about
the
the
ga
preparedness
ask
that
we
put
out
there
and
if
there's
anything
remaining
there,
that
you
think
could
use
some
more
eyes
and
also
I
mean
I
think,
we've
gotten
some
pretty
good
coverage
here,
but
also
just
some
a
general
question
of.
F
When
do
we
call
this
closed
or
how
does
an
individual
say
that
like
to
give
them
a
thumbs
up.
A
A
A
Yeah,
should
we
just
leave
comments
at
the
bottom?
I
think
yeah.
That
would
be
fine,
I'll,
just
leave
a
comment
asking
everyone
like
when
you're
done,
like
please
post
a
comment
here
that
you're
done
and
if
you
still
have
free
cycles.
Please
look
like
try
to
pick
more
yeah.
I
mean
there
are
still
few
left,
but,
like
I
mentioned
in
the
beginning,
if
there
are
things
which
are
not
yet
taken
by
anyone
else
like
just
the
approvals
and
maintenance
will
just
split
it
between
themselves.
A
So
yeah
I
mean
we
still
have
like
few.
I
mean
I'm
not
very
worried
about,
like
z
pages
or
from
appears,
because
we
may
hold
off
z
pages
and
prometheus
is
about
matrix,
so
it's
relatively
less
priority,
but
we
still
need
jaeger
and
yeah.
Jager
is
a
biggest
one,
which
seems
to
be
missing
yeah
and
there
is
grpc
client
this
one
we
already
talked
about
like
we
will
be
delaying
from
release,
and
I
think
allen
is
already
helping
with
this.
Even
though
he's
officially
not
taken
this
yeah
I'll.
A
Okay
yeah,
so
next
is
I
mean
what
was
the
name
of
the
new
person?
Let
me
go
back
to
my
zoo,
yeah
kevin.
Do
you
want
to
just
introduce
yourself?
I
don't
recollect
seeing
you
before.
E
Yeah
hi
I'm
kevin
mckinley
very
new
to
exploring
the
open
telemetry
what
it
is
in
general.
The
concepts
are
not
terribly
unfamiliar.
I
spent
pretty
recent
three
years
at
new
relic,
not
working
on
the
the.net
team
with
alan,
but
working
physically
near
him
back
when
we
could
work
in
the
office
together,
yeah
and
it's
just
really
a
cool
project
and
and
interested
in
it
and
excited
to
be
able
to
help
with
it.
E
If,
if
I
can
so,
I
picked
up
a
couple
of
docs
doc
reviews
in
the
that
open
issue
for
ga
and
I'm
happy
to
pick
up
pick
up
more
more
pieces.
Whatever
is
important.
A
Okay,
I
mean
feel
free
to.
There
is
no
requirement
that,
like
only
one
person
should
do
it
like,
especially
for
the
ap
and
sdk
the
more
ice,
the
better,
because,
in
fact,
the
open
element,
these
two
projects
are
the
foundation
on
which
everything
else
is
built.
So
if
you
can
get
more
than
one
person
to
review
it,
that
would
be
also
helpful.
So
if,
if
you're
still
looking
for
like
ways
to
contribute,
then
please
take
things
which
are
not
yet
taken
or
just
take.
A
A
Okay,
yeah
I'll,
put
a
comment
here
to
like
make
sure
everyone
who
has
signed
up
is
aware
that
when
they
are
done,
they
should
be
putting
a
note
here,
and
still
I
don't
know.
When
do
we
close
it?
It's
thing
like
when
we
get
a
green
light
from
the
committee
about
1.0
being
okay.
A
We
just
need
like
two
or
three
days
before
that
to
close
this,
because
if,
if
there
is
a
major
issue,
then
we
want
to
address
right
away.
If
it
is
not
broken,
then
we
just
tag
it
as
a
bug
and
still
proceed
with
1.0.
So
only
if
it's
a
issue
which
requires
apa
changes,
we
would
be
like
actively
working
on
it.
It's
already
clear,
like
there
are
several
issues
which
are
marked.
A
I
mean
these
are
new
one,
but
there
are
still
other
issues
which
are
open
as
perk
it's
known
issue,
but
we
don't
really
require
any
apa
change.
I
just
put
it
like.
As
the
next
release.
I
don't
know,
benny
is
1.1,
so
it
really
depends
on
when
it
is
1.0,
but
at
least
I
made
it
clear
that
it's
not
going
to
be
fixed
in
1.0,
okay,
yeah
thanks.
Everyone
see
you
next
week.
Hopefully
we
have
good
clarity
on
the
release
date
and
we'll
be
ready
for
1.0.
Oh,
I
just
wanted
to
highlight
one
thing.
A
So
if
you
have
like
internal
applications
or
any
other
applications
which
you
intend
to
instrument,
there
is
no
need
of
waiting
for
open
elementary
to
release,
because
the
recommendation
from
us
is
to
use
diagnostic
source,
which
has
g8
two
weeks
back
or
number
ten,
so
you
can
still
do
instrumentation
with
just
the
diagnostic
source
package,
so
you
only
need
the
open
telemetry.
If
you
intend
to
propagate
context
or
read
context,
but
for
everything
else,
you
should
still
be
like
able
to
do
instrumentation.
A
We
are
still
trying
to
get
it
into
a
couple
of
instrument.
A
couple
of
common
libraries.
I
did
not
have
anything
like
we're
trying
to
get
it
into
like
the
kafka
client.
There
was
one
more
library
which
we
were
trying,
but
those
libraries,
it's
typically
you
can
just
use
the
open
element
with.
Without
you
can
without
open
telemetry
api,
you
can
still
do
the
instrumentation.
So
if
you
have
like
libraries,
you
can
start
on
coding
right
now
without
waiting
for
us
being
1.0.