►
From YouTube: 2020-11-19 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
A
B
Hey
yeah,
I
think
I
see
some
people
adding
their
names
to
the
agenda.
Please
do
so.
If
you
haven't
already
and
any
items
that
you
would
like
to
discuss,
add
to
the
agenda
section
and
probably
wait.
Another
minute
start
into.
B
B
David,
are
you
out
of
australia
by
chance.
C
Are
but
okay,
currently
out
of
sunnyvale
california,.
B
Cool
yeah
we
can
jump
into
it.
Then
thanks
everyone
for
joining
just
kind
of
start
us
off.
Look
at
a
little
bit
of
a
status
update
on
the
project
board.
From
last
time.
You
could
definitely
see
some
solid
progress.
Getting
things
done.
B
I
also
noticed
that
we
had
a
lot
of
stuff
come
in
and
move
through
this
board
pretty
quickly.
So
thanks,
everyone
for
being
diligent
on
that
and
the
pr
reviews
and
comments
and
responses
are
much
appreciated,
good
way
to
move
with
going
forward.
I
think
that
there's
also
been
a
lot
of
interest,
I've
kind
of
noticed
in
the
past
few
days
from
the
good
con
event
and
the
community
days
which
I'm
really
excited
about.
I
hope
the
community
can
grow
and
grow
and
grow,
but
yeah
other
than
that.
B
I
only
had
a
few
things
to
talk
about
this
week,
so
it's
gonna
be
a
pretty
short
meeting.
If
no
one
else
has
anything
else
to
talk
about,
but
we
can
kind
of
just
go
into
it.
The
administrative
side
of
things
I
wanted
to
start
off
by
asking
next
week's
meeting
is
on
thursday
in
the
u.s.
B
I
know
it's
thanksgiving,
which
means,
I
also
know,
I'm
not
going
to
be
here
to
run
that
I
don't
know
if
there
are
other
people
that
wanted
to
run
that,
but
I
was
proposing
we
probably
should
cancel
the
meeting
next
week.
B
I
second
that
okay,
cool
I'll
sync
with
morgan
after
this
to
cancel
the
calendar,
invite
but
yeah,
I
think
that's
that's!
That's
gonna
carry
cool
next
thing
on
the
agenda.
Is
we
had
in
the
past
meeting
talked
about
this
api
refactor,
which
I
think
it's
not
closed.
Oh
no!
Okay.
It
is
closed.
B
It's
done
crazy
has
not
done
the
call
he's
in
berlin
right
now,
so
it's
very
late,
but
yeah
fantastic
job
from
him
on
that,
and
I
think
having
that
done,
I
was
looking
at
some
of
the
other.
You
know
kind
of
outstanding
pull
requests
just
kind
of
briefly.
B
I
was
wondering
if
there's
anything
else
we
wanted
to
try
to
get
in
before
we
try
to
push
out
a
release,
hopefully
today
or
tomorrow,
which
I'm
happy
to
facilitate.
D
B
B
This
is
where
I
get
confused
on
how
to
actually
check
what
version
we're
pegged
to
well.
There's
a
hash.
B
Okay,
oh
yeah,
yeah,
sorry,
yeah,
look
at
that
turns
out
reading
it's
an
important
skill
to
have
yeah.
I
I
think
we
should
probably
try
to
make
an
effort
to
update
to
v06.
I
can
try
to
take
that
action
item
on
this
evening.
Actually
I
think
I
have
enough
time
to
try
to
get
a
pr
out
for
that,
and
then
we
can
try
to
get
a
release
done
tomorrow.
B
If
that
sounds
good
anthony.
What
do
you
think.
F
B
B
That's
a
that's
a
really.
I
am
super
excited
about
this
development,
so
yeah.
A
B
Think
that
you
nailed
it
that
sounds
like
pretty
spot
on
for
those
that
aren't
kind
of
following
along
anthony
is
mentioning
the
fact
that
theater
is
this
new
repo
that
was
just
created
this
morning
around
that's
intended
to
house.
It
has
nothing
in
it
right
now,
but
it
was
intended
to
house
all
the
generated
proto
code
in
it,
which
is
really
cool.
We
didn't
want
to
actually
expose
this
from
our
internal,
build
of
this
and
there's
a
few
other
places.
This
is
used
and
there's
a
lot
of
other
issues.
B
It's
actually
a
pretty
deep
history
of
issues
here
so
having
this
is
really
ideal
to
have
a
centralized
place
for
the
build.
So
the
goal
here
is
to
try
to
get
this
really
structured.
B
I
think
I'm
probably
just
going
to
continue
on
with
our
old
method
of
using
a
sub
model
module
just
for
the
interest
of
getting
this
next
release
out
at
a
somewhat
faster
cadence
and
then
building
this
repository
up
is
kind
of
on
my
agenda.
B
I
don't
know
how
we're
gonna,
how
we
wanna
have
overlap
with
this
sig,
given
the
fact
that
this
and
the
collector
both
use.
This
proto
repo,
I'm
guessing
they're,
going
to
both
use
this
puerto
rico.
We
should
just
probably
sync
with
the
proto-
are
the
collector
guys
to
see
if
they're
collecting
people
to
see
if
they're
interested
in
overlapping
with
this
but
yeah?
I
think
that
this
is
kind
of
just
a
cool
opportunity
to
centralize
on
this
and
build
some
automation
around
it
as
well.
B
So
that
being
said,
I
guess
I'm
asking
if
anybody's
interested
in
helping
build
this
out
or
has
a
lot
of
experience
doing
these
kinds
of
things.
It
would
be
greatly
appreciated
if
you
wanted
to
help
contribute
to
this.
Otherwise,
it's
gonna
be
me
bumbling,
along
and
probably
making
some
mistakes
which
I'm
I'm
good
at
doing.
B
Also
good
at
admitting
him
hopefully
but
yeah,
so
I'm
excited
about
that
yeah.
Here's
one
of
the
other
relevant
issues
yeah.
I
think
that
was
actually
all
I
kind
of
had
I've
been
digging
pretty
hard
through
the
backlog
of
issues.
Lately,
I
think
we're
actually
doing
pretty
good
for
rc
track.
Maybe
we
can
talk
a
little
bit
about
that
if
there's
nothing
else
to
kind
of
go
over.
I
think
that
we
kind
of
looked
at
this
last
time
and
it's
looking
pretty
good.
Yes,
how
do
I
do
this.
B
Yeah
this
has
gone
up
in
the
past
few
days.
I
think
that's
also
for
me
to
blame.
I'm
picking
that
up.
I
think
we
were
at
six
last
time.
We
took
a
look
at
this,
but
some
of
these
are
a
little
bit
old.
I
know
that
krezmir
has
taken
some
interest
in
trying
to
address
some
of
these
in
the
long
term,
so
having
kimbo
kind
of
double
down
on
some
of
these
should
help
move
some
of
these
longer
spending
points
forward,
but
yeah.
B
I
think
I
think
it's
tractable.
I
don't
know
if
it's
going
to
be
tractable
in
the
next.
You
know
four
weeks
or
something
like
that,
but
that's,
I
think
I
think
a
reasonable
pace.
I
don't
know
if
anyone
else
saw
it
in
the
getter
channel.
Yesterday
josh
was
saying
that
the
tc
had
kind
of
loosely
pegged
the
end
of
january
to
have
the
metric
specification
locked
down.
I
know
yesterday
they
also
merged
the
tracing
specification
freeze,
so
the
1.0
freeze
or
the
potential
really
scan
it
for
the
1.0
freezer.
B
The
specifications
should
be
out
so
that
shouldn't
actually
change
anymore.
B
So
I
think
we
need
to
do
probably
some
better
auditing
and
just
making
sure
that
these
numbers
are
more
accurate,
going
into
the
feature
and
then
really
just
double
down
and
getting
the
work
done,
I
think,
is
the
idea,
but
just
a
kind
of
kind
of
a
heads
up
if
anybody's
scared
of
carrying
the
pulse
of
the
project
right
now,
I
think
there's
a
fair
amount
of
work
to
do,
but
we're
we're
making
good
progress
kind
of,
as
I
started
out,
where
we're
still
cranking
through
issues,
and
hopefully
a
lot
of
that
kubecon
energy's
going
to
keep
us
new
contributors
moving
forward.
B
So
yeah
that
sounds
good.
B
Is
there
anything
else
anybody
else
wanted
to
bring
up
prs
david?
I
saw
that
you
had
some
questions,
maybe
about
the
the
bridge
yeah.
C
Sure
happy
to
discuss
that
yeah.
So
the
interesting
piece
is
that
I've
written
a
an
open
census,
binary
propagation
propagator
for
open
telemetry,
with
the
idea
being
that,
if
I
would
like
to
use
or
to
have
an
open
telemetry
based,
go
application
talk
to
over
grpc
to
an
open,
census-based
one
I'd
like
them
to
be
able
to
communicate
in
the
same
propagation
format.
C
C
Yeah,
I
pasted
it
okay,
so
and
it's
the
question
at
the
bottom.
I
think
yeah
so
basically
I'd
like
to
add
an
example
and
this
example
would
rely
on
the
open,
telemetry,
grpc
library,
from
the
contrib
repo,
as
well
as
all
the
rest
of
the
code
in
this
repo.
So
the
question
is
basically.
C
C
Yeah
the
bridge
would
be
in
this
repo
because
essentially
it's
a
bridge
of
the
sdk
right
and
then
the
the
propagator
would
be
next
to
the
hotel
grpc
code,
because
it's
really
sort
of
a
way
to
hook
up
open,
telemetry's
grpc
to
open
census
grpc.
So
in
a
way,
it's
almost
a
bridge
for
that.
C
B
B
Gotcha,
okay,
so
anthony
kind
of
sounds
like
you're
leaning
towards
this
number
two
option
here.
I
didn't
actually
think
too
hard
on
it,
but.
F
B
Through
number
three
here,
yeah
gotcha,
I
think
that
that
kind
of
was
like
where
I
was
thinking
of
going
as
well
yeah.
I
actually
think
that
that's
probably
a
good
idea.
D
I
would
I
would
put
the
propagator,
I
think,
there's
a
propagators
package
in
contrib,
so
I
think
it
would
live
there
and
then
the
example
would
maybe
go
with
it
rather
than
with
the
grpc
instrumentation.
B
This
yeah
it
should
this
b3
and
this
jager
propagator,
actually,
I
think,
need
to
get
moved
back,
not
which
is
a
thing
things
change,
so
I
feel
bad
to
anyone
who
was
part
of
that
move
to
get
it
here,
but.
B
I
think
if
we
put
the
the
binary
propagator
for
open
census,
I
think
that
is
that
is
an
appropriate
one,
as
defined
by
the
specification
currently
to
fit
into
this
contrib
repo.
So
that
would
make
sense
to
me.
B
B
80
generators
sounds
good.
Yes,
this
is
great.
I
just
was
taking
a
look
at
this.
Not
this
there's
a
open
census,
I'm
sorry
aws
x-ray
apr
for
a
trace
id
generator
and
just
can't
keep
people
up
to
speed.
I
haven't
actually
read
this
ticket
yet,
but
I'm
guessing.
This
is
what
I'm
talking
about.
The
pr
here
included
some
interface
to
allow
for
people
to
provide
their
own
custom.
B
Id
generators
should
be
pretty,
I
guess,
transparent
from
the
pr
title,
and
this
is
definitely
something
that
is
a
duplicate
of
this
internal
id
generator.
B
So
I
think
that
the
idea
here
is
that
I
asked
for
this
pr
for
this
issue
to
be
open
to
expose
that,
because
I
do
think
that
this
id
generator
interface,
if
we're
going
to
have
that
so
people
can,
you
know,
write
custom
instrumentation
around
that,
as,
as
is
being
done
with
this
aws
x-ray
id
generator.
We
want
to
have
that
as
an
interface
in
the
sdk
itself,
that
you
would
then
implement,
and
then
you
could.
B
A
B
Anthony
made,
you
know
just
for
some
backstory
on
this,
like
anthony,
made
some
good
points
that,
like
you,
know,
having
a
panic
in
this
id
generator
is
probably
not
ideal,
and
maybe
maybe
these
methods
need
to
be
flushed
out
a
little
bit
more
is
so.
I
think
we
should
probably
spend
some
time
on
design
work
on
that.
D
Yeah,
I
don't
know
if
the
the
methods
there
need
to
change,
it
would
probably
more
be
the
mechanism
for
registering
an
id
generator
to
use
right.
You
need
some
way
to
indicate
that
id
generation
failed
and
allow
or
the
creation
of
that
id
generator
failed
and
allow
the
application
to
decide
what
it
wants
to
do.
D
It
may
be
that
the
application
wants
to
say.
Oh
if
I
can't
generate
ids,
I
want
to
panic,
but
I
don't
know
that
we
should,
from
within
our
library,
be
deciding
to
panic.
I
think
that's
been
a
guideline
that
we've
tried
to
stick
to
fairly
closely.
Is
that
we
won't
panic,
we'll
just
tell
you
and
you
decide
if
you
want
to.
B
Yeah,
I
completely
agree.
I
try
to
stay
away
from
panics
as
much
as
possible.
That
being
said,
I
think
I
was
the
person
who
recommended
we
have
a
must
function.
So
maybe
I
can't
talk
too
much
yeah
I
think
I
saw
will
wilbur
on
the
call.
I
don't
know
if
wilbur,
you
have
to
say,
yeah.
B
F
No,
I
think
you
guys
basically
summed
it
up
pretty.
Well,
I'm
a
bit
new
to
those.
So
I
didn't
know
that
panic
wasn't
like
a
good
idea
to
handle
the
errors
and
stuff
but
yeah.
I
can
change
that
for
the
future.
If
we
decide
on
a
specific
way
that
we
should
handle
it.
B
Yeah,
I
think
I
think
anthony's
got
a
really
good
point.
So
if
we
expose
this
id
generator
interface
in
the
sdk,
then
it's
just
all
about
registering
it.
So
that's
how
it
would
be
registered
and
in
that
registration
process
it
should
be
able
to
like
you
know
the
id
generator
return,
an
error
on
registration
or
something
or
on
creation,
or
something
like
that
and
the
sdk
should
be
able
to
handle
that
error.
We
do
have
a
little
bit
of
an
error
handling
pipeline
in
the
hotel
system
right
now.
B
D
Yeah
and
the
the
must
function
that
you
mentioned
right.
That
may
be
an
option
that
we
provide
so
wilbur
if
you're
not
terribly
familiar
with
go
and
how
panics
are
sometimes
handled.
D
Is
that
if
you've
got
a
thing
that
your
application
absolutely
needs
to
work
in
order
to
function
oftentimes,
there
will
be
a
must
do
thing
kind
of
function
that
that
takes
a
function
that
might
return
an
error
and
if
it
does
it
just
panics
and
gives
that
error
back,
but
that
takes
it
out
of
that
creation
or
initialization
of
the
id
generator
and
moves
the
choice
about
whether
to
panic
or
not
up
to
the
application.
That's
trying
to
initialize
it.
D
So
if
someone
says
yes,
I
absolutely
need
this
id
generator
to
work,
and
if
it
isn't
going
to
work,
then
my
application
shouldn't
even
continue
to
start.
They
can
say,
must
register
id
generator
and
otherwise,
I
can
say,
create
this
id
generator.
I'm
going
to
look
at
the
error.
If
there's
an
error,
then
I'll
register,
this
backup
type,
you
know,
maybe
I've
got
another
id
generator
that
will
work
just
as
well,
but
I
prefer
this
one.
D
B
Cool,
I'm
gonna
add
this
as
a
required
for
ga
because
it
seems
pretty
useful,
but
if
anybody
has
different
opinions,
seeing
a
nod
so
cool
yeah,
I
think
it's
a
useful
feature
set.
I
don't
think
that
you're
the
first
one
to
ask
for
this
either
wilbur
so
yeah.
I
appreciate
you
diving
in
because
it's
something
I
think
a
lot
of
users
could
potentially
use,
although
I
think
there's
probably
some
potential
to
misuse
it
as
well,
but
we'll
we'll
see
how
that
goes.
B
And
we'll
try
to
prioritize
what
we
did.
I
did
prioritize
cool.
B
Cool,
I
think
that
was
it
for
my
agenda
again
just
kind
of
open
up
the
floor
again.
If
there's
any
other
issues,
you
guys
wanted
to
talk.
C
B
Okay,
cool,
if
that's
the
case,
then
I'm
gonna
say
we
can
probably
end
it
early
and
everyone
is
I'm
sure,
pretty
drained
from
attending
a
conference
virtually
or
even
just
a
long
week,
but
yeah
I'm
also
just
a
heads
up.
I
am
going
to
be
out
most
of
the
week
next
week.
I'll
probably
be
somewhat
infrequently
online,
just
a
heads
up
on
that,
but
then
I'll
make
sure
the
meeting
itself
is
is
canceled
for
next
week.
B
So
yeah
I
imagine,
most
people
are
going
to
be
out
next
week
as
well.
So
if
you
are
looking
for
rapid
responses,
I
maybe
reset
your
expectations
is
all
I
can
say:
okay,
well,
cool
thanks
everyone
for
joining,
and
I
will
see
you
all
virtually
or
maybe
in
the
metrics
meeting
in
37
minutes,
but
yeah
cool,
bye.