►
From YouTube: 2020-10-08 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
Oh
yeah
yeah,
okay,
yeah,
I'm
not
I've
never
played
eve,
but
I
have
a
lot
of
friends
to
do
and
that
game
sounds
intense,
like
yeah.
A
That's
probably
pretty
fair
yeah.
I
heard
you
have
to
like
spreadsheets
a
lot.
If
you
want
to
play
that
game.
Yeah
yeah.
B
They
actually
just
added
an
option
to
totally
remove
the
graphics
entirely
and
just
have
a
spreadsheet,
basically
effectively
spreadsheet-based
ui,
which
at
times
is
the
best
way
to
interact
with
it.
A
A
Oh
man,
cool
yeah,
I'm
on.
I
also
see
that
you're
on
the
call
I
don't
know.
Let
me
pull.
Oh
there's
only
three
people,
so
it's
all
three
of
us
yeah,
if
you
guys
want
to
add
to
your
name
to
the
attendees
list
anthony.
I
know
you've
already
had
it's
up
to
the
agenda,
but
I'm
not.
If
you
had
anything
else
you
wanted
to
talk
about
today.
A
We
probably
get
started
in
a
little
bit.
Usually
oh
it's
the
afternoon
sessions.
These
are
not
as
intended
that
much.
B
Yeah,
I
think
paul
osmond
from
honeycomb
said
he
was
going
to
show
up
to
talk
about
the
the
issue
that
he
opened.
I
I
think
that's
probably
pretty
straightforward,
but
who
just
wanted
to
since
it
was
so
delicious
to
say,
get
it
on
the
agenda
and
make
sure
that
we're
all
aligned
that
it's
doing
the
right
thing,
yeah
cool
yeah.
I
kind
of
wanted
to.
A
B
Guess
we
could
just
we
want
to
wait
for
him,
let's
give
another
minute
or
two
yeah,
let's
give
him
another
minute,
and
then
we
can
get
started
yeah.
That
sounds.
A
A
A
A
There
we
go,
there's
paul,
hey
folks,
how's
it
going
going
well,
yeah
you're,
giving
me
a
minute
to
show
up.
I
want
to
kind
of
make
sure
we
had
everyone
available
for
talking
a
little
about
things.
I
think
there's
a
pretty
light
agenda
for
the
day,
so
nothing
that
we
couldn't
wait
for
yeah.
I
realize
I'm
not
sharing
my
screen,
which
probably
isn't
going
to
matter
too
much,
but
we
can
pick
it
up
cool
yeah.
We
can
just
jump
into
it.
A
Only
four
people
on
the
call
don't
need
to
go
too
deep
into
the
project.
Boards
definitely
was
impressed
with
some
of
the
numbers
we're
doing
really
good
at
making
some
progress
in
the
run
down
to
ga.
So
I'm
really
excited
about
that.
Don't
need
to
dwell
too
much
on
it
as
most
people
on
the
call
already
know
about
it,
but
it
just
makes
me
excited.
So
I
just
want
to
share
that
excitement.
Yeah.
B
Yeah,
so
I
threw
up
the
initial
pr
for
that.
Last
night
looks
like
it's
already
got
the
approvals,
but
I
just
wanted
to
wait
until
the
meeting
today
to
see
if
there's
anything
that
we
thought
needed
to
be
merged
before
that
release.
There
wasn't.
B
When
I
put
it
up,
but
then
you
got
the
the
trace
api
migration
up
and
it
looks
like
it's
been
approved.
I
think
there
are
a
couple,
others
that
landed
that
are
ready
to
land.
A
Yeah
yeah,
I'm
glad
we're
talking
about
it.
I
don't
want
to
merge
the
trace
api
changes
before
that
release.
I
I
was
hoping
to
make
the
trace
and
metrics
one
kind
of
go
at
the
same
time,
so
I
don't
have
a
metrics
pr
up
so
donate
that,
but
the
protobuf
one
for
hotel.
I
was
thinking
about
that.
That
might
be
a
good
one
to
go
with
this.
I
let
me
just
actually
open
up
the.
A
I
think,
if
that's
best,
that
could
go,
I'm
also
of
the
opinion
that,
like,
as
is
it's
totally
fine,
we
can
just
merge
this
in
the
next
one.
The
reason
for
that
is
because
this
is
a
little
bit
of
a
gray
area,
how
we're
handling
the
the
summary
right
now.
A
I'm
sorry,
the
maximum
count
right
now
we're
still
using
in
the
old
version
we're
like
sending
it
as
a
summary
data,
which
is
what
the
old
product
supports
and
technically
actually,
I
don't
know,
is
the
collector
on
the
v05
that
could
really
dictate
which
one
we
do.
B
Yeah,
I
wasn't
sure
if
that
was
a
backwards,
incompatible
change
or
not.
I
know
they
had
talked
about
their
potentially
being
one,
but
I
thought
they
ended
up
not
needing
to
break
the
proto.
That's
a
good
question!
Oh
man!
I
can't
ever
really
keep
this
all
right.
Well,
it
sounds
like
we
should
probably
then
just
ship
that
release
as
it
is
and.
A
B
A
B
A
The
next
one
then
yeah
cool
that
sounds
good,
so
yeah,
then
why
don't
we
just
jump
right
into
the
next
issue
with
paula
being
here.
Maybe
I
can
open
it
up,
give
a
little
refresher
on
this
one.
B
Yeah
so
paul
raised
this
one
because
it
looked
like
the
he
was
creating
a
sampler
and
it
wasn't
having
its
attributes
added
to
all
of
the
spans
that
he
expected
to
see
it
added
on,
and
it
looks
like
that's
because
we
were
trying
to
avoid
invoking
samplers
when
we
knew
the
parents
was
either
sampled
or
not,
and
thus
we
would
need
to
disable
or
not.
But
I
think
our
expectation
at
the
spec
level
about
how
that
should
be
handled
is
changed.
B
Based
on
the
parent-based
sampler,
that's
been
added,
which
has
the
the
configuration
for
the
different
hooks
for
disabled
and
not
remote
and
parent
or
remote
and
local,
all
of
those
possible
options,
and
it
also
is
awkward
for
samplers
like
paul's,
where
they
want
to
attach
sampling
information
to
every
span,
even
if
they
maybe
don't
necessarily
make
a
different
decision.
Every
time.
A
Okay,
I
see
yeah.
I
think
that
you're
right,
I
kind
of
that's
how
I
kind
of
remembered
it
when
they
were
like
wondering
about
a
like
a
security
issue
or
something
like
that
where
you'd
want
to
resample
or
you
would
not
want
to
like
pass
it
through
the
sampler
after
the
fact,
but
everyone
was
kind
of
like
well,
but
doesn't
that
negate
like
what
the
parent
sampling
thing
was
supposed
to
be
doing
in
the
first
place?
A
I
wasn't
a
part
of
that
conversation
too
much
when
it
was
happening
or
prior
the
old
one,
but
I
think
I
think
you're
right.
I
think
this
change
here
specifically
is
kind
of
making
this
a
little
bit
more
in
line
with
what
you're
talking
about.
So
I
yeah
I
agree.
I
don't
see
like
that
being
too
much
of
an
issue.
B
B
It
probably
been
a
performance
optimization.
I
don't
know
it
predates
my
involvement
with
the
project
that
that
short
circuit
yeah,
but
this
here
clearly
the
spec
expects
that
you
will
invoke
the
the
remote
parent
local
parents,
even
when
it's
sampled
or
not
when
the
parent,
disabled
or
not.
So
I
think
we
just
have
to
invoke
the
sampler
on
every
spain
creation.
A
A
B
I
think
it
was
calling
update,
name
or
set
name
after
this
ban
was
created,
but
the
the
spec
there
has
been
updated
to
to
say.
Yeah
all
bets
are
off
if
you
do
that,
sampler's
not
going
to
get
updated.
So
good
luck.
A
Okay,
all
right
yeah.
Okay,
maybe
I
need
to
update
that
issue
and
saying
we're
just
not
gonna
address
this
until
after
the
ga,
okay
I'll
try
to
dive
into
that
part.
Okay,.
A
B
Yeah,
I
think,
just
to
remove
those
guards
in
in
the
make
sampling
decision.
A
Yeah,
okay,
that
all
sounds
good
to
me.
Yeah
paula,
makes
sense
too
yeah.
Absolutely
I
got
one
ready
to
go
so
I'll
just
submit
the
full
request,
then
yeah,
okay,
cool
that
sounds
sounds
good.
Sounds
like
something
new
yeah
cool
awesome
yeah.
A
We
can
move
on
unless
you
guys
wanted
to
talk
a
little
more
about
that
one
cool
next
thing
I
just
wanted
to
kind
of
sink
on
the
shutdown
method
that
was
added,
I
think
we're
probably
in
line
based
on
who's
on
the
call.
This
was
just
adding
a
shutdown
method.
A
The
backstory
here
is
all
the
samplers
have
are
batch
span,
processors,
all
the
spam
processors.
Sorry
there's
a
lot
of
things
in
this
pipeline.
They
all
have
shut
down
method,
but
there's
no
unified
way
to
shut
them
down
from
the
trace
provider.
It
makes
a
really
good
sense
to
have
that
thing
shut
everything
down,
because
the
operator
would
be
the
one
night
interacting
with
the
trace
provider,
so
makes
a
lot
of
sense.
A
Only
issue
I
had
was
this
adds
it
to
the
api
and
I
don't
think
we
can
do
that
prior
to
the
specification
getting
updated
or
if
it
gets
updated,
which
seems
to
have
some
buy-in
scientific
plus
oneness.
But
the
other
thing
was
this
function
signature.
I
asked
that
the
context
be
included
as
a
call
argument
and
an
error
be
returned.
This
is
to
kind
of
like
mimic
what
we're
doing
in
the
exporter
right
now,
mostly
just
to
return.
A
The
error,
I
think,
is
you
know,
pump
it
upstream
and
let
the
caller
actually
determine
what
they
want
to
do
with
this
and
then
the
context,
I
think,
is
really
useful,
not
as
it's
implemented
right
here,
but
if
we
ever
go
into
a
situation
where
we're
concurrently
trying
to
shut
things
down
or
trying
to
resolve
some
sort
of
like
syncs,
really
quick,
that
we
can
pass
the
context
with
the
deadline
and
that
deadline
can
then
be
the
thing
that
dictates
how
long
this
is
called
and
just
just
hang
the
process
itself.
A
Usually
that's
a
good
idea.
It's
also,
like
you
know,
sid
kill
which
can
end
the
process
pretty
well.
So
I
was
on
the
fence
in
the
past
change,
but
since
we
did
it
in
the
past,
I
just
wanted
to
like
make
sure
everyone
was
so
good
with
doing
something
like
here.
If
it
made
sense.
B
B
A
B
Yeah
and
should
provide
a
way
to
let
the
caller
know
whether
it's
succeeded
failed
or
timed
out.
Oh,
it
kind
of
sounds
like
it
needs
to
return
an
error.
Yeah.
B
Yeah
battle
sounds
correct:
oh
also
completer
abort
within
some
time
out
so
yeah.
We
need
some
way
to
get
that
in
and
out
and
the
canonical.
A
B
A
Cool
I'll
take
an
action
item
on
this
to
open
an
issue
on
it.
After
this.
A
Cool
yeah.
That
sounds
good.
I
think
the
only
other
thing
I
wanted
to
talk
about
today
was
just
kind
of
giving
a
little
bit
of
an
update
on
this
prototype
looks
like
we're.
Pretty
close,
the
is
just
moving
that
api
package
and
kind
of
splitting
it
out
into
the
hotel
package.
I'm
pretty
excited
about
like
the
direction
it's
heading,
I'm
liking
the
direction
that
it's
kind
of
forming
in
the
code.
The
go
package
changes
themselves.
I
don't
have
a
server
running
locally.
I
think
that.
A
A
The
metrics
is
next
to
my
agenda
kind
of
like
what
I'm
saying
I
was
hoping
to
get
those
in
the
same
release,
and
I
imagined
global's
really
quick,
because
I
didn't
figure
out
a
way
to
split
that
yet
because
there's
a
lot
of
coupling
between
that
and
these
two
other
packages,
so
that
may
be
an
add-on
after
the
this
issue
and
then
the
api
test
itself.
I
was
going
to
try
to
merge
this
api
test.
The
metric.
B
A
And
the
tracing
trace
test
all
into
a
singular
hotel
trace
package
is
kind
of
the
direction
of
going
on
that
one
but
yeah,
I
don't
know
if
anybody
had
any
feedback
on
that,
I
probably
was
just
getting
the
status
update
more
for
transparency.
I
don't
think
you
guys
care
too
much,
but.
B
Yeah
I
read
through
the
trace,
move
pr
and
that
all
looks
good.
It
does
look
like
it'll
be
fairly
straightforward
from
the
consumer's
perspective.
They
just
changed
some
some
package
names
and
I
think
that's
about
it.
For
the
most
part
right.
A
Yeah,
which
has
annoyed
me
for
the
longest
time,
but
I
understood
why
it
wasn't
tricity
as
well,
yes,
yeah
yeah,
and
for
what
it's
worth,
I
was
updating
my
local
copy
of
the
contra
package
just
to
get
things
working,
and
this
makes
it
just
a
lot
nicer
to
work
with
personally
yeah,
that's,
okay,
cool!
No,
I
love
that
kind
of
feedback
because,
like
one
of
the
number
one
complaints
we
get
is
like,
why
are
there
three
trace
packages
and
why
do
they
always
conflict?
It's
like
yeah
yeah.
A
It
was
just
a
lot
less
confusing.
Reading
the
code
felt
a
lot
better,
a
lot
less
verbose,
a
lot
clearer,
okay,
cool
yeah.
I
love
that
feedback,
cool
awesome
thanks
cool
yeah.
I
think
that's
it
for
the
agenda.
I
don't
know
if
anybody
else
had
anything
we
wanted
to
talk
about.
I've
got
plenty
of
other
reviews
to
be
doing
and
work
to
be.
You
know,
so
I'm
happy.
B
A
A
But
awesome
well
paul
thanks
for
joining
yeah,
we
love
the
the
contributions
and
any
sort
of
feedback
as
well,
so
definitely
value
it
and
yeah.
It
goes
all
around
cool
yeah
for
sure
I'll.
Get
that
full
request
up
and
submitted
awesome.
That
sounds
good
cool
everyone.
A
I
think
we
think
we're
all
stuff
for
the
day
and
I
will
stop
sharing
and
we
can
call
it
for
the
agenda
and
I'll
see
you
guys
all,
maybe
in
the
metric,
the
45
minutes
or
next
week
see
you
later
see
ya.