►
From YouTube: 2020-10-14 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).
B
I
heard
an
interesting
thing
that
it's
very
easy
right
now
to
get
a
visa
visa
to
work
remotely
in
estonia.
C
No
idea:
well,
you
don't
need
yeah,
it's
it's
quite
easy
to
get
e-residents
in
estonia.
So
you
can.
You
can
be
like
electronic
resident
in
estonia,
open
business
working
work
here,
yep
that
you
can
do.
B
Yeah,
I
know
that
there
and
the
other
the
other
country
that
is
supposedly
really
easy
to
get
remote
work.
Visas
right
now
is
barbados,
which,
let
me
tell
you,
the
weather
is
probably
better
in
barbados.
B
I
think
if
I
were
younger
and
single,
I
would
definitely
be
heading
to
barbados
to
work.
C
B
B
B
D
E
B
Yes,
actually,
I
I'm
working
a
lot
right
now
on
kind
of
doing
some
stress,
testing
on
the
sdk
and
the
otlp
exporter
and
finding
some
weird
behavior
so
there's
definitely
some
possibility
of
some
fairly
critical
fix
being
needed.
B
B
But
it
definitely
appears
to
be
the
case,
I'm
using
toxoproxy
to
introduce
occasional
failure,
occasional
timeouts
and
if
you
get
a
timeout
on
the
initial
one,
everything
just
doesn't
work,
so
it
doesn't
try
to
reconnect
properly
anyway.
I'm
still
investigating
still
trying
to
figure
out.
What's
going
on.
B
B
If
you
want
to
chat
about
things
and
add
yourself
to
the
attendee
list,
if
you
could
so
just
want
to
give
a
quick
update
was,
I
guess,
was
sort
of
already
doing
that
on
the
ga
burndown
in
the
ga
project
got
11
p2
issues
in
to
do
I've
not
started
adding
p3s
yet
so
we
might
get
to
p3
if
we
can
get
done
with
all
the
p2s.
B
B
Yeah-
and
I
will
say,
the
things
that
are
to
do
are
not
necessarily
prioritized,
so
if
anyone
wants
to
pick
anything
up,
they're
all
open
seasons,
yeah
there's
a
big.
I
think
what
I
think
is
probably
the
last
like
super
significant
proposed
api
change,
to
the
way
that
our
open,
telemetry
global
works
and
that
pr
link
there
1737.
B
honorable
has
made
it
so
that
the
open
telemetry
instance
can
be
in
used
with
dependency
injection
and
injected
and
configured.
It
closes
like
four
issues
that
have
been
opened
for
a
really
long
time
and
based
on
some
early
feedback
from
some
folks
working
with
spring
and
micrometer.
This
will
be
a
really
big
help
for
integrating
with
spring
and
with
micrometer.
B
B
B
H
Can
you
summarize,
can
you
summarize
I
mean
I'm
all
for
having
an
object
that
includes
all
the
the
the
providers,
but
I'm
I'm
not
sure
if
we
need
to
add
yet
another
helper
for
get
meter
get
everything.
B
D
H
It's
most
likely
needed
for
things
that
they
need
both
the
trace
provider
and
the
meter
provider,
correct
and
propagator
yeah
yeah,
so
so
for
modules
that
need
multiple
parts
of
open
telemetry.
This
is
a
good
change,
but
just
for
micrometer,
I
think
you
can
provide
they
can
pass
only
the
meter
provided.
D
Yeah,
I
think
for
micrometer.
I
agree
on
the
instrument
on
the
instrument
on
the
tracing
instrumentation.
We
tend
to
need
both
the
tracer
provider
and
the
propagators.
H
H
Actually
I
don't
know
let's,
let's,
let's
make
it
happen
this
object,
but
I
think
we
need
to
to
document
how
we
expect
people
to
pass
this
object
for
sometimes
or
to
pass
only
the
meter
part,
because
I
feel
like,
for
example,
for
micrometer
and
people.
Don't
that
don't
care
about,
let's
say
seeing
at
all
from
open
telemetry.
They
can
just
use
the
meter
part
and
they
are
done.
The
meter
provided.
H
That
I
I
do
understand
this
problem,
and
but
you
have
also
to
keep
in
mind
that
the
the
plug-in
of
this
is
either
happening
by
the
auto
instrumentation
module
that
puts
everything
together
or
by
the
final
user,
and
I
wish
the
final
user
will
know
that
hey
I
have
this
open,
telemetry
object
and
micrometer
requires
only
the
meter
provider.
So
I'll
do
open
telemetry,
get
meter
provider
and
pass
it
to
micrometer.
B
B
I
don't
think
it's
documented
very
well
right
now,
but
I
think
we
need
to
you
know
really
get
that
documentation,
super
crisp
and
clear,
based
on
which
usage
pattern
you're,
using
whether
you're
in
the
agent,
whether
you're,
writing,
instrumentation
or
whether
you're
an
application
developer
or
operator,
and
make
sure
that
those
usage
patterns
are
very
well
defined
so
together.
Otherwise,
I
think
we're
going
to
be
in
for
a
very
difficult
support
story,
trying
to
talk
through
how
people
are
supposed
to
be
using
it
in
different
circumstances,
so
yeah.
I
agree.
D
I
liked
your
yeah
your
proposal
of
breaking
it
out
by
user
by
kind
of
user
role
or
what,
because
yeah
it's
very
different
for
somebody,
who's
writing,
instrumentation
versus
somebody,
who's
just
instrumenting,
their
app
versus
also,
potentially
somebody
who's
just
using
the
java
agent
and
wants
to
instrument
a
few
extra
things.
B
H
D
G
H
Marketing
implementations,
but
but
yeah,
but
in
this
case
you
you
have
to
to
have
a
logic
somewhere
that,
because
this
is
not
a
concrete
implementation,
it
may
do
other
things
than
what
you
expect
correct.
H
H
G
We
do
in
the
current
implementation.
We
always
return
the
same
tracer
provider.
H
B
H
H
The
last
part
is
the
way
how
we
tell
people
to
change
the
the
things
here.
With
the
we
have
we
talk,
we
tell
people
to
create
a
new
trace,
sdk
provider
like
if
you,
if
you
look
at
the
line
35
in
the
quick
start,
if
somebody
can,
can
you
press
on
that
thing?
Yeah,
I
don't
know
who
is
presenting,
but
that's
me
it's
you
yeah.
So
if
you
go
to
the
file
change
to
the
changes.
B
H
Correct
but,
but
what
do
I
do
so
so
one
of
the
first
thing
in
my
in
our
documentation
would
be
setting
up
the
sdk
I'm
setting
up
this
sdk.
What
do
I
do
with
it?
Because,
right
now,
implementations
may
use
open
telemetry
get
and
what
I
did
do
here
at
line.
45
is
not
going
to
reflect.
What
open
telemetry.get
will
return.
B
D
Well,
I
think
it
actually.
The
way
that's
implemented
currently
in
the
pr
at
least,
is
the
first
one
that
gets
created
also
gets
set
as
that
global,
which
is
a
pattern
that
anurag
liked
from
there
was
some.
He
made
a
comment
that
it
was
from
brave
or
somewhere
else
that
it
had
worked
well.
H
But
how?
How
do
you
ensure
that
static,
initialization
order
is
correct,
because
somebody
may
call
you
know
static
variable,
somebody
may
call
opentermite.get
and
you
have
an
undefined
behavior,
because
that
that,
based
on
the
static
initialization
order
that
may
be
called
first
before
you
do
this
in
your
main
function.
B
I
D
I
Ahead,
oh
sorry,
yeah,
so
I
just
wanted
to
say
that
in
open
tracing
when
like
initialization
returns
a
boolean
or
something
like
that,
but
that's
kind
of
ugly,
but
that
works.
You
know
at
least
you
know
that
initialization
failed.
B
B
Which,
maybe
is
just
like
a
that's
a
bug
in
the
library?
Don't
we
should
put
in
a
pr
to
get
that
fixed
so
that
people
don't
do
that
possibly,
and
this
is
why
I
think,
having
that
those
recommendations
very
if
we
want
to
go
this
way
very
strongly
documented
like
if
you're
writing
a
library.
Don't
do
this.
Do
this
and
like
let
people
configure
you
with
with
the
instance
rather
than
trying
to
grab
it
out
of
the
global
or
trying
to
grab
it
out
of
the
global
yourself
yeah.
D
D
Right,
so
what
what
would
the
pattern
we've
been
following
so
far?
Is
we
allow
injection
of
the
currently
tracer,
but
bogdan
was
pointing
out
correctly
that
we
should
be
that
should
be
the
tracer
provider
and
now
in
the
future,
this?
Hopefully
this
open
telemetry
instance,
but
we
also
have
a
a
way
to
do
it
where
it
uses
the
global,
but
in
either
case
the
user
has
to
call
the
library
instrument.
Oh,
are
you
talking,
though,
about
libraries
embedding
this
directly
in
their.
B
H
Right,
yeah,
yeah
so
anyway.
I
also
think
that
this
pr
tries
to
solve
more
than
it
should
in
one
pr.
I
think
I
think
the
initial
scope
of
this
pr
was
to
add
the
new
object
that
encapsulates
all
the
providers
and
all
the
things
which
is
great,
but
it
tries
also
to
mess
with
the
global.
In
the
same
pr,
I
would.
H
But
this
leads
to
other
things
and
usually
I'm
against
prs
that
have
a
main
scope
which
it
says
here
make
open
telemetry
an
instance,
so
it
can
be
injected
but
also
message
with
other
things
and,
and
you
cannot
think
of
all
the
edge
cases
that
that
can
come.
So
so,
as
my
comment
here
would
be
just
focus
prs
like:
let's
make
that
an
object
and
then
change
the
the
way
how
global
is
initialized
and
so.
D
B
Could
we
well
right
now
the
instance
is
kind
of
is
called
default,
open,
telemetry
right,
so
if
we,
if
we
could,
we
could
start
by
the
first
pr
could
be
to
convert
it
open,
telemetry
and
interface
and
then
to
introduce
the
implementation
and
then
to
mess
with
the
global
that
the
open,
telemetry
and
the
open
telemetry
interface
have
been
exposing.
Is
that
is
that
possible
or
not
positive?
B
G
H
H
B
H
But
that's
what
I
think
that
can
be
moved
in
the
default
open
telemetry,
but
we
don't
mess
with
the
contract
of
this.
So
far,
we
don't
add
the
get
method
on
this
and
stuff
like
that.
So
we
just
made
it
an
interface.
We
we
have
the
implementation
and
we
keep
the
public
api
that
we
have
right
now,
even
though
it's
not
the
best
or
whatever,
and
then
we
change
this
public
api
in
the
way.
How.
B
Yeah,
I
think
that
I
think
I
think
that
makes
sense.
Can
you
just
put
a
quick
comment
on
that
pr
for
honorable,
so
he
can
pick
it
up
tonight.
I
will
do
it
awesome
thanks
yeah,
I
think
that's
a
good
idea.
I've
been
trying
to
rein
in
some
of
the
prs
also
to
make
them
smaller.
G
B
B
I'm
getting
this
really
like
make
sure
that
people
use
this
and
don't
shoot
everyone
else
in
the
foot
while
they're
at
it.
So
cool
other
topics.
Anything
else.
People
want
to
talk
about
hi
ken
how's
it
going
down
there
down
under.
B
And
I
don't
know
ken,
have
you
been
doing
any
of
the
work
on
micrometer
or
has
that
been
mostly
delegated
over
to
erin
for
right
now,.
F
B
D
Honorable
mention
he
invited
some
folks
to
tomorrow's
evening,
pacific
time
morning,
asia-pacific
time
from
my
calendar.
Oh.
D
B
D
I
think
ken
is
tommy
is
that
a
micrometer
person
yeah.
F
D
Of
micrometer
yeah,
okay
yeah,
that's
he
mentioned
tommy.
He
invited
for
that
meeting.
B
Cool
yeah
people
aren't
aware
we
have.
We
do
have
there's
four
total
java
meetings
for
open
telemetry
now
every
week,
which
is
a
whole
lot
of
meetings.
Let
me
tell
you:
we've
got
wednesday
morning,
pacific
thursday
morning,
pacific
I'm
going
to
miss
tomorrow
morning,
but
then
we
also
have
tuesday
late
night,
pacific
and
thursday
afternoon
or
evening
pacific.
So
lots
of
opportunities
to
talk
about
open
telemetry
for
anybody
around
the
world.
I
B
I
think
if
we
can
get
honorard's
release,
automation,
pr
done
and
in
and
the
secrets
there
it'll
become
way
easier
to
do
releases
and
I'm
very
feasible.
B
H
Let's
do
it,
I
think
we
also
do
on
the
hotel
collector.
We
do
it
twice
a
week
twice
a
week
twice
a
week.
No
sorry,
sorry,
every
two
weeks.
J
H
H
I
Yes,
batch
releases
that
I
agree
that
I
agree
with,
but
now
it
could
be
like
more
for
testing.
You
know
it's
like
I
don't
know
alex.
Who
is
a
python
maintainer
told
me
about
this?
How,
like
you
know
allowing
people
to
actually
use
their
latest
stuff?
Wouldn't
you
know
like
show
that
some
things
were
not
perfect
or
you
know
there
was
something
slightly
broken
with
the
design.
So
that's
very
useful.
B
B
I
do
a
release
so
yeah.
I
think
I
would
be
totally
happy
with
bi-weekly
releases
once
the
automation
is
in
place.
I
Sweet
okay
by
the
way
box,
can
I
send
you
a
message
on
either
about
that?
Just
let
me
know.
B
About
you,
we
need
honorary
and
I
probably
need
access
to
secrets.
So
we
can
get
the
automation,
the
github
action
automation
in
place.
I
mean
you,
don't
need
the
secrets.
H
Oh
by
the
way,
yeah
talking
carlos
carlos
sorry
to
answer
your
question,
usually
what
we
did
we,
we
have
maintainers
with
a
maintainer
role
for
the
ripple,
but
in
cases
like
this,
we
either
help
them.
One
of
the
tc
member
helps
them
or
we
make
them
at
me
for
a
couple
of
days
or
whatever,
until
they
figure
out
and
then
change
back
to
maintain.
D
Some
some
repos
have
admin
maintainers
as
admins
on
the
repo.
I'm
not
really
sure.
Why
that's
a
problem.
It
makes
it
much
easier
for
maintainers.
H
Probably
the
main
reason
is
if
we
have
an
upset
maintainer
to
not
be
able
to
delete
the
repo
I'm
joking,
but
I
think
this
is
how
it's
documented
right
now
in
the
community
and
probably
trust
it's
a
good
point
and
I'm
happy
if
you
file
an
issue
in
the
community
repo
and
say
that
maintenance
should
have
admin
for
these
purposes.
I'm
I'm
happy
to
approve
that.
H
H
You
help
the
entire
community
by
by
by
filing
an
issue
to
the
community
specs
and
say
that
this
needs
to
happen
everywhere,
I'm
joking,
but
seriously
john.
Can
you
file
an
issue
in
the
community
and
say
hey
for
these
reasons,
I
think
I
believe
we
we
want
to
have
maintainers
with
the
admin
role
for
setting
up
these
things,
which
is
fine,
but
I
want
I
just
don't
want
to
have
it.
I
want
to
have
consistency
everywhere
and
documenting
the
the
roles
and
in
the
power
and
stuff.
K
G
H
For
java,
I'm
going
to
make
you
guys
pro
at
means
for
the
moment,
because
I
know
we
did
that
in
the
past
just
to
solve
this
issue,
and
then
in
the
mean
after
you,
you
finish
this
john,
please
ping
me
so
I
can.
I
can
revoke
this
until
the
the
community
changes
cool.
I
B
Yeah-
just
I
know,
especially
with
honorable
being
in
the
other
side
of
the
world
since
he's
getting
this
automation
in
place.
Making
sure
that
cycle
goes
quickly
will
definitely
be
good,
rather
than
oh,
I'm
missing
a
secret.
Now
I
have
to
wait
12
hours.
That
would
be
very
annoying
so
cool.
Thank
you,
bogdan
and
cool
anything
else
to
talk
about
carlos.
You
were
supposed
to
leave
six
minutes
ago.
I
don't
know,
maybe.
K
D
H
G
I
H
K
B
I
think
that
well,
I
could
be
wrong.
I
just
looked
at
this
for
the
first
time
this
morning
I
thought
it
was
in
preparation.
I
thought
there
was
a
spec
issue
about
this.
B
Yeah,
I
want
to
get
rid
of
tracing
contacts
utils,
but
I
think
this
I
I
think
we
should.
I
think
it
would
be
better
if
this
was
marked
as
a
draft
pending
what
happens
with
this
spec
issue.
H
H
Fine,
if
we
remove
the
other
two,
but
this
pr
does
not
do
that
so
yeah
right
now
we
just
add
the
third
one
and,
and
my
feedback
would
be-
let's
add
the
fourth
one,
because
it's
not
enough
tree.
D
Yeah,
I
I
really
like
the
span
current
though
like
as
the
it's
very
reflective
of
context,
current
very
easy
to
remember,
but
I.
F
H
D
One
shortcut
logged
in
how
about
I
mean
just
as
far
as
this
coming
back
to
the
last
pr
that
you
were
asking
to
break
it
out
into
steps.
Is
this
pr?
Would
you
be
happy
with
this
pr?
Just
with
like
saying
the
follow-up
prs
would
be
removing,
or
do
you
want
them
removed
in
this
pr?
Also
just
trying
to
give
honorable
clear
advice.
H
I
think
I
think,
based
on
reading
the
comment.
There
is
no
intention
to
remove
the
others.
Actually,
the
the
comment
says
there
that
the
we
want
to
keep
this
on
the
tracer
as
well,
and
that's
that's
where
I'm
getting
confused.
So
if
the
intention
says
yes
add
this
and
then
remove
all
the
usages
clearly
stated
in
the
description,
I'm
fine.
If
there's
a
plan
or
file
an
issue
where
it
says:
okay,
we
are
going
to
move
everything
to
spam.
This
is
the
first
pr
to
towards
that
direction.
B
I
agree
yeah,
and
I
also
called
out
that
I
wanted
other
people
to
comment.
I
wasn't
planning
on
merging
this
until
we
had
greater
consensus
cool.
I
will
put
some
additional
comments
on
this
as
well,
and
also
anyone
who
has
an
opinion
feel
free
to
please
do
so.
B
All
right:
did
you
see
my
pr
to
set
gauges
to
introduce
gauges.
H
Problem
with
that
at
all,
okay,
I
will
comment
on
that,
but
I
just
want
to
make
sure
we
track
and
usually
the
way
how
I
work.
By
the
way,
when
I
split
prs,
I
follow
I
file
an
issue
with
the
final
state
where
I
want
to
be,
and
then
just
link
prs
to
that
issue
just
for
people
to
understand
that
hey.
This
is
part
of
this
whole
effort,
even
though
I'm
not
gonna
produce
a
5000
line
change.
I'm
I'm
doing
this.
D
I'll
chat
with
I'm
curious
to
chat
with
onorak
tomorrow
or,
if
there's
more
dialogue
on
this
issue
before
then,
but
I
remember
was
remembering
something
he
mentioned.
I
think
on
monday
night
about
this
about
the
open
kind
of
and
the
with
the
open
telemetry
instance.
He
was
thinking
of
that.
The
get
current
context,
kind
of
made
sense
living
there
and
the
get
current
span
made
sense
living
on
the
tracer
and
that
you
know,
sort
of
from
a
dependent
di
perspective,
that's
nice
and
from
a
scoping
perspective.
D
So
I'm
not
sure
I'm
wondering
if
he's
still
thinking
about
this.
Oh
this.
This
was
very
recent,
though
just
11
hours
ago,.
B
H
D
Yeah
I
like
that
that
that
proposal
of
issue
that
you
know
to
talk
through
end
state
first
and
then
prs
along
the
way.
J
H
B
H
H
By
the
way,
john,
did
you
tell
a
talk
to
everyone
about
or
or
were
you
in
the
yesterday
meeting
nikita?
Was
there.
B
H
Them
updates
no
awesome.
Okay.
Yesterday
we
had
a
meeting
with
the
reactor
people.
We
talked
about
context
and
cortex
propagation
in
reactor
and
one
of
the
things
that
we
decided
as
a
follow-up
is.
H
We
will
wait
for
for
one
of
the
reactor
person
to
give
us
some
patterns
that
they
believe
are
hard
to
instrument
with
a
tread
local
propagation
mechanism,
and
we
would
like
to
to
work
with
them
and
figure
out
if,
if
that
works
or
not
with
the,
we
tread
local
and
with
our
context
as
a
starting
point
of
discussion
discussing
how
we
can
have
a
better
context
proposition.
H
D
That
summarizes
very
I
had
given
on
our
our
slack
channel
to
honor
og
had
asked
for
an
update
and
that
that
pretty
much
summarized
the
update
I
gave
him
plus
we're
gonna
chat
more
about
it
tomorrow.
D
C
B
D
Meeting
notes
would
be
nice
and,
and
recordings
of
the
meetings
will
be
nice.
Also
I'd
I'd
asked
for
that.
I
D
The
recording
part,
but
meeting
notes
would
be
nice,
I'm
not
sure
I
yeah
some
kind
of
documentation
of
of
it.
F
H
So
I
don't
I
don't
mind
doing
that
in
terms
of
putting
in
the
calendar.
I
don't
think
we
have
very
the
current
meeting
or
anything
they
are
very
ad
hoc,
so
we
can
still
put
them,
but
it's
a
bit
of
a
maybe
just
publish
them.
I
think,
or
whatever.
F
B
D
That's
why
I'm
hoping
to
get
those
recorded,
because
it's
a
very,
I
think,
there's
a
lot
of
people
interested
in
this
discussion,
because
it's
a
one
that
seems
very
unsolved.
D
D
Sort
of
two
different
discussions
going
on
in
these
meetings-
one
is
the
standardization
of
the
context,
which
is
the
one
you're
referring
to
of
there's
so
many
different
ones,
but
the
issue
that's
unsolved,
still
is
even
how
to
do
context.
Propagation
in
react
in
reactor.
D
Yeah,
so
I
think,
there's
kind
of
two
different
discussions
and
bugged,
and
I
don't
know
what
your
thoughts
on
on
you
know
moving
forward.
Those
two
like,
I
guess,
we'll
kind
of
see
how
that
that
plays
out
with
the
two
different
sides.
I
know
that
we're
like
nikita
and
I
are
particularly
interested
in
the
just
even
how
to
I
mean
our
current
immediate
block.
It
blocker
is
even
making
context
propagation
and
reactor
work
for
auto
instrumentation.
D
D
But
I
know
you're
interested
in
the
very
interested
also
in
the
consolidation
of
the
contacts
to
avoid
having
to
sink
bridge,
merge
those
the
overhead
involved.
There.
H
D
Right
yeah,
I
just
wanted
to
call
out
that
there's
sort
of
two
different
two
different
problems
that
we
want
to
discuss
with
them
and
hopefully
solve.
B
H
B
H
Yes,
and
no
so
I
think
I
think
I
think
the
first
step
towards
that
is
they
should
be
able
to
move
an
object,
called
context
or
something
around
across
all
the
streams
and
all
the
execution
things
which
they
are
not
able
to
do
right
now,
because
there
are
many
implementations
of
the
interfaces,
they
have
very
small
interfaces
for
subscriber
consumer
and
so
on
and
that's
their
problem.
I
think
once
they
have
that
and
because
they
ask
kind
of
us
to
to
help
a
bit
with
that.
H
D
D
H
L
I
time
warped
so
bad,
I'm
so
sorry
I
don't
know
if
everybody
saw
my
pr
I
I
haven't
been
able
to
look
at
it
for
a
few
days,
because
I
got
mad
and
needed
to
walk
away
and
think
about
it.
But
I
have
a
bunch
of
tests
working
the
in-memory
exporter
for
for
making
sure
the
tests.
All
work
is
what's
going
to
be
the
problem,
because
it's
super
awkward,
so
I'm
chewing
over
how
to
do
that.
B
Yeah
I've
actually
I've
been
I've,
been
starting
to
use
the
in-memory
exporter
myself
for
doing
a
bunch
of
stress
testing
on
the
sdk
and
finding
it
also
extremely
awkward
to
interact
with
so
super.
L
Awkward
so
like
I,
I
have
a
feeling
if,
if
I
just-
and
it's
like,
I
just
want
to
verify
that
it
gets
called
when
it's
supposed
to
get
called
and
in
the
case
where
you're
working
you
know,
live
runtime
in
the
agent
or
whatever
is
doing
what
it's
doing
like
you
don't
need
that.
I
need
it
to
just
know
that
it's
going
all
the
time
yeah,
maybe.
B
Even
I
could
talk
later,
you
might
need
a
different
kind
of
test
utility
that
operates
at
a
lower
level
that
that's.
B
B
L
L
I've
got
to
work
with
john
and
tommy
about
how
to
turn
some
of
those
things
off,
but
other
things,
it's
just
you
know,
didn't
get
called,
did
the
metric
and
what
they're
testing
is
more
the
function
of
this
registered
med
registry
within
the
context
of
how
all
registries
function
right.
So
it's
like
if
I
hit
a
counter
in
the
top
of
the
composite.
Does
it
hit
this
specific
registry
because
it
registered
properly
and
do
the
filters
work
right?
L
You
know
what
I
mean
so
like
trying
to
restructure
the
entire
micrometer
tck,
because
it
wants
to
update
a
counter,
and
our
in
memory
exporter
doesn't
seems
lame,
because
it's
not
really
what
it's
testing.
It's
just
testing
to
make
sure
all
of
the
pieces
about
micrometer
linkage
are
correct,
not
so
much
what
the
you
know.
What
I
mean
like
what
the
registry
actually
needs
it
to
do
later
is
a
different
problem,
so
I
need,
I
think,
that's
just
like
any
alternate
in
memory.
L
Exporter
kind
of
thing,
john,
is
where
I'm
I'm
thinking
just
something
that
approximates
enough,
that
this
open,
telemetry
micrometer
meter
registry
that
I
have
can
pass
that
tck,
because
again,
that's
mostly
about
linkage
with
that
system,
and
then
I
don't
know
whatever
we
think
needs
to
come
next
after
that,
but
that
we
might
be
able
to.
B
Do
you
might
be
able
to
do
something
like,
rather
than
trying
to
have
a
like
one,
true
testing
utility
to
rule
them
all?
It
might
be
useful
to
just
have
the
ability
to
return
alternate
instrument,
implementations,
and
you
can
just
verify
that
they're.
The
right
thing
is
happening
with
them.
Potentially.
B
I
think
there's
I
think
that
there's
lots
lots
of
opportunities
in
here,
for
you
know,
making
it
making
it
easier
to
work
with
yeah.
B
L
Terms
of
having
having
the
I
would
like
some
people
to
review
the
pr
that
I
have
over
in
the
micrometer
space
only
from
the
sense
of
does
the
open,
telemetry
metered
metric,
whatever
type
that
I
picked,
makes
sense
like
there's
one
corner
case
where
micrometer
allows
you
to
track
two
separate
functions
and
again
it's
the
kind
of
case
where
we
didn't
it's
like.
I
didn't
go
in
and
re-instrument
this
third-party
library,
but
I
know
they
have
these
two
functions.
L
L
L
B
All
right
well
we're
out
of
time
I'm
going
to
take
off
thanks
everybody
for
helping
and
working
with
us.
Thank
you.
B
Yeah,
I'm
going
to
take
a
look
at
that
pr.
Yep
perfect.