►
From YouTube: 2021-06-15 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
I
think
we
only
took
about
25
minutes
total,
maybe
not
even
quite
that
much
it
turns
out.
There
was
an
instrumentation
sig
kickoff
meeting
that
was
not
well
publicized.
A
I
hope
it's
at
a
time
I
can
join
yeah,
there's
a
so
in
slack,
there
was
a
link
that
riley
posted
to.
Where
is
it?
The
instrument
hotel
instrumentation
channel?
There
was
a
board
created,
and
since
you
wrote
that
otep
I
would
feel
free
to
lobby
ted
for
a
time
that
you
could
make
it.
A
I
keep
on
keep
on
pestering,
hopefully
people
hoping
people
will
show
up.
A
The
was
he
on
the
call
this
morning,
I'm
not
sure
he
was
so
anyway.
Yeah.
There's
two
issues
on
that.
The
board
that
riley
linked
to
you
could
add
one
like
make
sure
the
instrumentation
meeting
is
at
a
time.
I
can.
B
B
A
A
B
A
A
Common,
I
don't
know
if
it
belongs
there,
but
it's
not
it's
just
another
thought.
A
B
A
A
A
B
B
B
B
A
A
Well,
we've
run
into
that
before
is
not
having
that
be
public,
it
was
a
little
was
a
little
annoying
for
those
spi.
I
mean
I
like
sbi's
having
that
config
properties
fastened
into
them,
though
I
mean
just
my
dependency
injection
inclinations,
make
it
make
more
sense
rather
than
having
to
that.
Well,
also,
it's
a
lot
more
self-discover
discoverable,
rather
than
people
having
to
know
that
they
can
go
and
reach
into
this
class
to
get
properties
yeah.
A
A
I
just
wonder
if
there's
some
sort
of
option,
I
don't
think
there's
an
easy
way
to
do
this,
but
some
sort
of
option
where
we
could,
where,
like
the
ken
and
what
he
needs
for
crawl
and
whatever
the
other
thing
is
that
he's
working
on
could
get
like.
Oh,
I
guess
I
mean
I
guess
the
other
option
is
he
could
just
you
know,
do
the
work
that
we've
talked
about
stripping
out
the
code
stripping
out
the
classes.
A
I
don't
know,
I
don't
think
it
doesn't
feel
like
there's
a
great
answer.
Nope.
B
A
Right
it
calls
get
primary
sources
in
the
auto
configure
module
yeah.
So
one
thing
that's
a
little
strange
is
that
the
get
method
is
only
going
to
be
called
in
the
auto
configure
module,
even
though
it's
something
that's
exposed
by
that
spi
module
right,
that's
a
little.
That
feels
a
little
weird,
but
maybe
that's
okay.
B
A
And
then
where's
the
implement,
it's
been
a
while
it's
been
a
week
or
so,
since
I've
been.
A
B
B
A
B
A
I
mean,
I
guess
one
one
nice
thing
about
making
config
properties
an
interface
wherever
it
lives.
Is
that
implementers
like
if
somebody
wants
to
build
their
own
configuration
options
like
their
own
sbi
and
their
own
whatever,
then
they
can
continue
to
use
it.
Well,
I
guess
it's
in
the
update.
They
have
to
depend
on
the
spi
module,
though
it
doesn't
really
help
very
much.
B
I
mean
this
is
just
a
possible
use
case.
I
don't
even
really
expect
it,
but
spring
boot,
for
example,
like
sleuth,
could
have
an
option
itself.
Exporters
which
is
like
a
comma
separated
list
of
sp
implementations
and
so
you're,
just
using
their
same
sp
interface,
and
they
would
of
course,
implement
config
properties
using
spring
boot
stuff
and.
A
B
A
A
Would
we
make
so
the
spi
module
would
also
be
alpha
until
we
feel
like
it's
more
stabilized,
so
it
has
metrics
in
it
right.
So
it
has
to
be
alpha
at
the
moment.
A
And
that'll
make
your
pr
smaller
right,
because
your
your
api
diffs
will
get
small
get
smaller.
B
B
A
B
B
A
A
A
When
I
was
I
was
on
a
team
when
I
was
at
new
relic
like
four
or
five
years
ago,
probably
five
years
ago,
where
the
one
of
the
founding
members
of
the
team
felt
very
strongly
that
gradle
scripts
should
be
like
literally
a
page
long
like
you
shouldn't,
have
your
main
build
gradle
be
longer
than
a
page,
so
he
did
an
enormous
amount
of
work
to
factor
it
out
and
make
and
componentize
it,
and
it
was
great,
like
you
could
read
the
build
like
the
gradle
was,
like
bam.
Bam
bam,
bam
done.
B
A
Yeah,
no,
I
think
it's,
I
think
it's
great,
especially
when
our
build
like
phil
gradle's,
like
1800
2,
000
line
long
or
some
crazy
thing
now,
the
main
one
350.
B
A
Do
we
really
have?
Are
there
really
a
whole
ton
of
unused
imports
in
the
build
gradle
now
or
is
idea
taking
a
long
time
to
optimize.
A
A
Makes
sense,
unfortunately,
it
should
work,
that'd
be
nice,
so
josh
surith
has
been
hard
at
work.
Doing
metrics
working
on
metrics.
It
sounds
like
he
may
build.
A
new
module
sounds
like
he's.
Building
a
new
module
like
he
tried
to
do
the
change
in
place,
and
it
was
just
going
to
be
too
messy,
especially
with
the
dependencies
that
the
sdk
has
on
the
metrics,
like
the
tracing
sdk
dependencies
on
metrics.
We're
going
to
make
the
break
in
place
very
painful.
B
A
A
B
A
Yeah
yeah
I've
used
them
personally
to
verify
that
things
were
working.
It's
like
are
metrics
getting
anywhere.
I
don't
know,
oh
hey,
there's
that
metric.
So
now
I
know
there's
a
mentor
like
I
wasn't.
There
was
no
instrumentation.
It
was
just
that,
so
I
could
see
that
there
was
actually
you
have
some
benchmarks
that
use
them
right
there.
Yes,
I
did
have
benchmarks
that
yeah
there.
B
A
B
A
A
You
know
we
have
a
whole
bunch
of
old
ers.
We
should
just
come
out
of
here,
like
we
should
close
the
draft,
the
batch
observer,
the
two
batch
observer
pr's,
given
that
that
stuff
has
been
pulled
out
of
the
spec
for
right
now
there
are
no.
There
are
no
batch
processors
in
the
or
batch
metrics,
and
then
we
should
close
tyler's
dd
sketch
because
he
hasn't
worked
on
that
thing.
Since
november
yeah.
B
A
I
think
it
needs
to
go
through
its
respect,
yeah
and
it's
also
very
old,
but
we
need
to
go
through
respect
to
get
that
to
work.
I
think
I
like
it.
I
still
like
the
idea,
especially
as
people
are
talking
about
adding
multiple
exporters
back
in
the
environment,
variables.
B
A
That
was
actually
a
discussion
this
morning.
I
think
it
was
kind
of
useful.
I
think
that
was
this
morning.
I
don't
see
it
in
the
agenda,
but
it
sounds
like
josh
surith
in
particular,
is
getting
uncomfortable
with
how
complicated
the
environment
variable
configuration
story
is
and
would
prefer
that
we
think
more
about
a
comprehensive
way
to
do
configuration
whether
it
be
like
a
protobuf
for
configuration
that
that
can
be
mapped
to
via
yaml
or
vo
something,
but
thinking
about
it
more
and
especially
as
instrumentation
becomes
more
formalized.
A
Being
able
to
configure
instrumentation
is
always
also
going
to
be
important.
So
josh
was
saying
he
thinks
that
the
instrumentation
sig
should
probably
also
pick
up
like
the
more
comprehensive
view
towards
configuration,
rather
than
just
adding
ad-hoc
environment
variables,
like
crazy
all
over
the
place,
like
we've
done
so.
A
B
A
Yeah
well
so,
if
you're
a
part
of
the
instrumentation
state,
you
can
help
make
your
dreams
come
true.
B
I'm
very
pessimistic
that
the
meetings
will
be
scheduled
because
I
know
nikita
is
also
reactive
on
the
yeah.
That
is
true,
so
it
needs
to
be
10
p.m.
Specifically,
like
I
mean
it
could
be
in
every
other
week,
approach
like
some
yeah,
like
the
collector,
was
trying
that
for
a
while,
but
since
the
asia
version
I'm
the
only
one
that
joins,
then
people
just
get
a
bit
tired
of
that.
I
think
so.
A
And
you
all
need
to
get
your
button
gear
and
get
the
agent
released.
I'm
gonna
get
that
1.3
yeah.
It's
like
the
longest.
The
longest
lag
after
an
sdk
release
in
a
while.
A
Well,
maybe
it
just
feels
like
there's
a
lot
more
like
scrambling
to
get
things
in
at
the
last
minute,
this
time
with
nikita,
it's
all
nakita's
stuff
and
I
think
mateo
has
his
spring,
something
something
instrumentation
yeah
and
also
the
there's.
I
there's
been
so
much
activity
in
my
inbox
for
open,
telemetry
instrumentation
in
the
past
three
days.
It's
crazy
just
bonkers.