►
From YouTube: 2021-02-17 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
Hopefully
this
will
be.
That
was
the
last
one
that
is
all
just
java
maintainers
debating
1.0.
D
A
A
But
yeah
the
plan
will
be.
The
plan
is
currently
that
that
will
be
a
minor
version,
since
we
will
not
be
breaking
any
existing
apis.
A
D
D
A
Her
yeah
at
some
point
I'll
just
close,
get
her
and
people
will
have
to
fend
for
themselves.
A
Trying
to
be
a
good,
a
good
helper.
A
Community
howdy
tyler:
do
you
have
do
you
have
snow
and
ice
in
florida
too?.
E
I
saw
this
graphic
the
other
day
like
every
state
was
freezing
cold
except
florida,
and
they
said
national
hate
florida
day.
E
A
All
right,
it's
the
mic
on
this
shitty
camera
that
I
bought.
A
It's
one
of
the
built-in
so
easy
to
select.
I
put
this
on
when
I
was
interviewing
somebody
earlier
in
the
week
and
forgot
to
remove
it.
A
All
right:
well,
it
is
five
minutes
after
so,
let's
get
going.
I
only
have
one
agenda
item.
If
anyone
has
anything
else,
please
please
add
it
to
the
agenda.
A
We
have
a
few
more
breaking
changes,
they're
all
very
minor.
It's
a
few
renames.
It's
not
anything
super
significant
we're
going
to
be
removing
a
couple
of
at
least
one
thing:
that's
not
in
the
spec
at
all
still
being
worked
on
and
we're
removing
one
thing
that's
needed
for
the
for
the
one
of
the
samplers
again,
I'm
tired,
my
brain
is
not
working
great
anyway,
removing
some
code,
that's
only
needed
for
a
sampler
removing
that
internally
and
not
going
to
expose
it
for
now.
A
A
A
B
So
logistically
does
that
mean
that
there's
effectively
like
a
code
freeze
on
main,
I
don't
think,
there's
a
code
freeze,
no
there's
a
freeze
on
breaking
changes;
okay,
so,
but
what
we're
calling
the
release
candidate?
If
that
gets
promoted?
And
we
say
yep,
that's
the
one
I
mean
there's
still
another
build
that
has
to
happen.
Yeah
we
have
documentation,
we
need
to
add
still
there's
going
to
be.
I'm.
B
A
Yeah
yeah,
it
will
definitely
be
another
build
we
don't
have.
We
don't
really
have
a
way
to
promote
like
there's
not
a
way
to
do
that
in
maven
central
right
either
way
it
has
to
be
a
new
build
and
release
yeah
and,
as
I
said,
we
still
have
some
documentation.
There's
documentation
to
write
here
and
there
nothing
gigantic
and
documentation
can
always
be
added
after
1.0.
It's
not
that's.
That's
totally
fine
as
well,
but
yeah
josh.
Just
to
catch
you
up.
A
The
plan
is
have
that
done
tonight
or
tomorrow
morning
and
then
let
it
bake
for
a
week
in
a
day
and
release
next
friday.
The
26th.
G
Yeah
I
commented
on
those
oh
yeah.
Those
are
the
changes
I
regret
the
most
just
fyi.
Like
always
always
always
those
are
the
ones
I've
regretted.
I
will
let
you
chat
with
bogdan,
I
yeah
I
have
been,
but
anyway,
hopefully
that's
not
the
case
here.
I
just
want
to
call
that
out
of
like
anyway,
if
it
makes
you
feel
better
scholar,
three
is
trying
to
go
to
rc
and
they
just
changed
the
entire
freaking
import,
syntax
nice,
like
two
weeks
before
rc1.
G
A
No,
I
think
I
mean
we
debated
like
I
said
we
like,
we
were
on
the
call
for
an
hour
and
a
half
last
night,
going
through
each
of
these
changes
individually
and
making
sure
that
we
are
all
on
board
with
it
and
that
it
was
the
right
change
and
something
we
wanted
to
do
so.
A
We
did.
We've
spent
the
last
two
weeks
kind
of
wrangling
in
back
channels
about
all
this
stuff,
so
yeah,
I'm
sure
there
are
things
we're
going
to
regret,
but
yeah
we'll
move
forward,
we'll
we'll
work
with
it.
It's
it's
all
good,
I'm
bored!
It's
fine!
You
know
like
yeah.
We
all
agreed
that
none
of
these
changes
are
so
like
it's
not
anything
we
couldn't
have
lived
with,
but
if
we
have
an
extra
week,
we
just
wanted
to
do
some
last
few
cleanups.
A
Yeah,
that's
a
good
question
has
been
merged,
has
only
been
approved
by
two
two
people,
so
I
don't
know
what's
going
on
with
that,
I
mean
I
think
effectively.
Bogdan
has
taken
us
through
that
process,
just
not
with
the
official
checklist.
A
But
yeah
I
I
do
not
know,
hopefully,
since
we've
had
a
tc
member
involved
very
heavily
throughout
this
whole
process,
that
we'll
be
okay,
but
I
suppose
the
tc
could
say
no
at
some
point.
A
The
real
question
is:
will
I
we
do
have
like?
There
are
still
a
couple
things
we
have
to
do
like.
We
really
do
need
to
get
the
the
compliance
matrix
updated
in
the
spec
there's
a
few.
There
are
a
few
last
I's
and
t's
to
dot
and
cross
before
we
want
to
release,
but
that's
why
we
give
ourselves
a
week.
A
Hopefully
it'll
be
enough,
and
hopefully
the
power
will
stay
on
up
here.
Any
other
questions
comments
on
1.0.
G
There's
a
the
metrics
api
working
group,
which
I
think
some
people
here
are
in
are
trying
to
come
up
with,
like
a
set
of
use.
Cases
that
we
make
sure
are
super
super
super
well
supported
and
john.
You
already
commented
on
this,
so
thank
you
so
much.
I
think
anurag
did
as
well,
and
I
apologize
for
all
of
my
typos.
I
actually
did
this
while
speaking
at
a
conference,
so
you
know
I
don't
do
two
things
at
once.
Anyway.
This
is
a
it's
it's
I.
G
What
I
would
like
from
this
community
is
to
make
sure
that
I
am
doing
things
the
way
you
expect
it
to
be
done
in
java
that
it's
following
best
practices
and
that
it's
kind
of
how
you
want
the
api
to
get
used
as
best
as
possible
and
then
also
comment
on
all
the
works
that
you're
unhappy
with,
because
that's
the
idea
for
the
metrics
api
sig
is
there.
They
want
to
use
these
kind
of
this.
G
This
use
case
there's
going
to
be
another
one,
but
I
I
haven't
had
time
to
do
it
to
like
drive.
You
know
improvements
to
the
api,
so
I
just
want
to
make
sure
that
we're
all
on
board
with
what
that
looks
like
to
represent
java
and
like
java
isms,
for
example,
you
know
you
can't
just
like
statically
link
in
something
and
assume
that
these
statics
run.
G
So
that's
not
a
good
design
for
an
api.
You
can't
put
that
in
the
spec,
and
I
want
to
make
sure
that
doesn't
happen
so
so
trying
to
call
that
stuff
out.
You
know
the
notion
of
global
versus
not
global
and
using
it.
There
was
a
discussion
on
there,
so
just
want
to
make
sure
that
I
understand
exactly
how
the
java
one's
meant
to
be
used
in
that
discussion
and,
more
importantly,
so
that
the
metrics
api
sig
can
kind
of
talk
through
it.
A
Yeah
yeah,
the
global
versus
instance
thing
is
something
that
we've
you
know.
We've
evolved
our
thinking
on
that
over
the
past
six
months,
especially
as
spring
sleuth
started,
integrating
with
hotel
providing
a
hotel
back
end
for
sleuth,
it
became
more
and
more
clear
that
the
global
was
not
going
to
be
something
that
was
generally
applicable.
A
So
we
worked
super
hard
to
make
that
be
the
not
the
happy
path,
but
the
if
you
absolutely
must
path.
The
the
counter
point
to
that
is
that
the
agent,
the
auto
instrumentation
agent,
basically
works
with
the
global.
So
if
you
want
to
be
adding
custom
instrumentation
to
your
agent
managed
code,
the
global
is
the
way
to
do
it
and
it
manages
the
sdk
and
sets
the
global
and
sets
that
all
up
for
you.
So
you
don't
have
to
ever
worry
about
it.
A
So
I
think
there
was
one
honorable
had
one
small
example
of
a
library
that
could
only
be
instrumented
via
global.
There
was
no
way
to
get
an
instance
into
it.
I
don't
remember
which
library
it
was,
but
he
had
ones
that
he
had.
He
had
worked
on
instrumenting
in
a
previous
project
where
the
goal
required.
A
What's
the
plan
for
like
jdbc
libraries
yeah,
I
think
it
was
a
jd.
It
was
like
it
was
like
my
sequel
in
jdbc.
It
was
one
of
those
okay.
There
was
no
way,
and
so
the
idea
is
that
auto
instrumentation
is
the
only
way
you're
going
to
get
that
stuff
in
there.
There's
no
way
to
put
in
there's
no
way
they
don't
provide
hooks
for
adding
instrumentation
so
by
code,
manipulation
and
auto
instrumentation
is
the
only
way
to
go,
and
the
agent
already
is
using
the
global
internally.
G
Do
you
think
I
need
to
put
that
in
an
example
to
call
it
out
for
people?
Oh,
like?
Is
that,
like
an
example
using
the
global
as
a
fallback
like
or
in
the
example,
I
could
say
like
never
use
global
here's
one
instance
where
you
might
have
to
that
is
probably
common.
Okay
people
still
use
jdbc
right.
That's
not
like
a
didn't
die.
I
know
the
scala
hipsters
like
totally
got
rid
of
it
right,
but
I
like
in
java,
you
still
use
jdbc
right.
G
A
So
should
you
put
an
example
there
that's
a
great
question.
I
don't
know,
especially
since,
in
your
example
you're
not
going
to
be
able
to
be
demonstrating,
auto
instrumentation,
where
it
is
the
real
place
where
it
applies.
G
I
I
want
to
do
another
one
of
these
with
auto
instrumentation
with
it,
because
I
think
like
calling
out
the
unification
of
the
two
is
really
important,
and
I
think
java
has
the
most
complex
example
of
how
it's
set
up.
So
it's
like
really
important
to
get
it
right
and
then
everyone
else
can
simplify
right.
But
I
that
wasn't
the
goal
in
this
initial
pr.
The
initial
pr
is
just.
How
do
we
want
apis
that
are
direct
users
to
look
like
initially
yeah
yeah?
I
think
it's
also
really
interesting.
A
Because
my
recommendation
for
people
using
the
agent
who
want
to
do
custom
instrumentation
on
top
of
the
auto
instrumentation,
is
to
really
not
use
much
of
the
the
regular
open
telemetry
apis
at
all
like
they
should
be
using
the
annotation.
The
width
span,
annotation
to
add
spans,
and
then
they
should
just
be
accessing
the
span
out
of
the
context
and
doing
stuff
from
there,
rather
than
getting
tracers
and
trying
to
do
all
of
the
the
normal
hard
work
that
you
have
to
do
to
actually
do
make
that
happen.
A
A
Maybe
not
maybe
I
mean
it'd
be
tricky
to
actually
get
it
implemented,
because
you
have
to
go
into
the
agent
and
do
some
do
all
that
agent
bite
cody
stuff,
but
it
might
be
really
cool
to
write
up
an
example
of
like
how
we
could
build
that.
To
do.
To
add
timings
and
then
maybe
like
a
rough,
a
rough
sketch
of
what
the
implementation
under
the
hood
would
look
like
around
doing.
Adding
timing
metrics
to
things
via
annotations
in
the
agent.
G
Yeah,
that
would
be
very
cool.
I
can
see
how
to
do
the
observer
based
metrics
pretty
easily
there's
almost
direct
correlation
with
jmx
in
the
worst
possible
way,
but
you
could
do
that
with
annotations
too
right
timings
are
hard
interesting.
I
like
it.
I
will
I'll
make
a
comment
in
the
thing
about
it.
As
like
a
to
do
cool
yeah.
A
And
if
you
want,
I'm
sure
me
or
nikita
plum
or
any
of
the
people
who
work
over
in
the
instrumentation
project
could
show
you
how
the
instrumentation
is
written
for
the
width
span,
so
you
at
least
have
an
idea
of
what
that
code
looks
like
it's
not
actually
very
complicated.
A
All
you
need
to
do
is
be
able
to
start
a
spam
and
then
you're
kind
of
stick
it
in
the
context
and
you're.
That's
really
all
the
instrumentation
does
in
that
case,
so
might
be
illuminating
to
at
least
c
and
then,
given
that
you
could
say,
oh
if
we
want
to
do
a
timing
thing
like
this,
then
we
would
do
something
similar,
like
I
don't
know,
wrap
it
in
some
sort
of
timing,
logic.
A
G
A
Over
the
past
year,
because
I
want
this
desperately-
the
real
question
is:
do
you
do
it
ahead
of
time?
Do
you
do
some
sort
of
aspect
thing
like
what?
What
does
the
implementation
look
like
to
you?
I
I
don't
know
like
this
is
something
you
could
definitely
imagine
building
into
using
springs,
annotation
framework,
but
it's
or
would
you
use
dynamic
proxies
to
do
it
like?
What's
the
implementation
look
like,
I
don't
know,
and
I
haven't
had
the
time
to
sit
down
and
try
to
implement
it.
A
I
actually
made
sure
that
the
java
agent
has
a
feature
where
you
can
turn
off
all
the
instrumentation
and
just
get
the
width
spanned,
so
I
can
basically
take
it,
so
I
don't
have
to
do
that.
G
Well,
if
you,
if
you
want
to
prototype
what
I've
done
in
the
past,
that's
really
easy.
Is
you
make
an
interface
with
annotations,
so
imagine
an
interface
called
like
foo
metrics,
where
it'd
be
like
report.
This
report,
this
report,
this
and
and
then
you
have
annotations
for
like
the
metric
descriptor
and
labels.
G
That
would
be
really
easy
to
annotation
processor
into
your
apis,
like
really
really
really
really
easy.
So
would
you
do
that,
though,
at
build
time,
yeah
yeah?
That
would
be
a
build
time
so
it'd
be
like
I'm
writing
an
instrumentation
for
a
library
I'm
going
to
make
this
interface,
and
then
I
can
auto
inject
it
and
have
kind
of
like
auto
value,
annotation
processor
right.
So
I
don't
want
to
inject
this
interface
and
it
would
just
use
the
api
in
a
way
where
I
don't
have
to
remember
how
to
set
up
everything
right.
A
Yeah
yeah,
no,
I
absolutely,
I
think
that
would
be
very
cool
to
have
around.
Would
you
do
it
via
just
like
dynamic
proxy,
or
would
you
do
it
via.
G
No,
you
can
you
can
fully
synthesize
all
that
code
yeah.
So,
like
I
like
again,
I'm
you
would
you
would
like
the
lombok
way
more.
G
Yeah,
like
dagger
lombok
sure
I
I
think,
lombok
actually
hijacks
the
compiler
itself,
so
it'll
be
less
invasive.
Oh
yeah
yeah,
but
I
mean
that's.
That's
that's
anyway.
That's
my
my
background
is
all
static,
nothing
dynamic.
So
if
you
want
dynamic,
I'm
not
the
best,
but
that
the
apt
ahead
of
time
kind
of
auto,
auto
value,
slash.
I
forget
the
other
example
anyway
yeah
I
I
can
only
think
of
google
names
of
things.
So
I'm
terrible.
I
I'm
like
useless
in
the
real
world
again.
A
Yeah,
that
would
be
a
fantastic
thing
to
have
somewhere
in
some
contrib
module
available.
It'll
be
red.
G
All
right
I'll
put
that
on
the
back
burner
as
like
a
thing
we
could
do
to
help
simplify
the
api
if
we're
not
happy
with
it,
but
more
more
concerned
about
like.
If
you
look
at
this
and
you
look
at
how
people
adopt
it
any
kind
of
like
things
we
could
make
nicer,
you
know
anything
we
can
do
or
that
like
violate
expectations,
we'd
really
love
to
have
that.
So
thank
you.
A
Yeah,
absolutely
I
will.
I
need
to
take
a
deeper
look
at
it.
I
kind
of
just
took
a
cursory
pass
through
it.
I'm
also
probably
a
bad
person
to
do
that,
because
I've
you
I've
been
living
and
breathing
those
apis
for
a
long
time.
It'd
be
much
better
for
people
who
are
unfamiliar
with
the
apis
to
take
a
look
and
see
how
it
feels
and
see.
What's
unexpected
and
surprising,.
F
F
Is
there
any
appetite
in
and
granted?
I
don't
know
whether
this
would
be
a
spec
thing
or
just
no-till
java
thing
in
terms
of
providing
annotations
for
doing
similar
to
what
some
of
the
spring
stuff
has
out
of
the
box
like
creating
a
new
span
on
this
method,
changing
the
name
of
the
method
kind
of
thing,
so
you
don't
have
to
get
access
to
the
tracer
and
the
span
to
be
able
to
do
some
of
those
things.
A
A
A
I'd
love
to
love
to
have,
I
don't
know
whether
it
lives
in
java,
contrib
or
in
the
instrumentation
project,
or
where
that
code
would
live.
Maybe
it's
a
whole
separate
module
like
an
annotation
processing,
open,
telemetry,
annotations
project
where
that
could
be
built,
but
I
think
it'd
be
great.
I
think
it'd
be
fantastic.
I
want
to
use.
G
F
Well
well,
for
quarks,
we
don't
do
that.
We
just
write
special
methods
that
just
spit
out
by
code,
so
that
during
the
build,
so
we
get
avoid
all
that
kind
of
nastiness
but
yeah.
I
was
thinking
more
from
the
perspective
of
like
just
general
annotations
as
to
whether
that's
something
open
telemetry
wants
to
have,
but
obviously
it
would
need
some
kind
of
implementation
or
people
to
implement
it.
F
A
I
think
there's
definitely
I
mean
none
of
this
is
in
the
none
of
the
annotations
are
in
the
spec.
This
is
purely
a
java
thing
at
the
moment,
but
I
think
that
I
certainly
think
it's
great,
I
think
it'd
be
valuable.
Okay,
I'm
on
the
record
that
I
hate
annotations,
except
when
I
love
them.
So
no
as
long
as
we
don't
try
to
implement
something
as
insane
as
jax
rs
is
that
annotation
language,
then
I'll,
be
reasonable.
Yeah.
F
A
What's
that
the
documentation
swagger,
the
swag.
B
Well,
I
mean
I
was
also
just
thinking
of
other
languages
that
have
any
form
of
annotation
processor
dot
net.
Has
that
has
something?
Doesn't
it
I
think
so,
and
then
I
found
they're
called
decorators.
I
think
yeah
same
thing,
though,.
A
A
So,
anyway,
yeah
this
is
the
one
thing
that
always
scares
me
away
from
closures.
They're
like
and
you
get
meta
programming.
I'm
like,
I
don't
want
metaphorical.
I
just
want
programming
programming.
A
Yeah
exactly
cool
all
right.
Well,
I
I
we
should
add
some
notes
here.
Somebody
should
have
notes.
Somebody
just
find
me
annotations,
more
less.