►
From YouTube: 2020-09-16 Java SIG
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
F
E
E
G
G
G
You
should
you
should
really
move
the
the
java
sig
meeting.
A
Yeah
I
100
agree.
I
will
bring
it
up
on
friday
at
the
very
early
meeting
as
a
topic
of
discussion,
especially
with
honorable.
There.
E
G
Oh
yeah
yeah,
you
should
start
yeah.
How
do
you
do
that
with
the
honorary
needs
very
early
or
late
in
the
day.
H
A
E
E
G
Yeah
yeah,
because
the
friday
morning,
meetings,
I
know,
are
super
bad
for
europe,
because
it's
it's
like
we're
done
for
the
week.
G
G
Okay,
so
I
I
had
with
there
was
an
email
chain
about,
because
dinah
trace
was
the
only
vendor
who
seemed
to
need
java,
7
support
in
the
sdk,
and
so
we
had
reached
out.
I
was
discussing
with
sarah
novotny
at
microsoft,
who's
on
the
open,
telemetry
governance
committee
and.
G
She
knows
alois,
so
she
had
a
recommendation
to
she
had
reached
out
to
him
to
discuss
sort
of
what
the
dynatrace's
specific
needs
were
for
java
7.
G
and
whether
a
fork,
whether
we
could
fork
the
open,
telemetry
sdk
repository
and
have
a
seven
version
of
that
repo,
a
java,
7
version
of
the
api
and
sdk
and
make
you
know
dynatrace
folks,
could
be
maintainers
of
that
repo
since
you're
the
ones
who
need
that
support,
and
it
seemed
like
that
was
acceptable
to
aloise.
G
F
A
G
Yeah
and
we've
had
I
mean
we've-
we've
had
very
that
conversation
in
there
was
the
issue
that
that
anurag
opened
a
little
bit
ago,
also
where
that
kind
of
discussion
happened.
So
I
guess.
G
Yeah
we
just
wanted
to
chat.
We
were
planning
to
chat
tonight
about
sort
of
next
steps.
A
We
can
yeah
with.
I
mean,
I
think,
that
if
we
could
get
pinned
down
alois
and
get
a
like
a
an
actual,
like
yes
go
ahead,
and
do
this
or
maybe
we
should
just
tag
him
on
the
issue
and
see
if
we
can
get
some
like
get
a
confirmation.
There.
G
Yeah
I
was
thinking
also,
potentially,
we
could
open
a
community
an
issue
on
the
community
repo
asking
to
create
open,
telemetry
dash
java,
7
repo.
A
G
Oh
yeah,
so
it
would
be
very
similar
to
like
the
java,
instrumentation
repo
and
the
dd,
the
datadog
repo
right
right.
So
it's
kind
of.
G
Even
if
it's
official
fork,
though,
how
does
that
make
it
easier
if
it's
an
official
fork,
I
thought
you
still
had
to
do
all
the
merges.
A
E
Do
we
still
have
any
ones
correlation
context?
Does
this
have
to
be
finished
before
we
can
even
do
any
port.
F
G
We
want
to
do
it
before
ga,
specifically,
because
if
we
wait
till
after
ga,
then
we're
stuck
in
the
the
api
for
backward
compatibility,
we
can't
really
make.
F
F
A
F
E
A
Yeah,
the
the
grpc
context,
one
is
going
to
be
like
somebody
needs
to
grab
that
and
be
able
to
actually
spend
some
cycles.
Some
significant.
E
A
A
E
A
E
E
A
A
A
G
G
I
think
that
grpc
issue,
that
quote
from
the
spec,
because
I
found
it
particularly
enlightening
or
I
agree
it
was
confirmation
by
us.
A
E
E
G
C
G
Of
the
friday
java
sig
meetings-
and
I
assume
if
because
if,
if
it's
not
in
the
notes-
and
he
I
don't
think
he's
ever
really
responded
much
on
those
in
the
issues.
So
it
must
have
been
in
the
meeting
that
he
said
that
the
spec
could
be
that
he
could
change
the
spec
yeah.
E
Yeah
right,
I
so
my
suggestion,
rather
than
shading
your
pc
context
as
is
or
anything,
we
should
write
something
it
can
be
the
same
api
without
the
deprecated
stuff
and
with
the
better
service
provider
mechanism
than
what
they
have,
because
I
really
hate
that
mechanism
and
other
than
that.
If
it
looks
the
same
migrating
or
instrumentation
also
shouldn't
be
hard
right.
I
guess
it's
abstracted
mostly
by
tracing
contextuals
anyways,
so
yeah.
G
I
have
a
couple
from
an
instrumentation
perspective.
I
think
I'd
put
a
couple
notes
in
there
as
far
as
my
wish
list,
if.
G
There
was
something
that
a
couple
things
that
would
make,
I
forget
what
they
were
yeah.
E
G
Was
a
to-do
comment
about
that,
but
the
I'm
thinking
more
about
instrumentation
points
for
like
the
constructor.
E
A
Well,
honorable,
is
there
a
way
that
you
and
I
could
collaborate
on
getting
that
grpc,
like
lifting
the
code
kind
of
working
together
to
get
that
you
know,
get
the
pieces
we
want
out
of
there
and
the
deprecated
parts
gone
and.
E
Yeah,
what
would
be
one.
E
That's
a
great
strategy
for
me,
that
is
fair.
A
B
Yeah
one
one
more
like
vicious,
I
don't
know,
I
don't
know
if
that's
too
much
of
a
scopes
creep,
but
current
context
is
very
very
very
thread:
local
centric
and
for
the
last
two
weeks
I
was
struggling
with
project
reactor.
B
H
E
A
Yeah,
if
you
think
you
can
get
that
cranked
out,
that'd
be
awesome.
Yeah.
G
Fantastic,
I'm
not
volunteering
for
anything,
but
it's
worth
just
a
quick
read
of
this
because
I
I
kind
of
like
you
know
the
type
like
thought
through
some
of
the
issues
around
typed
keys
or
untyped
keys,
dreadlocal,
not
thread
local
and
yeah.
A
G
G
Yeah,
I
know
I
I
I
think
I've
been
I've
been
persuaded
to.
I.
I
thought
that
this
was
a
good
idea,
but
I've
I've
since.
E
A
E
A
E
A
B
C
C
C
E
G
Oh,
you
guys
have
been
cranking
on
the
instrumentation
yeah
yeah,
it's
been
a
it's
been
yep,
it's
been
hard
to
keep
up.
Yeah.
C
G
A
G
E
A
Do
it
I
still
get?
I
still
gave
two
weeks
notice,
so
maybe
they'll
actually
pay
me
for
two
weeks,
even
though
I'm
not
working
there,
I
don't
know
how
it
works
exactly,
but
and
splunk
said
that
if
they
actually
wanted
to
keep
me
for
two
weeks
that
I
I
could
delay
my
start
date,
but
I
would
have
been
very
surprised
they
have
all.
The
managers
have
strict
orders
that
if
anyone
is
going
to
a
competitor
that
they
need
to
be
immediately
terminated.
E
E
B
E
G
Yeah,
I
think
we
were
good
because
it
seemed
like
pretty
much
all
the
java
features,
except
for
like
completable
future
was
really
the
only
one
it
seemed
like.
We
got
all
everything
we
wanted
from
d
sugaring
default
methods,
yeah,
de-sugared,
okay,
yep,
static,
yeah,
static
methods
on
interfaces,
de-sugared.
E
E
E
Well,
no,
that's
not
so
if
we
separate
export
out
of
this
sdk
because
presumably
they
can
just
have
a
different
spam
processor,
but
this
so
the
spam
processor
will
still
need.
G
Yeah
we
should
get.
I
wish
we
had
somebody
who
was
your
co-worker,
who
came
to
the
meeting
john
one
of
the.
G
E
G
Yeah
just
go
while
you're
at
it
go
ahead
and
do
the
convert
the
agent
to
do
compile
time
instrumentation
and
of
android
apps.
E
Could
have
discussed
it
a
bit
and
like
lambda
does
have
pre-provisioned
functions,
so
you
can
still
use
an
agent
if
you
really
want
to,
but
for
on-demand
users,
it's
impossible,
and
so
at
some
point,
if
there's
aot
that'll
help
the
server-less
cases
and
grow
vm
users.
I
guess
everyone
is
going
to
need
this
at
some
point,
so
we'll
need
it
at
some
point.
I
guess
yeah.
E
G
It
would
be
great
for
for
a
lambda.
G
Since
you
control
the
the
infrastructure
you
can,
I
mean
you
have
all
the
code,
you
could
instrument.
E
G
It
sounded
like
the
among
the
similar
to
and
the
the
way
that
we
I
mean
is
that
j
the
statements
are
json,
and
so
I
would,
I
would
say,
let's
just
do
the
same
thing
that
we're
the
same
scrubbing
that
we're
doing
in
which
is
essentially
preserving
the
json
structure
and
just
scrubbing.
A
B
B
B
G
I
might
close
it,
but
no
no
there's.
G
No
there's
some
caching
in
here
that
is
decent
that
I
would
like
to
bring
over-
and
richard
mentioned
in
the
last
comment
that
he
is
in
the
datadog
repo
he
moved
stuff
that
should
make
this
nicer
for
us,
so
I'm
gonna
I
haven't-
I
haven't,
looked
at
it
since,
since
he
did
that
so
I'm
hoping
that
makes
things
a
little
bit
better
and
then
I
need
to
catch
up
on.
61
is
out,
so
I
need
to
go
through
cherry
pick.
G
Some
commits
I'm
gonna
for
the
61,
I'm
gonna
do
individual
prs
for
each
one
perfect.
Thank
you.
Yeah.
G
Agree
force
flush.
We
chatted
about
that.
G
It
would
be
that
I
was
trying
to
think
of
how
the
it
would
be
nice.
We
wish
we
had
a
smoke
test
for
any
time.
We
have
like
shading
problems
like
this,
that
are
that
we
only
catch.
We
can
only
catch
via
smoke
test
currently.
G
B
Our
our
current
smoke
test
capture
the
output
from
containers
anyway,
so
we
can
augment
that
application
inside
the
container
that
we
monitor
to
produce
to
use
those
trace
id
and
fans
ide
in
logs
and
then
just
check
that
log
does
contain
that.
So
I
think
it
should
be
quite
quite
stable.
G
B
G
Yeah
I
I
noticed
that
I
was
wondering
what
to
do
about
that.
That's
a
good
idea.
G
G
B
Should
get
it
through
yeah
and
you
can
go,
you
cannot
assign
that
directly
to
material
right
away.
He
is
not
in
organization
yet.
G
Okay,
yeah,
because
we
had,
I
knew
that
the
the
shading
around
our
logging
instrumentation
is
fragile.
So
I
had
called
that
out
here
for
smoke
that
we
needed
smoke
test
for
it.
But
then
we
bumped
logging
instrumentation
for
post
ga,
but
then
we
brought
it
back
for
mdc.
So.
G
Yeah
I'll
I'll
resurrect
these
in
a
new
issue.
B
G
What
was
the
I
mean,
I
don't
really
have
a
preference
if
it's
a
single
or
broken
out
into
separate
ones.
I
mean
it
could
still
be
a
single
one
right
that
uses
all
four
different
logging
frameworks.
B
G
B
Pen
not
yet
so
it's
for
me,
it's
a
little
bit
a
separate
problem,
because,
even
if
I
use
type
spans
I
I
still
have
like
I
want
to
verify
and
like
document
which
semantic
convention
we
actually
use
because
a
lot
of
them.
B
G
What
you
didn't
just
suddenly
decide
you
wanted
to
hack
into
the
embedded
sdk
inside
of
the
make
the
first
instrumentation
that
talk
directly
to
the.
E
G
John
anything
else
on
the
java
side
that
you
want
to
chat
about.
While
we've
got
the
whole
team
here
or
a
big
chunk
of
the
team.
A
Only
just
to
I
think
we
got
nikita's
approval,
I
don't
know
the
honorary.
Did
you
get
the
type
typed
attributes
approved,
I'm
trying
to
remember
who.
A
E
A
I'd
like
to
get
that
done
and
not
think
about
it
anymore
after
months,
actually,
the
so
that
case,
insensitive
name
thing
that
giovanni
put
in
this
morning,
I
think,
looked
good,
but
I
thought
it
would
be
worth
another
eye
on
it.
I
think
it
looks
fine,
though.
A
A
Oh
yeah,
that
javadoc
thing
was
the
synthetics
yeah
I
didn't
realize
we
had
instructions
on
generating
javadoc
and
committing
it
to
the
repo.
I
think
we
should
never
do
that.
So
I
added
a
comment
to
that
effect,
generating
javadoc.
H
H
A
E
F
D
A
G
Hey
I've
almost
landed,
my
first
spec
pr.
H
G
All
right
anything
else,
anybody
wants
to
chat
about.
A
I
may
pick
up
since
honorary
if
you're
going
to
run
with
the
the
jrpc
context
removal
I
may
start
running
with
baggage.
A
E
A
I
mean
I
realize
there's
another
thing
in
the
spec
that
we
don't
have,
which
is
this
composite
propagator,
which
does
both
like
two
different
kinds
of
propagation,
so
like
baggage,
propagation
and
context
like
trace
propagation,
but
we
don't
have
a
thing
that
combines
those
two
together.
So
that'll
be
I'll
put
an
issue
in
tomorrow
for
that.
But
that's
another
thing:
we'll
need
for
ga
how's
that
different.
G
A
A
Yeah
I
actually
asked
the
question
in
the
spec
on
monday.
I
think,
but
there's
this
composite
propagator,
which
is
two
different
types
of
propagation
and
then
there's
the
multi-propagator,
which
is
multiple
propagations.
A
E
A
Yeah
though
it
doesn't
match
what
this
does,
the
spec
actually
has
two
different
id
concepts
for
this.
So
it's
all
very
confusing
to
me
to
be
totally
honest.
E
G
A
G
Ad
listenable
that
ad
the
guava
listenable
future
instrumentation
that
tyler
had
added,
and
then
he
had
added
a
test
for
it
in
the
cassandra
300.
But.
E
So
now
now
I
so
the
answer
to
my
question,
I
guess
was
that
we
added
listenable
future
instrumentation
and
that's
why
now
we
need
it
right.
Okay,.
B
G
G
Okay,
I
still
don't
think
it's
needed
in
production,
because
the
reiterating
this
is
like
when
we
go
through
at
startup.
We
go
through
the
existing
classes
that
have
already
been
loaded
and
we
instrument
them
and
that's
good
for,
like
like
java
tasks
like
run
a
couple
of
runnable
things
that
we
want
to
instrument
after
the
fact,
but
most
classes
we
don't,
but
in
during
tests,
there's
a
bunch
of
stuff
already
there
guava
in
particular.
E
G
Seems
like
it'll,
be
a
better
world
and
if
we
could,
if
we
yeah-
and
I
mean
I-
I
sort
of
hate
to
shade
at
the
same
time,
just
because
of
perform
like
performance,
but
man
that
would
that
would
also
catch
bugs
too.
So.
E
G
Oh
yeah,
that's
post,
post,
ga
yeah
is
the
yeah.
It's
nice
to
know
the
direction
we're
headed
because
then
we
don't
worry
quite
so
much
about
maybe
some
of
the
bumps
along
the
road,
because
we
know
it's
all
going
to
get.
G
Every
five
years
there's
a
new
telemetry
open
source
telemetry.