►
From YouTube: 2020-07-21 Java Auto-Instrumentation 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
B
B
C
Yeah,
I
am
having
all
kinds
of
problems.
Nikita
with
gradle
I
tried
to
run.
I
was
trying
to
run
those
those
jmh
tests
that
that
nikita
wrote
and
I
don't
know
something-
keeps
caching.
I
I've
deleted
now.
D
C
Hold.Gradle
folder,
I
rebooted,
I
don't
know
what
I'm
going
to
do
next.
A
D
C
B
C
You
took
part
you
did
the
the
issue
triage
in
the
java
repo
yesterday
for
you,
you
want
to
guide
us,
you
want
to
lead
us
here.
We
should
be.
A
So
I
have
created
the
new
labels,
this
following
specification
group,
so
the
they
are,
these
release
required
for
ga
release
after
ga
and
priority.
E
A
Priorities
is
very
simple:
p1
is
that
we
probably
do.
We
do
need
that
before
ga
and
it
is
more
important
than
p2
and
p3
is
wrap.
If
you
have
time,
probably
maybe
we
will
do
that
before
ga
p0
is
not
here
in
case
if
something
flow
flows
in
that
will
be
p0
or
something
like
cve
or
something,
and
so
what
andrew
did
with
java
java
group
yesterday,
just
we
can
split
the
issues
from
that
page
by
pay
page.
A
You
just
go
through
them
and
we
put
those
issues
before
gi
after
j
and
if
you
have
some
doubts
or
anything,
then
we
can
discuss
that.
But
the
idea
for
the
first
triage
is
just
to
put
issues
into
two
buckets
before
or
after.
C
Okay,
do
we
need
to
put
oh
I
so
we
put
after
so
that
we
know
that
we've
looked
at
it
so
that
if
it
has
neither
of
those,
we
still
need
to
triage
it?
A
C
And
then
we
mark
any
that
we
don't,
that
we're
unsure
of.
A
C
A
E
E
D
D
B
A
Confusing
I
don't
know
either
I
mean
I.
I
think
we
want
to
try
to
address
them
more
or
less
as
soon
as
possible,
because
they
are
actively.
A
A
C
C
I
haven't
marked
much
as
after
ga
is
this
a
problem.
A
A
C
C
Yeah,
that's
why
I
was
kind
of
like
the
the
priorities.
C
B
A
C
I
had
that
on
one
of
the
the
p1
having
that
or
at
least
benchmarks
yeah.
B
C
B
B
B
C
A
C
Did
you
already
finish
page
four
in
the
honorary.
B
A
A
B
C
C
A
Consider
tweaking
the
build
to
only
run
tests
for
affected
files
yeah.
So
do
you
have
any
idea
how
to
achieve
that.
B
Yeah
I
mean
in
other
builds
I've
done
it
where
you
just
look
at
the
get
diff
and
see
which
projects
are
affected.
It's
surprisingly,
not
so
complex
from
what
I've
seen.
Okay,
then,
I
believe.
C
Is
I
thought
that
that
was
going
to
be,
or
was
addressed
by
the
gradle
caching
stuff.
B
A
B
D
B
C
Although
I
guess
with
the
run
times
I
mean
nikita
with
the
new
runtimes,
maybe
it's
going
to
be
fast
enough
that
I
mean
if
it
only
takes
like
five
minutes
to
run
all
the
tests.
That's
I.
A
A
A
I
tend
to
like
them
it's
it's
kind
of
structure,
which
I
sometimes
think
it's
well
yeah.
We
we
try
to
predict
what
kind
of
issues
people
are
going
to
to
submit
and
we
try
to
help
them
with
this
trial
structure.
In
my
opinion,
the
best
github
is
the
best
issue.
Template
is
a
link
to
that
article.
How
to
ask
smart
questions,
read
that
and
then
submit
your
issue.
B
A
That
my
primary
use
case
for
templates
yeah,
that
is,
that
may
be
a
good
idea.
Yeah,
okay,
yeah
understood
yeah
68
issues
and
seems
like.
A
C
If
you
want
now
the
we
need
right,
we
need
some
kind
of
shaded
or
context
to
map,
because
we
can't
use
the
one
in
the
user's
class
path.
Why?
C
A
A
C
Now,
whether
we
you
know
how
we
store
it
and
that
kind
of
thing
is,
I
was
trying
the
reason
I
was
trying
to
run.
Those
jmh
tests
is
because
I
feel
like
something
is
wrong.
If
we're.
If
the
field
storage
is
expensive,
because
the
idea
of
the
field
storage
right
is
supposed
to
just
be
a
direct
store
into
a
field,
it
should
be.
B
C
Yeah,
it's
not
that!
That's
why
you
can.
You
can
only
use
that
instrumentation
context
inside
of
advice,
because
there's
a
byte
code,
instrumentation
of
that
code
that
replaces
instrumentation.get
and
then
those
two
classes
with
a
hard-coded
class
object.
That
then
does
the
field
direct
field,
mapping.
C
I
was
trying
to,
I
was
gonna,
run
the
jmh
test
and
then
also
dump
the
bite
as
a
flag
to
dump
the
byte
code,
so
that
then
I
could
look
inside
of
that
storage
benchmark
class
and
see
what
it
was
doing.
I
haven't
looked
super
carefully
closely
at
that,
so
I
definitely
could
be
missing
something.
A
D
D
A
A
Okay,
anyway,
so
these
certainly
should
be.
A
A
A
C
C
It
would
be
nice
for
it
to
not
lock
down
yeah.
I
think
yes,
I
think
this
is
required,
because
the
idea
is
for
it
not
to
bog
down
new
contributors
machine.
C
A
I
mean
I
can
I
kind
of
like
it,
especially
yeah.
I
can't
do
that.
No,
no
problem,
especially
considering
that
you
trust,
had
concerns
about
our
growing
build
bill.
B
C
A
B
A
Marked
that
it's
a
little
bit
too
much,
I
think,
but
I
think
from
now
on
from
this
we
have
two
things
to
do.
First,
is
all
new
issues
should
be
assigned
before
or
after
I
don't
know,
do
we
want
to
to
do
that?
A
A
A
A
C
Yeah
I
mean
so
one
of
my
thought
on
at
least
the
semantic
conventions.
The
semantic
attributes
was
that
we
would
only
implement-
and
I
guess
now,
they're
they're
doing
the
right
way
with
required
or
probably
doing
away
with
whether.
C
C
I
mean
there's
so
many
semantic
attributes
and
so
many
different
instrumentations
that
we
have.
B
C
A
B
A
Yeah,
probably
by
the
way
that
brings
me
to
a
question
that
migrating
from
all
decorators
to
my
new
tracers,
so
server
side
is
quite
easy.
It's
almost
done
so
I
think
I
will
do
that
completely,
but
we
have
a
lot
of
clients.
Client
decorators,
should
I
should
we
continue
like
rewriting
them
to
new
new
tracers.
A
So
is
this
not
that
important
for
for
for
me,
those
new
tracers
are
a
very
good
place
to
exactly
to
like
enforce
or
be
consistent
in
semantic
conventions
and
attributes,
and
and
that
so
like
they
are
much
more
much
better
suited
for
encapsulating
common
logic
than
current
decorators.
So
that's
why
I
kind
of
wanted
to
do
the
migration
server
side
server
spans,
but
clients
fans
were
a
lot.
B
C
The
cleanup
on
the
server
spans
was
really
nice.
I
mean
it
ended
up
cleaning
up
a
lot
of
stuff
producing
a
lot
of
duplication.
C
C
Let's
see,
there's
stuff
I
mean,
if
I
can
also
like
helen
hayams
is
joined,
is
starting
to
work
in
my
team
and
so
there's
stuff.
I
can
have
her
start
doing
some
work
also
and
help
her
with
that.
C
If
we
think
that
that
you
know
do
one
of
the
I
mean,
especially
if
you
did
like
one
of
them,
then
that
would
be
something
that
she
could
take
to
other
client
instrumentation.
A
B
The
tracers
and
the
decorators
they're
both
part
of
our
api
for
instrumentation
right.
Yes,
so
in
that
sense,
if
we
don't
like
the
decorators
much,
if
we
don't
get
rid
of
them
before
j,
we
do
still
have
to
commit
to
sort
of
maintaining
them
for
a
while.
So
that's
one
reason:
if
we
don't
like
them,
we
want
to
get
rid
of
them
sooner
than
later.
A
A
Then
I
I
think
what
I
will
do
is
I
will
write
a
small,
not
designed
document,
but
not
something
similar
to
that
to
explain
my
idea,
why
tracers
not
decorators,
what
difference
between
them
and
water
like
guidelines?
If
you
want
to
write
new
tracers,
then
it
will
be
easier
to
spill
that
work
on
other
contributors.
A
C
C
A
A
C
Yeah,
I
just
went,
does
the
do
we
want?
Can
we
use
that
to
help
us
drive
our
decision,
making
like
having
a
consistent
decision-making
process.
A
A
C
Cool
I'll
put
in
tomorrow
your
the
feedback
that
you
both
gave
and
then
I'll
change
it
from
p0
starting
at
p0
to
starting
at
p1
and
then
should
I
combine,
I
mean,
should
I
say,
p1
p2
would
be
all
the
contributor
experience
and
then
combine
p3,
p4
and
p5
all
into
p3.
A
Cool
yeah
and
then
probably
after
this
document
is
merged,
we
really
can
assign
priorities
to
issues
in
like
uncoordinated
fashion
like
we
do
that
on
offline
and
then,
if
we,
if
we
will
be
far
enough,
then
next
week
we
can
meet
again
to
just
review
our
priorities
and
or
discuss
any
conflict
in
our
opinions.
C
Good
to
me,
do
you
think
that
I
should
remove
the
road
map
part
or
replace
the
roadmap
by
this
or
have
both.
A
A
B
C
I
will
yeah
I'll
probably
bring
it
up
in
the
thursday
sig,
but
then
I
need
to
wait
for
john.
C
C
Okay,
I
don't
want
to
make
any
decisions
without
him,
the
the
snapshot
repo
stuff
did
that
make
sense.
Does
that
sound.
C
A
C
B
C
A
B
A
Yeah,
so
I
I
propose
this
plan,
so
I
will.
A
I
will
do
that
later,
but
I
will
do
that
because
during
the
experiment
I
have
fixed
several
potential
flaky
mess
sources,
so
I
will
do
the
probably
separate,
oh
in
this
puller,
because
I
don't
know
yet,
but
mostly
I
will
leave
it.
It
probably
is
a
little
bit
flakier,
but
I
promise
I
will
address
that
later.
B
C
Going
to
include
the
the
before
class-
yes,
okay,
awesome.
C
All
right:
well,
you
all
have
good
days.
I'm
gonna
turn
in
here.