►
From YouTube: 2021-08-24 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
A
Materials,
do
you
know
if
nikita
is
back
this
week.
A
A
The
okay
http
thing
for
sure,
since
it's
a
regression
I
figure
I
should
pull
in
the
at
least
the
fixing
of
the
publishing
for
the
agent
for
testing.
A
If
you
think
of
anything
else
today,
let
me
know
I
can
go
ahead
and
churn
the
release.
The
patch
release
out
tomorrow.
A
A
Yeah,
I
mean
technically,
this,
I
don't
know
was
the
agent
for
testing
might
not
have
been
broken.
A
A
Broken
packaging
palm
yeah,
so
technically
this
one
is
not
a
regression,
but
we
had
fixed
it
specifically
in
we
thought
we
were
fixing
it
specifically
in
one
five.
A
B
A
A
A
God,
oh
my
god,
I
always
feel
so
so
unlucky
with
everything
I
touched
with
shadow
jar.
I
could
not.
I
literally
could
not
believe
that,
like
let
me
try
deleting
these
lines
of
code
that
worked
yeah
credit
for
that
I'm
sure
goes
to
mateo.
Something
matish
did
that
he
doesn't
know
he
did
when
he
refactored
that
agent
for
testing.
C
C
B
B
I
think
its
biggest
problem
seems
to
be
that
it
just
adds
stuff
to
your
java
publishing
config
automatically
with
no
flag
like
if
that
was
up
it's
hard
to
remove
stuff
from
publishing.
That's
been
added
like
riddles.
Aprs
aren't
very
good
for
that,
and
so
once
it
does
its
thing,
it's
very
hard.
If
that's
not
the
behavior,
you
want
that's
the
biggest
problem.
A
A
Oh
this
one,
I
figure,
I'm
sure
materia,
should
take
a
look
at
this
today.
A
Oh
yeah,
we
talked
about
this
last
thursday
in
the
sig,
so
I
thought
about
it
a
little
bit
and
the
the
instrumentation
I
didn't
feel
like
the
instrumentation
type
worked
out
that
great,
partly
because
there's
no.
We
have
that
for
the
instrumentation
type.
We
have
that
nice,
auto
detection
based
on
the
attribute
extractors.
A
A
I
wasn't
sure
what
you
thought
about
the
no
op,
like
I
kind
of
wished,
although
it
worked
ended
up,
especially
with
this,
it
ended
up
working
out
pretty
nicely
just
calling
that
set
disabled
as
opposed
to
having
to
have
a
conditional
and
then
constructing
a
no
op.
Instead,.
A
We'll
see
there
were
two
more
sporadic
failures
today,
one
in
and
one
in
elasticsearch,
so.
B
A
Yeah
but
things
that
we
maybe
like,
I
know
they.
I
think
that
the
data
dog
repo
ended
up
listing
or
restricting
more
which
rentables
would
be
instrumented.
A
Well,
like
I
think
we
have,
we
have
a
sort
of
deny
list
of
runables
that
we
don't
instrument.
So
I
mean
we
could
you
know,
do
something
like
that
and
keep
you
know,
whack-a-moling
them.
A
A
A
B
B
B
A
I
directed
somebody
from
app
dynamic,
oh
yeah
matt.
He
had
joined
a
couple
of
the
thursday
sigs.
I
directed
like
keep
directing
more
people.
B
B
B
Yeah,
you
could
all
extend
from
that
fits
into
that
sort
of
direction
of
the
instrumentation
api,
otep
and
whatnot,
also
where
somehow
we
have
to
make
configurations
sort
of
consistent
across
instrumentation.
First,
eventually,
we
also
think
about
across
languages
and
using
some
file.
But
for
now
I
think,
having
this
sort
of
builder
is
important.
Just
for
within
our
code
base.
A
Yeah
yeah.
That
sounds
good,
oh
yeah!
I
had
thrown
this
here
because
I
didn't
quite
understand
what,
but
it
sounded
cool
the
idea
of
out
of
test
suite
for
the
semantic
conventions
implemented
out
of
process.
B
Yeah
I
mean
mostly
about
http.
I
think
I
don't
know
if
it
applies
to
other
libraries,
but
maybe
it
could
depending
how
it's
implemented,
but
I
mean
we
have
http
server
test
https
those
are
building
the
test,
server
or
client
whatever
inside
java,
but
it
doesn't
have
to
be
right
to
be
out
of
process,
in
which
case,
if
the
spans
are
also
exported
to
some
other
process
collector,
then
those
could
be
validated
in
a
standard
way
not
having
to
do
it.
Language
just.
A
B
A
I
remember
when
we
were
when
I
was
initially
evaluating
the
datadog
repo,
and
that
was
actually
one
of
the
I'm
pretty
sure
I
I
think
I
ever
could
probably
find
the
commute.
The
open,
telemetry
issue,
where
I
put
my
comments
and
I'm
pretty
sure
that
was
one
of
my
comments
that
I
liked
the
the
those
reusable
http,
client
and
server
within.
B
The
code
base,
but
yeah
we
can
even
go
further
than
that
taking
yeah,
I
mean
I
think
our
tests
are
a
good
inspiration
for
that,
because
it
would
be
sort
of
just
assuming
that
we're
following
this
meant
just
correctly
but
yeah.
He
would
literally
just
be
extracting
that
code
into
some
separate
process.
A
Else,
how
are
things
going
over
in
the
sdk
repo
in
a
lot
of
good
good,
a
lot
of
activity.
B
B
B
Like
the
product
of
dependency
and
the
generated
produce
dependency
like
even
our
agent,
jar
is
probably
going
to
get
quite
a
bit
smaller
next
place.
Yeah
yeah.