►
From YouTube: 2021-10-27 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).
C
A
A
D
Android-
and
was
there
anything
else.
A
This
one
john
was
saying:
todas
was
not
urgent,
that
they
could
stay
on
1-6
or
he
said
not
to
push
out
something
quickly
for
that.
But
this
one
was
a
real.
We
we
couldn't
say
this.
One
was
not
oh,
I
guess
we
could
say.
Okay,
dp
is
alpha,
but
I
don't
know.
A
A
A
Fix
for
that
one,
yet
right
we
have
a
fix
for,
and
oh
this
one
honorable
convinced
me
was
not
was
unlikely
and
I
I
agree.
It's
unlikely
to
cause
problems
yeah,
but
we've.
D
Yeah,
it
should
be
a
pretty
quick
fix.
I
mean
it's
just
bumping
the
version,
if
you
don't
count,
adding
any
tests
for
that,
of
course,
because
we
have
no
tests,
no
smoke
tests
for
metrics.
I
believe.
A
No,
I
think
you're
right
yeah,
I
I
I
would
vote
for
that.
Quick
patch
version.
A
Cool,
do
you
mind
putting
that
in
today
matters
and
I
will
kick
the
I
can
take
care
of
this.
A
And
then
we
will
did
this,
even
though
I
think
I
would
like
to
discuss
alternatives
long-term.
A
Cool,
I
think
this
patch
release
I
actually
kind
of
want
to
have
a
redo
on
the
on
the
patch
release.
Since
my
last
patch
release
before
this
was
so
not
good,
so
yeah,
I
think
we're
all
good
now
I
think
we
have
well
we'll
see.
A
A
Oh
did
you?
Yes,
you
already
were
just
all
right:
cool.
B
A
C
A
C
A
A
C
A
What
would
it
look
like?
I
mean
from
what
are
we
getting
from?
It
is
it
you
know,
I
think,
we're
just
using
it
for
pretty
simple
cash.
A
D
C
C
For
example,
we
have
the
statement
cache
and
I
think
we
have
some
of
these
caches
that,
like
they're
always
looked
up
on
the
hot
pack,
and
I
guess
the
question
is:
if
it's,
if
we
ever
hit
the
bound
on
our
caches,
if
we
don't
hit
that
down,
then
it's
the
same
as
a
concurrent
hashmap.
Basically,
and
so
maybe
one
thing
is
just
not
too
bounding-
I
don't
know
if
that's
safe,
but
we
probably
generally
want
to
cash
more
than
like
cash.
Everything
this
and
then
the
bound
is
probably
more
of
a
safety
measure.
A
C
A
Meaning
like
it,
it
predicts
best
what
the
least
recently
used
is
at.
C
C
C
As
if,
if
it
works
like
I,
we
might
want
to
split
out
our
projects
in
a
way
that
that
shaded
caffeine
has
its
own
animals
different
like
make
sure
that
doesn't
get
applied
to
the
whole
instrumentation.
Okay,
that's
definitely
a
problem
but
like
if
our
usage
of
caffeine
works,
I'm
still
sort
of
comfortable
with
using
it
with
those
exclusions.
As
long
as
those
exclusions
aren't
leaking
into
our
other
code
base,.
C
A
What
do
you
think
if
we
do
you
think
it
if
we
pro-guarded
caffeine,
that
it
might
we
might
get
lucky
and
if
we're
not
and
strip
out
its
usages
of
these
things,.
C
A
And
materia
you
were
saying
earlier,
I
think
something
about
still
using
caffeine
in
the
agent
potentially
as
the
backing.
D
So
we
currently
use
that
streak
when
we
don't
include
caffeine-free
instrumentation
api,
but
we
do
that
in
in
the
agent
and
we
put
it
somewhere
in
the
bootstrap
and
we
catch
a
note
has
the
founder.
So
if,
if
you
try
to
create
the
cash
and
caffeine-free
it's
not
on
class
path,
then
it's
okay.
It
will
work
without
that.
A
D
So
another
option
would
be
using
either
reflection
to
dynamically
get
the
class,
and
that
would
probably
that
wouldn't
need
any
even
compiled
reference
to
caffe
or
I
don't
know
some
kind
of
sbi.
But.
A
B
D
Problem
and
we
needed
in
needed
it
on
the
bootstrap
cluster
because
of
instrumentation
bi,
which
is
also
another
surprise.
Over.
A
D
A
B
C
D
Yeah
but
remember
that
the
last
time
I
mean
this
issue
actually
was
about
method
handle
being
used
in
caffeine,
free
and.
D
D
C
B
C
C
A
Oh
well,
this
here
is
before
laurie
doesn't
incorporate
laurie's
change
urged
laurie,
I
think.
Maybe,
if
you
just
it.
D
C
D
C
Sex
food
is
only
applying
to
caffeine
three,
so
it
seems
like
animal
sniffer
is
doing
its
job
you're
right
and
for
some
like
we
just
have
to
tweak
it
so
that
this
caffeine,
through
that's,
not
supposed
to
be
sniffed
anyways
just
doesn't
because
we're
not
including
it.
So
we
definitely
are
supposed
to
exclude
that
from
animal
sniffer.
I
think.
D
A
D
C
C
A
A
A
D
D
C
We
can't
exclude
these
classes
for
sure
I
think
like
we
could
potentially
exclude
caffeine
too,
which
has
transmitted
like
it
feels
a
bit
bad.
But
since
we
know
caffeine,
2
won't
be
released,
it's
also
sort
of
okay,
but-
and
maybe,
if
that's
like,
if
it's
just
a
one
line
and
then
it
works,
maybe
that's
still
an
okay
approach,
but
it
is
a
bit
fragile.
That's
true!.
A
A
Example,
if
we
really
thought
we
wanted
that
performance
on
the
on
the
bounded
path,.
C
Think
I'm
more
thinking
that
to
switch
to
a
different
like
concurrently
attached
map
or
something
it's
a
reasonable
work,
and
if
we
know
that
the
current
thing
does
actually
work
on
android.
Maybe
it's
not
worth
putting
in
time
to
do
a
big
change
if
we
can
just
figure
out
how
to
get
rid
of
these
ignored
classes,.
A
I
mean
if
we've
already,
there's
got
to
be
a
there's.
No
other
fallback
libraries
like
what
was
we
had
the
one
that
we
had
inherited
from
the
data.
C
A
C
A
C
D
A
Splitting
it
out
into
a
separate
artifact,
if
we're
not,
is
that
it
feel
worse,
because
I'm
wondering
what
we
even
use
that
cash
for,
like
you
were
saying,
matesh
in
okhtp
or
probably
oh,
but
we
do
do.
We
do
use
it
in
some
like
spam
names
cache
even
in
the
instrumentation
api
directly.
So.
D
Statement,
sanitizer,
http
header
attributes
and
that's
pretty
much
it
for
library,
stuff,
okay,.
C
E
B
C
E
Yeah,
it's
good.
I
haven't
worked
on
it
all
today.
I
did
a
bit
of
I
replied
to
some
comments
on
the
pr
yesterday,
but
yeah
it's
up
to
other
things
today,
but
yeah
it's
going.
Well,
I
think
it's
it's
getting
there.
There
was
a
bit
of
discussion
around
whether
to
use
into
along
for
the
the
indexes,
but
I
think,
probably
use
int.
So
I've
just
got
to
go
ahead
and
change
that,
but
yeah
nothing
too
new.
B
E
E
E
Stuff,
as
you
as
you
talk
as
well
as
just
working
on
stuff,
yeah,.
C
Like
if
we
published
artifacts
with
the
version
like
made
or
experimental,
then
then
it
makes
a
ton
of
sense,
and
maybe
we
should-
and
I
wasn't
sure
if
this
table
is
the
current
state
or
the
future
state
that
we
want
but
like.
I
would
argue
that
the
current
state
is
only
two,
at
least
on
the
java
side.
We
have
stable
and
alpha
and
we
don't
really
think
in
these
terms.
Right
now,
that's
my
opinion.
I
think.
A
To
me
yeah,
I
I
feel
like
beta
maps
best
fret
in
our
world,
like
experimental
feels,
like
I
don't
know,
just
trying
something
like
almost
proof
of
concept,
but
I'm
not
sure
what
what's
the
intended
or
do
they.
D
A
I
did
go
through
and
kind
of
just
does
count
the
what
find
find
grab
the
number
of
modules.
I
have
a
lot
of
modules
in
our
repo.
A
It
was
cool,
though,
like
I
mean
I
know
this
is
not
for
logging.
It's
not
the
logging
signal,
but
definitely
the
the
mdc
injection
stuff
is
part
of
the
logging
spec.
A
C
Sort
of
wondering
what
a
good
direction
for
that
is
one
thing
to
keep
in
mind
is
that
for
most
languages
I
think
the
sort
of
documents
will
just
be
in
the
open
company
operator
repo,
for
example,
node.js
or
python,
because
there's
no
like
normally
a
package
goes
into
your
app
if
you're
using
those
languages
and
reference
them
java
is
a
bit
unique
in
that
there's.
The
separate
binary
thing,
that's
totally
unrelated
to
an
app,
and
so
there
is
some
value
in
having
the
sort
of
base
image,
whereas
it
wouldn't
be
for
the
other
languages.
C
C
A
C
A
It
kind
of
makes
it
makes
sense
to
me
I
mean
since
most
of
the
then
it
can
be
potentially
more
consistent
with
the
other
languages.
Yeah.
C
E
I
had
a
quick
question
about
some
of
the
instrumentation
in
the
instrumentation
repo
seems
like
for
stuff
using
the
sdk
like
instrumenting
the
sdk.
It
has
kind
of
attributes
like
this,
like
aws
dot
operation,
I
think,
are
those
just
like
aws
like
endpoint
aws,
stop
agent
stuff
like
that.
C
E
Yeah,
I
I
read
the
library,
it's
kind
of
you,
yeah
I'd
like
an
interceptor
right.
C
E
C
This
yeah,
and
so
there
might
be
some
confusion,
because
some
attributes
are
properly
sort
population
for
legacy
like
those
are
ad
hoc
named
and
whatnot,
and
they
used
to
be
populated
but
they're
supposed
to
be
replaced
by
the
names
in
this
spec.
I
don't
know.
B
C
E
B
C
C
E
Yeah
cool
because
we
just
like
before
we
have
our
own
kind
of
like
convention
internally
for
metrics,
because
we
just
used
it
before
there
was
a
metric
semantic
convention
and
we
have
one
for
storage,
which
is
kind
of
like
basically
just
s3.
But
really
I
was
thinking.
Is
there
really
a
necessary
like
convention
for
storage?
It
could
probably
just
be
under
something
else,
but
yeah
that
was
just
around
that
yeah.
So.
A
D
I
think
that
the
worst
offender
here
is
camel,
which
is
so
complex,
that
rewriting
it
to
instrumental
will
be
painful.
A
Any
chance
we
can
get
jack
up,
jake
jacob,
I
doubt
but
I'll
try
asking
him.
No,
no,
no,
that
that
would
be
mean
the
instrument.
C
E
Yeah-
and
I
think
I
think
there
was
some
talk
about
in
the
semantic
convention
having
a
certain
place
for
like
things
that
extend
the
semantic
convention
almost
like,
like
the
aws
sdk,
I
suppose,
would
be
an
example
kind
of
structured
in
there
somehow.
So
we
were
discussing
that
instrumentation
meeting
a
few
weeks
ago.
Yeah,
that's
that's
another
discussion,
yeah!
A
Like
a
rabbitmq,
like
all
these
things
that
they
mostly
follow,
say
you
know
rpc
or
messaging,
but
they
also
all
have
some.
You
know
a
few
extra
bells
and
whistles
or
things
curious
to
see
if
this
becomes
like
the
contrib
repo
of
specs.
E
Yeah,
that
was
one
of
my
worries
actually
because
like
if
you
look
at
the
database
convention
currently
like
there
is
like
already
so
many
different
database
things
going
on.
Like
you
know,
cassandra
has
a
couple
of
extra
attributes
like
redis
has
a
couple
of
ones
and
it's
kind
of
like
well,
let's
just
separate
that
out
and
have
like
a
core
database
convention
and
then
other
things
can
extend.
It
could
maybe
keep
it
a
bit
more
sensible.
E
I'm
hoping
I'm
hoping
that
something
like
that
will
get
get
done,
because
otherwise
yeah
it'll
be
a
bit
chaotic,
like
you
say,.
E
D
E
A
D
A
All
right,
folks,
any
anything
else
before.
Let
you
all
get
on
with
your
days.