►
From YouTube: 2019-10-09 JavaScript 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).
B
B
A
E
G
G
C
Cool
I
think
we
should
get
started,
looks
like
we
have
enough
people,
so
the
first
agenda
is
about
the
release
that
we
did
yesterday,
so
we
release
0
dot
or
dot
zero
dot,
one
dot,
oh
version,
and
it's
available
now
to
use
I
mean
thanks
to
everyone.
This
is
really
awesome.
I
mean
kudos
to
everyone
which
is
really
awesome
job
there.
C
Mean
everyone
else
did
really
awesome
job
that
I
know
so
close
was
yeah
cool
and
then
then
I
have
added
link
for
the
next
milestone,
so
have
couple
of
alpha
v3
v4
milestones
to
achieve.
So
probably
we
can
talk
about
that
here
or
sometime
later,
I,
don't
know,
but
you
items
have
already
mentioned
here
and
then
we
can
talk
on
that
sometime.
This.
I
E
E
E
This
is
the
front.
Mayor
was
asking
about
browser
j/s,
and
my
question,
for
the
call
is
one
to
raise
awareness
about
the
work
that
just
got
committed
and
two
we
were
discussing
internally.
What
would
be
good
next
steps
for,
like
a
browser
backlog
I,
wanted
to
actually
pose
that
to
this
group
to
see,
if
you
all
had
ideas
about
what
you
know
see
if
we
could
get
a
backlog
going
through
the
browser
work,
as
opposed
to
today
as
the
nodejs
work.
Yes,.
J
So
basically,
the
next
step
is
the
document
load
and
starts
to
around
it
taken
from
the
performance
as
once
it's
done.
The
next
step
would
be
to
create
this
new
scope.
Manager,
which
will
eat
Azzam
scope
manager,
would
support
the
a
sink,
and
once
we
have
this
one,
then
we
can
start
instrumenting
things
like
xhr
I
mean
anything
that
is
basically
to
a
sink,
because
so
far
we
have
this
scope.
J
Manager
that
is
basically
like
keeping
the
reference
to
the
last
used,
but
it
won't
work
correctly
with
I
think
why
and
yeah.
So
that
is
basically
like
a
basic
stuff
that
is
missing
for
the
web,
and
once
we
have
this,
then
it
will
be
just
a
matter
of
adding
more
and
more
plugins
and
I.
Think
that
yesterday
just
proposed
that
the
next
thing
after
this
could
be,
they
react
and
start
see
what
we
can
get
from
the
react
framework
and
how
to
instrument
start
back
again.
E
J
E
And
I
wanted
to
check
in
it
where
I
know.
I
think
react
is
an
interesting
project,
but
I
was
wondering
if
there
was
more
just
like
more
basic
web
information,
we
could
be
getting
out
of
it.
100%
I'm,
not
totally.
It's
not
totally
clear
to
me
what's
already
baked
into
the
the
web.
Jay
s
frame,
work
you've
got
to
do
know,
there's
like
basic
stuff,
around
x8
xhr
timing
and
like
download
page
loose
stuff
like
that,
whether,
like
you
get
that.
J
And
basically,
what
exactly
do
we
want
to
collect?
Because
I
mean
the
document
load
is
just
like
the
way
I'm
implementing
this
one
is
really
doesn't
matter
whether
you
load
to
this
plug-in
after
the
page
was
loaded
before,
because
the
performance
is
already
taking
care
of
collecting
all
those
data.
So
it's
a
matter
of
either
hooking
up
and
checking
for
the
ready
States.
So
when
the
Dom
is
loaded,
then
you
simply
collect
information
from
the
performance.
J
If
not,
then
listen
for
the
events
until
tom
slow
that
intend
to
the
same
and
and
I'm
all
most
started
and
just
download
like,
for
example,
the
TLS
when
I
started
some
DNS
negotiation
and
and
stuff
like
this
re
assumed.
We
would
be
interesting
in
extracting
what
is
possible
how
this
will
be
presented
later.
It's
a
it's
another
question,
but
I
think
that
it's
going
with
to
collect
whatever
is
possible.
I
think
it
should
be
the
way
to
go
right.
Yeah.
E
It
seems
like
this
is
like
an
issue:
that's
less
less
obvious
or
less
important
on
server-side
stuff
right,
because
on
server-side
you
know
these
processes
tend
to
be
longer
running
when
it
comes
to
the
browser,
the
users
booting
it
up
and
there's
like
yeah
this
whole
suite
of
information.
You
might
want
to
capture
from
the
browser
around
that
initial
page
load
at
the
same
time
we're
talking
about
like
information
people
who
want
to
capture
in
production.
Obviously,
if
it's
just
development,
you've
got
all
of
like
the
native
browser
tools.
You
know
at
your
fingertips.
E
We
don't
need
to
recreate.
You
know
that
world
is
but
there's
a
question
of
like
okay,
like
for
production
usage,
you'd,
probably
want
something
more
like
open,
telemetry
feeding
you
back
this
information,
so
anyways
I
wanted
to
raise
it
as
an
interesting
like
question,
because
I
think
it's
there's
like
some
bigger
questions
than
necessarily
just
like
throwing
some
features
into
backlog
and
knocking
them
out.
E
So
I
don't
know
if
people
there
are
other
people
on
the
call
who
have
like
a
browser
background.
If
you're
interested
in
this,
this
problem,
I
just
kind
of
wouldn't
take
the
temperature
of
the
room
or
see
if
we
needed
to
get
more
like
browser,
focus
people
kind
of
involved
on
the
browser
side,
stuff.
J
And
one
of
the
key
tiles
or
I
see
yesterday
was
that
if
we
might
be
interested
about
collecting
the
stats
around
the
video,
because
basically
the
video
is
really
also
interesting
subject,
because
first,
it
can
be
run
only
in
the
browser.
But
then
there
is
really
a
lot
of
analytics
connected
with
photo.
Video
and
many
different
video
players
and
I
have
some
experience
with
this
one.
It
was
really
like
a
lot
of
work
to
be
done
to
integrate
any
analytics
and
letting.
J
Mean
not
only
what
I
know
know
I'm
talking
about
like
VOD,
II
or
DRM,
protected
content
and
stuff
like
this
yeah,
but
WebRTC
can
be
also
there's
a
nice
subject
for
that,
but
it
was
just
about
the
video
yeah,
because
the
market
might
be
really
really
big.
So
for
that
I'm
having
something
that
simply
works
almost
like
out
of
box.
That
would
be
nice
addition
something
to
consider
in
the
future.
So.
E
C
Go
ahead,
mayor
yeah,
so
actually
we
have
one
browser
expert.
His
name
is
Dave.
It's
gonna
join
in
few
minutes.
His
lady
there's
some
other
meetings.
He
did
really
awesome
job
on
the
open
census
website,
so
we
are
trying
to
put
a
bunch
of
code
from
there
as
well
and
he
is
working
with
the
bar
to
get
this
in
good
shape.
That.
E
That's
great
yeah
I
don't
want
to
distract
from
the
nodejs
work.
That's
also
important,
so
I
just
wanted
to
see
if
there
were
browser
people
who
either
as
part
of
this
call
or
we
could
have
another
call,
you
know
get
like
a
browser
backlog
together,
because
otherwise
we
can
just
kind
of
start
doing
things
ad
hoc.
But
if
there's
like
someone
in
this
group
who
has
a
plan
or
an
idea
I'd
like
to
respect
that?
Okay,
that's.
L
L
E
A
great
question
I'm
happy
to
field
that
one
yet
so
the
the
one
hang-up
with
the
point
to
release
of
the
speck
right
now
is
metrics.
That's
sort
of
like
the
most
important
thing
in
there
and
I
know
that
Bogdan
and
Josh
are
in
the
final
stage
of
like
hashing
out
the
final
details
between
each
other,
so
I
believe
josh
is
going
to
have
an
update
to
that
PR.
C
E
We're
the
goal
with
the
specs
is
I
mean,
and
we
don't
want
too
much
structure
here,
because
hopefully
we're
gonna
get
big
stable
soon
right,
okay,
thanks,
but
we're
gonna,
try
been
trying
to
say
as
soon
as
metrics
is
settled,
we're
going
to
have
like
a
snapshot
of
the
spec
called
point
two.
So
there's
at
least
like
some
stable
target
for
people
to
hit.
E
But
we've
been
saying
like
that
would
include
metrics
cuz,
that's
the
one
thing:
that's
actually
causing
the
SIG's
right
now
to
hold
off
like
context,
propagation
and
everything
else.
People
would
like
something
improved,
but
there's
like
nuts
hoppin
anybody,
but
metrics
is
the
one
thing
where
everyone's
like.
Could
you
please.
E
E
There's
a
point
three
release
expected
after
that,
and
that
basically
contains
the
rest
of
the
issues
that
we
don't
think
will
be
done
in
time
for
this
metrics
API
to
be
done.
I
think
that's
mostly
around
context
propagation
you're,
trying
to
fully
separate
out
this
like
context,
propagation
and
baggage
subjects,
there's
some
stuff
around
like
global
in
it
like
what
should
actually
be
the
like
order,
the
bootstrapping
order
when
these
programs
initialize
you
have
some
issue
there
around
like
what
happens
the
beginning,
but
I,
don't
think
it's
too
many
issues
beyond
that.
E
Yeah
I
think
I
know
so
4.3
the
stuff.
We
know
that
people
have
already
voiced
wanting
to
like
discuss
our
global
initialization
finalizing
context
propagation.
Getting
the
official
opens
lemon
tree
protocol
signed
up
on
I'd
like
the
data
protocol
and
then
the
semantics
conventions
feeling
like
we've
done
like
a
solid
pass
at
those
and
as
far
as
I'm
aware,
that's
like
it
as
far
as
spec
work,
I'm
sure,
there's
gonna
be
more
spec
work,
but
that's
that's
the
only
major
stuff
people
really
queued
up.
C
J
F
Yeah
Bart
I
I
just
made
a
it,
took
me
a
little
while
to
get
back,
but
I
just
made
a
comment
today
on
that
issue
basically
suggests
my
I
think
my
main
suggestion
would
be
for
the
browser
to
see
if
we
can
avoid
any
kind
of
hooking
into
the
module
loading
system
and
just
maybe
trying
to
use
the
plugins
as
something
we
import
and
past
the
tracer.
But
you
take
a
look
at.
J
J
G
E
C
H
C
A
H
For
the
exponent
just
to
know
what
left
like,
how
the
export
hooks
and
the
whole
exporter
like
interface,
will
look
like,
but
I
can,
as
its
it's
really
very
much
open
senses.
What
we
have
here,
I
can
just
go
ahead
and
port
it
over
I
started
it
anyways
already,
and
then
we
will
just
change
what
needs
to
be
changed
after
that.
Yeah.
C
G
M
C
The
next
one
is
I
was
thinking
to
add
few
dogs,
for
example,
to
let
people
create
their
own
exporter
or
propagator
SDK
that
sort
of
work.
So
we
already
have
one
doc
for
adding
the
support
for
the
new
plugin,
like
what
I'll
step
required
to
follow
in
order
to
add
support
for
the
plug-in
same
sort
of
work,
I
was
thinking
to
do
for
the
photos
and
SDKs.
D
Think
it's
good
I!
Guess
one
question
or
I
guess
one
thing
food
for
thought
is
like
at
what
point
do
you
want
to
like
split
these
out
into
like
an
actual
blog
post
type
saying
that
we
can
post
on
medium
or
something
along
those
lines?
I
started
getting
more
engagement
rather
than
this
being
sort
of
like
buried
inside
of
a
github
repo,
okay,.
D
I
remember:
we
were
doing
open
tracing
a
while
ago.
I
like
did
a
medium
article
that
was
like
open
tracing
in
15
minutes
and
I.
Just
still
think
that
it's
like
a
huge
amount
of
traffic
and
so
I
think
that,
like
this
kind
of
instructional
stuff
could
probably
like
be
put
on
the
repo.
But
then
it
should
also
be
split
out
and
put
on
the
blog
yeah.
E
Please
do
the
blog
also
the
website.
It
was
like
a
nascent
effort
to
start
adding
navigation
and
stuff
to
the
website.
I
haven't
been
bothering
the
Simms
to
contribute
to
that,
but
yeah.
You
know
getting
it
up
in
markdown
in
any
form
and
it
can
go
on
the
blog
and
then
get
ported
into
like
greenfield
documentation
on
the
website.
I
think
would
be
fab.
This.
F
E
D
F
H
E
Yeah
I
think
that
will
help
we've
been
trying
to
I've,
been
personally
trying
to
wait
for
this
metric
stuff
to
get
into
the
spec.
So
we
can
say
like
point
to
the
spec
is
out
and
then
I
was
going
to
start
bothering
people,
but
if
you
all
want
to
start
writing
blog
posts
and
start,
you
know
getting
documentation
together.
I
think
that
would
be
great
yeah.
E
C
D
The
specification
is
oddly,
is,
is
an,
albeit
is
very
Java
specific
and
the
way
that's
there.
The
specification
approach
to
name
tracers
is
that
you
have
some
sort
of
factory
that
you
can
pass
a
name
in
version
two,
but
factories
feel
kind
of
gross
and
I'm
curious
like
if
there's
a
better
way
to
do
that
in
JavaScript,
where,
like
what's
considered
JavaScript
bagless
the
practices
around
it.
C
D
So
understanding
what
the
name
tracers
work
is
that
let's
say
that
you're,
an
instrumentation
library
and
you
write,
say
inspiration
for
say,
like
I,
don't
know
MongoDB
what
you
want
to
be
able
to
do
like
semantically.
What
you
traditionally
do
is
that
when
anything,
any
poets
advanced
into
the
MongoDB
codebase
is
like
you
said,
attack
or
it's
like
component
MongoDB.
That
way,
you
can
see
that
these
fans
are
coming
from.
Like
a
library.
The
idea
behind
name
tracer
is
is
a
sort
of
like,
rather
than
it
being
a
tag,
it's
more
like
explicit.
D
H
Coming
basically
from
dynaTrace,
so
that's
your
initiative,
because
we
we
know
that
sometimes
you
just
need
to
switch
off
like
a
sensor
more
or
less
or
like
the
library
support,
because
something
is
wrong
there,
and
this
is
a
part
of
this
ability
topic.
So
you
need
some
way
to
provide
this,
which
kind
of
what
you
are
you?
What
I,
what
you're
currently
tracing
yeah,
what
it
doesn't
come
from.
D
D
So
the
way
that
I
currently
have
it
set
up
is,
rather
than
in
at
Global
tracer
you
sort
of
in
it
Global
tracer
Factory.
Let
them
be
like
open
or
open
till
March.
We
don't
get
tracer
that
still
works
because
part
of
the
so
part
of
the
specification
states
that,
like
in
the
event
that,
like
name,
is
null
or
like
empty,
you
can
provide
a
default
tracer
implementation
and
so
I.
D
C
C
So
the
one
more
item,
I
added,
is
for
the
benchmarking
framework.
So
one
last
week,
I
had
one
PR
to
support
this
page
mocking
this
case.
So
I
got
a
bunch
of
comments
there,
but
haven't
had
chance
to
revisit
that.
So
the
basic
idea
is
these
frameworks
will
help
us
to
run
some
samples
against
our
API.
C
So,
for
example,
the
tracer
API
with
a
start
span
and
bunch
of
other
parameters,
then
another
example
I
added
for
the
propagator,
so
it
will
run
a
bunch
of
samples
again
against
the
all
three
sort
of
propagator
that
we
have
binary.
Http
takes
and
then
b3
format,
and
it
will
give
you
some
numbers
like
how
much
time
it
sticks
with
these
kind
of
samples
and
all
so
do
you
guys
I
mean
I'm
kind
of
looking
for
the
options
like
is
it?
C
H
C
Do
you
want
to
run
this
every
pull
request,
or
once
in
a
while,
or
how
do
you
want
to
consider
this?
How
long
does
it
run
to
another?
So
it's
depends
the
samples
so
the
more
samples
you
provide.
It
will
take
longer
time
if
by
default,
I
added
five
samples,
so
it
it
will
not
take
more
than
made
or
to
that.
H
I
guess
bread,
please
also-
and
we
have
ten
on
here
on
Brandon,
mentioning
that
we
might
get
a
little
bit
more
resources
on
on
the
build
systems,
meaning
because
that,
if
we
want
to
run
it
with
every
PR,
then
we
would
beat
that
because
right
now
it
already,
if
you
start
merging
and
updating
the
branches,
everything
starts
taking
really
long.
Yeah.
D
Quite
figured
out
who
like
how
the
circle
CI
organization
ownership,
stuff
works,
like
do
I
have
like
I,
think
I
need
to
talk
to
someone
who's
actually
like
an
admin
of
the
open
to
a
morgue,
and
then
they
have
to
provide
a
card.
I
can't
just
like
get
the
card
my
manager
will
throw
it
in
which
is
fortunate,
yeah
I
think
we
have
a
go
ahead.
Let's
this
other
question.
To
that,
I
have
about
circle,
see
I
thought
like.
E
Don't
I
don't
have
a
credit
card
I'm
not
yet
on
the
governance
board,
but
I
am
happy
to
poke
people
if
it's
literally
a
matter
of
like
only
someone
with
admin
access
needs
to
put
the
information,
then
I'm
happy
to
be
HS
and
Sergei
and
Bogdan
till
it
gets
done
so
Brandon.
It
sounds
like
this
is
something
you've
got
a
handle
on,
maybe
yeah.
You
can
take
this
on
yeah.
H
Mean
such
a
benchmark,
always
just
maybe
it
should
be
a
separate
process,
because
the
the
benchmark
right
now
obviously
has
no
stage
right.
So
a
benchmark
makes
more
sense
if
we
have
some
way
to
track
like
progress
so
that
we
see
when
we
do
a
change
how
this
change
affects
the
overall
performance.
It
is.
C
Actually,
at
that's
good
point,
I
was
thinking
about
that
as
well.
One
of
the
options
is
to
compare
with
the
released
released
version
with
the
current
one
we
have,
and
then,
if
we
see
the
improvement
in
the
numbers,
then
that
might
be
the
good
indicator
that
okay,
the
things
that
we
changed,
it's
making
sense
in
terms
of
performance
and
optimization
yeah.
K
H
C
Another
option
I
was
thinking
to
release
this
benchmarking
number,
along
with
the
new
release.
Once
we
release
some
version,
we
can
put
all
sort
of
benchmark
numbers
that
we
have
in
countries
and
yeah
somewhere
in
the
readme
file.
I
saw
somewhere
some
sort
of
extra
file
and
then
at
least
we
have
some
numbers
to
compare
when
we
do
next
release
and
all.
C
And
one
interesting
fact,
I
found
is
the
HTTP
text.
Context
in
inject
method
specifically
is
kind
of
better
than
the
v3
format,
but
then
extract
method
for
the
beta
format
is
much
more
faster
than
the
should
repeat
text
format.
So
it's
kind
of
difficult
to
I
mean
I'm,
not
sure
whether
we
are
recommend
someone
to
use
any
particular
format.
But
if
you
compare
these
two
at
one
place,
one
is
better
and
other
place.
The
other
is
better.
So
that's
the
current
state
so.
H
C
One
comment
Dave
added
is
to
instead
of
having
on
the
outside
the
packages,
we
can
build
this
sort
of
benchmarking
inside
in
each
individual
packages,
so
next
time
onwards.
If
someone
wants
to
do
some
kind
of
benchmarking,
they
can
include
in
that
only
that
particular
package,
and
only
that
think
we'd
run
something
like
that.
So.
H
C
K
A
I
added
that
I
don't
know
if
it's
too
early
in
the
process
for
something
like
that,
but
I'm
relatively
new
to
the
project
and
one
thing
that
I
was
struggling
with
as
I
as
I'm
getting
ramped
up
here
is
just
like
a
basic
example
using
Jaeger
and
everything.
You
know
just
like
a
10
minute,
if
I've
never
seen
this
before
getting
started,
guide,
I
was
able
to
get
it
figured
out,
but
it
took
me
longer
than
I
would
have
hoped.
A
C
C
A
H
H
C
J
Is
just
one
more
question
and
do
we
have
any
any
specific
way
of
naming
response,
because
if
okay
now
I'm
doing
them
documented
all
and
all
stress
approaching
how
to
exactly
named
this
panel
and
to
your
head
guys
any
any
suggestion
for
that
or
basically
how
I
mean
there
is
I
mean
I
am
I
just
took
something
from
the
matrix
like
there's
like
connect
down
complete
and
so
on.
So
I
can
assume
that's
some
way
of
naming
fans
but
other
than
that.
I,
don't
know
be
honest.
What
will
be
the
correct
name
for
first
time.
I
I
D
There
yeah
yeah
I
feel
expand.
A
being
is
a
thing.
That's
like
there
are
no
really
good
best
practices
around
it
besides,
like
I,
feel
like
the
standard
trans
like
this
band
name,
should
be
the
action
of
work,
that's
occurring,
but
that's
kind
of
it
yeah
in
terms
of
like
whether
you
use
underscore
or
like
slashes
or
whatever,
it's
like
up
in
the
air
really.
D
H
E
E
E
H
I
E
C
C
C
E
H
H
E
E
C
E
C
E
Right,
yeah
I
have
one
thing:
I
threw
at
the
bottom
of
the
agenda
just
to
raise
awareness
around
there's
this
question
about
like
default
behavior
for
the
API.
If
you've
not
installed
an
SDK,
so
literally
we're
saying,
there's
no
SDK
installed,
but
somehow
someone's
installed
enough,
instrumentation,
perhaps
natively
that
they
are
able
to
propagate
headers
Sergei,
and
some
people
have
been
really
hot
on
the
idea
that
by
default
we
should
be
like
generating
span,
IDs
and
and
propagating
traces.
I
think
this
is
like
a
noble
suggestion.
E
It
does
suck
when
someone
adds
a
service
that
hasn't
been
configured,
and
so
your
traces
break,
but
I've
personally
been
concerned
that,
like
for
project
we're
proposing,
gets
embedded
in
literally
every
piece
of
software
having
an
on
by
you
know
on
by
default
or
behavior
by
default
is
like
not
a
good
default.
Yeah.
D
We
had
this
discussion
a
very
long
time
ago,
when
the
project
was
first
starting
off.
I.
Remember
that
one
of
rock's
big
concern
with
like
always
by
default,
is
that
like
they
have
customers
where
like
and
even
they
switch
off
the
service
like
any
speed
off
off,
and
so
the
overhead
by
like
propagating,
expand,
IDs
and
stuff
like
that
was
like
not
great
and
so
I
think.
That's
why
we're
default.
They
want
behavior.
It's
like
it's
truly
no
op
yeah.
E
E
E
N
C
So
this
is
just
a
kind
of
fun
fact
and
it's
not
kind
of
fun
fact,
but
this
group
is
like
really
motivated
and
working
on
the
plugins,
so
we
have
almost
like
three
to
four
plugins
already
in
place
and
we
are
already
working
on
the
post
creation
of
the
plugins.
So
how
do
you
feel
about
this?
Is
it
a
good
time
or
how
do
we
prioritize
days
and
all
I.
E
Mean
I
think
it's
great
I
mean
we
need
to
I'm
hoping
as
soon
as
we
get
an
alpha
at
the
door.
We
can
start
exercising
it
and
in
order
to
do
that,
we
need
to
be
writing
plug-ins
an
application
code
with
an
eye
through,
to
my
mind,
one
is
like:
does
this
just
not
work
like,
let's
make
sure
we're
looking
at
edge
cases,
I,
don't
know
what
those
are
for?
No
jeaious
and
python.
E
H
Yeah,
yeah
and
I
guess
also
I'm,
not
sure
real,
quick
about
other
languages
like
Java.
If
there
is
already
like
just
plugging
infrastructures
back
out,
I'm,
not
sure
about
that,
because
I
liked
it
and
I
think
that
we
can
learn
a
lot
from
how
node
is
doing
it,
because
this
plug-in
loader,
it's
like
you,
you
hook
into
the
plug-in
loading
more
or
less
and
then.
H
H
It's
more
about
like
having
plugins
in
place
and
then
the
DES
pet,
so
that
the
auto
instrumental
figures
out
which
thing
to
use
which
library
to
do
the
patched
and
yeah.
So
that
would
so
that
you
can
say:
oh
I,
don't
want
to
use
this
HTTP
instrumentation
from
OPA
telemetry
I
want
to
use
something
else
in
some.
E
Yeah
no
I
think
that's
awesome
related
to
that.
I
would
like
to
see
us
port
the
registry
over
so
in
open
tracing.
We
had
a
very
simple
thing
on
the
website:
I
like
to
improve
where,
regardless
of
where
you're
hosting
your
instrumentation,
you
can
just
add
it
to
this
registry,
so
people
can
find
it
and
yeah
then
going
forward
from
that.
E
If
there
is
hopefully
an
auto
installer
for
every
language
that
handles
like
I'm
trying
to
call
it
auto
installer,
because
when
you
say
dynamic,
instrumentation
I
noticed
some
people
get
confused
and
they
think
you're
talking
about
something.
That's
like
dynamically,
adding
instrumentation
points
as
opposed
to
just
installing
these
things,
but
something
like
that
in
every
language.
So,
let's
you
just
here's
my
pilot
plugins,
please
install
them
for
me,
don't
turn
on
the
HTTP
one
for
this
named
component.
Thank
you.
Something
like
that
would
be
really
great
cool,
I.
E
Go
doesn't
want
this
because
it's
full
of
Luddites
who
think,
if
you
take
a
photo
of
them,
you've
stolen
their
soul
and
Java,
has
like
so
many
Java
agent
options
that
it's
like
a
like
gang
war.
Trying
to
like
pick,
which
is
the
right
approach
like
no
Jess
hidden
it
right
in
the
middle
raid
that
sweet
spot.