►
From YouTube: 2022-05-18 meeting
Description
OpenTelemetry PHP SIG Meeting
D
B
Good,
yes,
I
was
confused
because
we
don't,
we
obviously
can
track
that
down
to
get
a
oh.
I
think
I
think,
if
you
said
good
day,
no
one
would
know
what
you
meant.
D
G'day,
it's
also
very
fun
to
say
yeah.
It.
E
B
C
D
C
It's
doing
going
great
lots
of
work
because
we
put
out
a
v2
version
too,
so
there's
a
lot
of
stuff
to
fix
until
the
end
of
last
week,
but
now
it's
calmer.
So
I
have
a
little
bit
more
time.
D
C
Yeah,
so
basically
it's
based
on
git
or
the
configuration
that
people
have
start
somewhere
and
get
on
there
on
github
or
something
like
that.
So
they
can
do
versions
and
branches
of
the
software
and
stuff
like
this.
So
basically
like
git
ops,
button
data
way.
C
D
So
today,
we'll
start
by
reviewing
our
open,
prs
and
open
box
and
then
looks
like
kashan
has
a
sas
update
on
instrumentation
with
symphony,
so
the
open,
telemetry
php
repo
has
three
pr's
open,
but
two
of
them
are
semi-tied
together.
I
believe
brett
made
a
beautiful
pr
for
concurrent
traces
and
then
nevaeh
at
its
additional
context
in
a
separate
pr.
But
do
you
want
to
take
that
away.
B
Yeah
yeah,
so
so
my
or
my
method
involves
creating
a
lot
more
storage
and,
to
be
honest,
I
was
kind
of
hoping
that
that
nevaeh
or
one
of
the
original
context
authors
would
pop
in
and
help,
and
so
it
looks
like
his
pr
that
the
secret
one
is
both
less
invasive
and
probably
more
elegant
than
than
my
own
solution.
So
I've
been
chatting
with
him
on
on
slack
and
he's
and
doing
some
testing
of
of
the
pr
and
it
it
seems
really
good.
D
Sweet
deal,
the
the
his
pr
is
still
in
draft
is
that
because
y'all
are
still
conking
through
some
of
the
implementation
to
each
other
anything
he
just
hasn't
changed.
B
No,
I
did
ask
him
about
that
earlier
today.
He
said.
Yes,
it
is
very
close
he's
he
might
drop
a
couple
of
things
out
of
it
just
just
to
keep
to
keep
the
scope
a
bit
smaller
and
sort
of
get
something
through,
but
it
sounds
very,
very
close
to
being
ready
for
review
and
merge
wonderful.
D
D
We
had
one
more
the
we
have
a
creating
events
and
listeners
api.
Pr,
that's
still
up.
It
looks
like
right.
You
you're
the
last
one
to
make
comments
about
that.
B
Yeah,
it's
like
I
don't
know
if
he's
spoken
to
anyone
else,
but
kumar
did
just
have
a
couple
more
questions
for
me
about
that.
Honestly,
it
looks
like
he's
still
still
working
on
it
a
little
bit.
It's
not
it's
not
dead
and
it's
not
abandoned.
B
D
C
B
I'm
pretty
happy
with
how
it
is
he's
got
a
few
things
that
that
he
needs
to
fix
up
from
you
know
the
last
time
we
reviewed
it
so
wait
for
it
to
get
get
those
all
sort
of
actioned
and
knocked
off,
and
then
we
can
have
another
look:
hey
dog.
D
Excellent
and
as
far
as
the
contrib
repo,
there
are
three
pr's
two
and
draft
state
one
ready
for
review
the
scpk
bunder
factory
team.
I
know
that.
That's
that's
a
big
thing
and
you'd
mentioned
that
you
had
other
other
priorities,
but
I'm
just
mentioning
that
one's
there.
C
Yeah
to
me
my
my
vr.
D
And
then
cedric
had
opened
one
with
a
instrumentational,
an
instrumentation
bundle
to
capture
signals
from
symphony
apps
out
of
the
box.
I
followed
up
with
him
earlier
this
week
and
he
said
he
has
everything
except
for
time
and
I
can
empathize
with
that.
C
D
He's
working
he's
working
through
that,
I
believe
and
then
there's
one
additional
pr
by
w-a-d-j-e-I
and
he
is
working
on
getting
his
cla
signed
so
that
we
can
merge
that
pr
with
a
bundle.
Kernel
listener
example.
C
That's
just
the
documentation
right.
I.
D
A
Sure
so,
basically,
I
just
needed
status
update
on
where
we
are
with
symphony
and
auto
instrumentation.
In
that
case,.
B
A
I
wanted
update,
I
don't
have
anything
from
mine.
Basically,
I
have
just
set
up
a
symphony
framework.
The
symphony
is
a
framework
example
that
is
mentioned
in
on
our
page,
and
I
ran
into
some
issues.
So
first
thing
was
to
get
the
root
span
and
that
I'm
working
on
found
a
workaround.
But
let's
see
if
that
works
or
not
I'll
talk
to
brett
about
that.
D
I
don't
think
that
auto
instrumentation
is
directly
front
of
mind
for
anybody
right
now.
Is
there
anybody
in
the
car?
Thinking
about
that.
D
D
C
C
Well,
I
could
could
need
some
additional
things.
Time
is
an
issue
as
well.
C
A
I
can
do
that
as
well,
but
many
thing
is,
I
assume
that
instrumentation
would
be
something
that
would
be
targeting
for
ga.
I
guess
I'm
not
sure
if
you
are
targeting
it
for
beta.
C
Well,
the
thing
with
the
instrumentation
of
this
stuff
is:
we
have
to
some
sort,
have
a
stable,
sdk
and
api
to
run
against
us,
because
the
issues
which
we
already
had
well,
we
had
a
lot
of
issues
with
keeping
up
updating
the
control
people
with
all
regarding
to
the
changes
we
have
done
in
sdk,
so
yeah.
Once
we
have
everything
in
place,
then
it
should
be
easier
to
actually
do
something
with
spreading
words
and
things
like
that,
because
there
won't
be
so
much
work
when
you're
changing
the
other
way
around.
C
C
C
Yeah
and
for
for
these
tutorials
on
in
the
repo
we
have
discussed
a
while
for
that,
we
may
want
to
abandon
them
because
they
are
teaching
some
things
which
are
not
really
best
practices,
and,
let's
put
it
that
way-
and
I
don't
want
to-
or
we
probably
don't
want
to
give
out
examples
which
maybe
people
copy
and
then
think
this
is
like
a
recommended
way
to
set
up
stuff.
C
That
was
my
concern
and
I
could
put
it
or
you
know,
go
into
detail
what's
actually
about,
but
you
know
that
was
actually
what
was
in
my
mind
since
a
long
time.
D
Yeah,
I
agree.
I
agree
with
you
team.
I
think
that
these
aren't
perfect
examples
and
I'm
always
conflicted
with
that,
because
there
you
know,
as
a
brand
new
user
to
this
it
would
be
nice
to
be
able
to
come
to
our
repo
and
go
oh.
This
is
how
you
can
use
our
library
quickly
and
easily,
but
if
the
practices
aren't
perfect,
then
it
that
kind
of
falls
into
that
anti-pattern
right.
We
want
something
that
is
universally
accepted
as
a
reasonable
implementation,
and
if
we
have
to
remove
them,
I'm
all
in
favor,
for
it.
C
Yeah,
well
I
mean
the
issue
is
right:
there's
no
real
instrumentation,
so
you
could
set
up
this
everything
because
it's
basically
just
using
a
global
scope
which
you
can
do
in
the
framework
you
shouldn't
and
you
can.
You
could
basically
use
it
everywhere
and
everywhere.
It's
not
really
a
good
way
to
deal
with
things.
C
Yeah,
that's
just
to
that's.
You
maybe
see
those
tutorials,
so
it's
not
like.
I
would
see
them
as
very
good
examples
or
tutorials
for
beginners
but
yeah.
Maybe
we
should
have
something
additional
in
place,
which
is
a
little
bit
more
to
more
exp
or
targets
more
experienced
developers,
because
for
me
personally,
one
of
the
issues
would
be.
This
is
to
way
too
handholding
for
me
to
read.
You
know.
So
it's
telling
me
a
lot
of
stuff
about
which
I
already,
which
is
basically
for
me,
it's
just
information
and
noise.
C
A
Yeah,
I
agree
with
that.
Actually,
the
examples
that
we
have
on
the
readme
page:
they
are,
they
work
only
with
hotel,
phps,
sticker
version
0.0.3
and
below.
Currently
we
are
at
0.0.10
this
to
just
get
it
to
work.
C
Yeah,
that's
the
other
issue
that
they're
are
basically
outdated,
so
this
version
is
about
a
year
old,
which
is
it's
depending
on
so
that
that's
just
another
thing
so.
B
Could
just
put
up,
for
you
know
new
contributors
looking
for
something
to
do
there,
I'm
sure
enough
people
with
laravel
experience
that
they
could
build
a
better
example
and
update.
D
C
D
B
B
C
D
C
D
Okay,
those
are
the
those
were
the
things
that
were
on
the
agendas
anybody
yeah.
C
Justin
requirement,
maybe
for
the
tickets,
that
we
make
a
requirement
to
use
actually
some
books
or
entry
points
or
whatever
you
have
there
for
from
a
framework
and
what
you
know,
some
trickery
which
may
work.
So
there
are
ways
in
most
frameworks
that
you
can
integrate
stuff
like
this.
You
know
without
overwriting
the
front
controller
on
some
global
stages.
D
C
Well,
I
don't
know
if
you
already,
the
documentation
is
really
good,
but
it's
as
it's
hard
to
follow,
sometimes
what
it
does,
because
there's
so
much
magic
in
it.
You
know
global
methods
and
functions
and
things
that
try
to
be
smart
and.
B
C
B
C
Didn't
take
any
notes,
but
I
know
what
you
know:
what's
basically
on
my
table
more
or
less
so
the
most
important
thing
which
I
will
put
out
and
let's
do
since
a
long
time,
is
basically
the
speed
and
the
splitting
of
or
creating
packages
and
splitting
the
two
repositories
into
separate
packages.
So
my
initial
idea
before
I
started
my
work
is,
I
quickly
finished
the
symphony,
the
rework
pr
which
I
had
there
and
then
once
this
is
done.
C
I
create
the
packages
first
for
the
contract
people
because
it's
easy
to
test
and
then
do
the
same
for
for
the
manual
repository
still
my
plan,
though
I
might
might
do
the
packet
splitting
before
I've
done
a
lot
of
testing
in
the
past,
so
it's
basically
just
implementing
this
stuff
and
then
yeah.
C
My
my
additional
plan
was
to
adding
some
more
stuff
on
the
tremendous
plus.
This
will.
I
will
postpone
so
first,
basically
do
the
split
for
the
packages
and
during
the
stuff
we
need
by
hand
first,
you
know
so
later
there
will
be
some
magic
if
there's
something
some
new
package
created
in
the
controller
preview,
for
example,
to
set
up
the
split
package,
repository
and
composer
or
packages
on
entry
and
stuff
like
this.
So
but
I
think
we
can
work
on
this
later
and
just
make
the
basic
functionality
works
yeah.
C
So
I
will
on
the
symphony
pr
I
don't
know.
Actually
I
wrote
a
lot
of
custom
code
because
this
configuration
thing
and
since
for
me,
is
kind
of
well
it's
hard
to
digest
for
all
the
static
package
managers
what's
going
on
there
and
it's
actually
yeah.
C
So
I
wrote
a
lot
of
custom
stuff
to
make
it
easier
and
I
actually
started
with
the
last
the
complaints
there
were
by.
I
don't
know
who
that
was
fan
or
hp
stand
or
whatever,
but
I
think
I
will
maybe
postpone
it
because
there's
a
lot
of
stuff
to
do
it's
kind
of
nice
but
yeah.
C
So
I
will
basically
just
dirty
fix
the
issues
by
just
suppressing
them
and
then
make
this
pr
work
and
I
can
do
the
other
stuff
later,
which
I
still
have
a
lot
of
open
code
on
my
site.
So
it's
half
finished,
but
the
last
half
is
a
little
bit
tricky
and
not
super
exciting.
So
I
didn't
finish
it.
C
Yeah
sure,
but
maybe
we
can
you
guys,
can
update
me
on
what
the
actual
plans
are
for
better
release
and
things
like
that
and
what
you
have
done
in
that
regards.
D
So
right
now
the
plan
for
we,
we
don't
have
a
a
concrete
plan
for
a
beta
release.
We
do
have
a
bunch
of
issues
that
are
linked
and
labeled
as
required
for
beta.
So
I
think
that
we
need
to
get
all
of
the.
We
have
seven
tickets
in
there
and
those
are
the
ones
that
we're
aware
of
that
are
required
for
beta,
but
there
may
be
more
so
if
you
have
any
input
there,
that
would
be
really
great.
Otherwise.
D
I
think
that
we're
getting
to
a
point
now
where
we
really
do
need
to
get
more
feedback,
and
I
think,
by
putting
this
project
in
a
beta
state,
we're
going
to
get
a
little
bit
more
feedback
and
that's
a
good
thing
and
probably
get
a
lot
more
contributors
that
way
too,
which
is
also
a
good
thing.
So
I
think
that
that's
something
that
we
can
do
and
then
part
of
the
process
when
we
go
from
the
way
that
the
open,
telemetry
php,
the
open,
telemetry
maintainers
group
works,
is
they
will
have.
D
Somebody
from
the
group
do
a
code
review
of
our
repository
and
give
feedback,
and
then,
from
that
point
we
can
go
to
a
release
candidate
and
do
a
couple
iterations
on
the
release
candidate,
and
then
we
can
do
our
general
availability
release.
So
I
think
that
we'll
need
to
burn
through
these
tickets
that
are
in
the
required
for
beta
open.
D
Our
you
know
create
our
first
beta
release,
solicit
feedback
from
both
the
other,
both
the
maintainers
from
other
sigs
and
from
the
public
and
then
release
candidate
and
then
general
availability
and
then
start
working
on
metrics
and
logs.
C
Okay,
so
who's
gonna.
Will
this
person
be
who's
doing
the
code
review.
D
D
So-
and
you
know,
they're
all
very,
very
good
programmers,
so
they
can
sort
of
attempt
to
reason
through
the
php
and
they
also-
and
we
also
will
attempt
to
find
somebody
who
is
php
fluent
to
do
it
in
that
group,
and
I
think
we
have.
We
have
a
lot
of
people
who
have
had
previous
experience
with
php
in
that
group.
So
I
think
we
should
be
able
to
find
somebody
that
can
review
that
effectively.
But
that's
something
that
I
will
ask
for.
I
can
ask
for
that
in
the
maintainers
meeting
next
week,.
C
Okay,
let's
hope
yeah
one
thing:
brett,
we
should
maybe
in
the
very
near
future,
have
for
a
long,
a
planned
meeting
regarding
the
configuration
and
did
you
have
any
progress
in
this
regard
lately?
No,
I
don't
have
any
progress.
Okay,
then,
let's
just
pick
up
where
we
left
where
we
wanted
to
have
a
meeting,
and
then
we
can
talk
through
what
we
have
and
maybe
or
we
maybe
use
the
configuration
from
symfony
just
temporary.
Anything
else
would
be
too
much
work
or
too
opinionated
or
something
like
that.
B
C
Me
to
chase
that
one
up.
Well,
I
mean
anyway,
we
I
don't
think
we
should
or
could
wait
on
this
to
happen.
If
there
are
some
changes,
we
can
maybe
incorporate
them,
but
we
need
to
have
a
solution
for
that,
because
it
will
also
make
make
it
easier
to
work
with
frameworks
and
stuff
like
this.
If
you
can
just
reuse
configuration
yeah,
that's
basically
all
the
sdk
and
symfony
bundle
does
at
the
moment,
because
there
needs
to
be
a
way
to
configure
the
sdk.
C
You
know,
and
so
you
just
have
it,
you
can
have
it
as
a
dependency
and
yeah.
That's
something
we
probably
need
in
a
lot
of
different
frameworks
and
libraries
as
well.
So
yes
and
make
this
the
other
way
thing
I
was
thinking
about
is
we
need,
and
I
don't
think
we
have
this
correctly
in
place,
so
we
need
a
way
that
well
essentially,
it
goes,
but
we
have
thing
to
think.
C
I
think
over
that
we
need
a
way
to
for
instrumentations
to
only
run
against
the
api,
so
they
should
not
use
well.
They
can
use
the
sdk
just
to
set
up
something,
but
the
dependency
should
be
strictly
on
the
api
and
it
should
do
something
but
no
break
anything
so
just
to
put
it
very
simply
so,
did
you
guys
have
a
look
or
thought
about
that?
C
If
we
actually
could
do
that
or
we
maybe
should
create
an
example,
because
at
the
moment
we
are
way
too
focused
on
a
lot
of
things
on
the
sdk
right,
so
we're
not
really
running
the
api
so
yeah.
Maybe
we
should
look
again
at
this
if
it's
actually
spec
on
client
and
how
we
that's
wrong,
what
do
you
think
yep.
B
Yep
definitely
I
have
seen
in
like
in
the
past
couple
of
days
there's
at
least
one
to
do
where
you
know
we
need
to
return
and
go
up
something
so
that
so
that
the
api
itself
can
can
work
without
an
sdk.
C
Should
yeah?
Okay,
then
we
should
look
into
this.
Basically,
if
you
depend
on
an
interface
but
the
main
issue,
which
is
well,
you
can
yeah
okay.
We
should
look
into
this,
but
basically
you
should
be
able
to
just
depend
on
api
on
interfaces
which
are
defined
in
the
api,
and
it
should
not
break
anything.
If
there's
no
sdk
and
if
there's
an
sdk
or
another
implementation,
then
you
should
make
basically
or
be
able
to
really
instrument
something
so
right.
C
D
C
E
How
are
you
did
you
make
any
progress
on
the
v3
header
thingies?
Oh
yes,
I
have
made
some
changes.
Sorry
in
a
day
or
two
I
will
disappear.
C
Okay
right
so
yeah,
if
you
have
any
questions
also,
I
should
have
more
time
now
so
feel
free
to
ask
me,
and
otherwise
you
know
just
create
a
pr
and
then
you
can
work
out,
but
I
don't
think
it's
too
complicated
once
you
get
into
it
right.
A
No,
it's
not
so
yeah
I'll
raise
appear.
You
can.
C
Okay,
cool,
thank
you
yeah
and
different
things.
What
what
do
we
do
with
the
documentation
issue?.
C
I
mean
the
the
ticket
says.
Well
I
mean
it
was
changed
a
lot
of
in
scope
because
our
documentation
has
been
moved.
Let's
put
it
that
way
and
well.
The
ticket
is
kind
of
we
have
some
documentation
there.
So
there's
no
real
target.
You
know
what
which
we
can
aim
at.
C
D
D
C
Yeah
on
the
side-
yeah
not
much
there,
it's
basically
just
the
the
basic
hello
world
thing,
but
on
the
other,
other
languages
don't
have
much
more
on
the
website
so
but
it
would
be
good,
so
you
know
at
least
have
some
bullet
points
which
should
be
there,
so
we
can
mark
it
as
done
you
know,
we
can
probably
further
improve
on
the
future,
but
some
documentation
is
not
really.
You
know
the
criteria.
Answer.
D
D
C
C
D
B
I
did
yeah
just
a
couple
of
very
quick
updates
sort
of
side
projects
of
open
telemetry
one
was
the
roadrunner
runtime.
I
worked
with
their
developers
to
sort
of
test
and
and
provide
feedback
on
their
open,
telemetry
implementation.
So
that's
all
done
and
released
and
sort
of.
I
could.
B
Yeah
so
under
the
hood
it
uses
the
go
open,
telemetry
implementation,
but
so
the
way
it
works
is
because
you've
got
the
roadrunner
server
and
it
will
instrument
the
incoming
request
there.
B
B
C
Yeah
convoy
or
a
site
proxy,
I
wanted
to
say
cool
so
and
does
it
you,
how
does
it
pass
the
trace
information
to
the
actually
to
try.
C
Transparent
header,
but
so
it's
the
w
c
three.
B
B
Sounds
cool
yeah
and
the
other
one
was
psr22
which
fan
wants
to
jump
in
and
have
a
look
there's
a
bit
of
activity.
Starting
in
discord
is
where
that's
all
happening.
Okay,
then
I
will
have
a
look.
B
Yeah
yeah.
It's
just
really
just
started
to
pick
up,
probably
the
last
day,
they're
actually
starting
to
to
talk
interfaces
a
bit
so
get
some
progress
happening.
There
you're
welcome
to
jump
in
if
you
have
an
opinion
or
if
you've
already
volunteered
to
do
so
too.
B
B
B
B
D
All
right
that
was
the
last
thing
on
the
agenda.
Anybody
else.
D
Once
going
twice
sold
all
right
thanks
for
thanks
for
showing
up
today,
everybody
glad
to
have
y'all
here
and
we'll
see
you
on
the
internet.