►
From YouTube: 2019-10-16 PHP 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
B
D
D
A
I
sent
a
fun
email
to
Morgan
McLain.
Yesterday,
to
you
too,
we
had.
We
had
a
wait.
A
meeting
with
with
Etsy
yesterday,
that's
sort
of
like
I,
don't
want
to
call
it
a
venting
session,
but
that's
what
it
kind
of
felt
like
about
open
census
and
how
we
feel
like
we
can
make
sure
that
it
gets
improved
with
open
telemetry
and
it
spurred
the
question
like
hey
we're,
paying
alimony
for
stackdriver
shouldn't.
E
A
B
Yeah,
that's
probably
the
sticking
point.
It's
been
like
our
issue.
It's
like
you
know.
We
want
all
of
these
things
to
work
at
light
step,
but,
like
light
step
itself,
is
like
a
go
shock,
so
you
know
everyone
we
you
like
to
hire
is
like
hired
because
they
want
to
work
on
go
and
they
know
how
to
do
that,
and
so
it
means
like
yeah,
we
kind
of
like
it's
not
necessarily
easy
for
us
to
like
staff.
All
of
these
things
directly
at
this
time
right,
north.
B
We're
trying
to
hold
down
open
to
a
tree
in
a
major
way,
like
you
know,
we're
like
kind
of
an
originator
of
all
of
this
nonsense
from
the
beginning.
So
we
see
it
as
like
aa
responsibility
to
make
it
work,
but
it
is
true
that,
like
Google
and
Microsoft
have
orders
of
magnitude
more
employees
than
we
do
yeah
so
yeah
I
would
like
to
see
them
contribute
a
lot
of
resources.
I
have.
D
E
D
B
A
A
He
said
that
he
was
gonna
come
today,
but
let
me
see
if
I
was
chatting
with
him
on
get
her
earlier.
Oh.
A
So
what
we
know
now
is
Dimitri's
working
on
getting.
He
I
think
what
happened?
Is
he
inadvertently
signs
his
BR
with
whose
work
you
helped
account,
rather
than
his
individual
github
account?
So
when
he
did,
that
is
said
that
the
CLA
was
not
signed
for
him,
even
though
he
had,
he
had
signed
it
with
his
work
account
and
then
he
had
committed
with
his
personal
account.
So
that's
where
the
conflict
was
so.
C
A
So
once
that's
where
once
up
to
merge,
will
have
a
pretty
good
base
to
start
on
he's
instrumented
like
a
very
basic
tracer
and
a
very
serious
manner,
and
a
very
basic
like
spam
context,
propagator
and
obvious
things.
My
plan
next
is
to
work
on
the
sampler
or
the
or
PHP
I.
Think
that
that
would
be
that's
our
next
step
and
then
we
can
continue
following
down
the
API
specification.
That's
available.
F
A
Then
after
and
we
also
have
our
issue,
our
issue
collector
updated,
which
is
good,
and
we
have
what
else
we
got.
We
got
read
me
up.
They
did.
We
have
some
examples
and
read
me
wants
to
me
trees
that
work
like
Linda
mentioned
so
I
think
we're
honor
like
we're
getting
the
ball
moving
from
nothing
to
something.
It's
just
a
matter
of
getting
it
like
working
with
me
to
get
that
first.
B
A
A
F
B
B
B
B
A
A
A
If
y'all
want
to
start
taking
a
look
at
Dimitri's,
PR
and
thinking
and
thinking
about
how
to
build
on
that,
that
would
probably
be
helpful
I'm
trying
to
think
of
anything
else
that
we
need
to
just
I,
don't
think,
there's
a
whole
lot
else.
We
need
to
discuss
tab.
We're
going
to
keep
tracking
through
on
get
ur.
Dad
I
am
I.
Think
that's
the
best
place
for
us
to
chat
a
synchronously
here.
D
A
Yeah
yeah,
so
yeah
Demetri
put
in
a
pickup
like
a
thousand
one
PR
for
the
first,
like
yeah
I've
gone
through
and
I've
read
the
code
and
I've
asked
some
questions
about
it
and
it
seems
like
a
very
good
start
to
me
and
then
I
think
we're
just
until
we
are
having
a
difficult
time
getting
started
with
like
getting
momentum
to
start
the
project,
so
I
think
I'm,
hoping
once
that
it's
in
then
people
can
start
building
off.
Of
that.
A
Because
I
think
we're
already
a
little
behind
based
on
like
compared
to
the
other
six,
because
we
started
much
much
later
than
them.
So
it's
expected
that
we're
behind,
but
I
think
it's
it's
important
to
recognize
that
we
can't
just
like
sit
on
our
hands
for
a
while.
We
need
to
get
clacking
away
at
the
keyboard.
B
Yeah,
it's
really
just
the
issue
that
you
know
until
open,
telemetry
gets
up
to
parity
with.
You
know,
open
tracing
and
open
census.
Then
it's
kind
of
like
I've
been
calling
it
like
the
valley
of
death
right
where
we're
kind
of
like
saying
hey.
These
old
projects
are
end
of
life,
but
like
the
new
ones,.
B
F
C
B
Iterating
on
what
was
in
open,
tracing
or
open
blowing
things
down
from
a
aspect
standpoint,
just
to
get
you
up
to
speed.
One
was
the
metrics
API,
the
open
census
creators
and
everyone
felt
like
the
metrics,
API
and
open
census
could
use
a
lot
of
work.
So
there
was
a
write
of
that
that
everyone's
a
lot
happier
with
right
now
there
was
a
desire
from
the
dynaTrace
people.
B
For
knowing
what
library
or
component
is
being
traced
or
producing
metrics
in
particular,
so
that
you
could
turn
things
off
or
change
things
like
what
kind
of
headers
you're
propagating
and
do
stuff
on
the
server,
we
didn't
really
have
a
mechanism
for
doing
that
in
open
tracing
or
open
census,
and
the
final
thing
we're
trying
to
iterate
on
right.
Now,
it's
X
propagation.
B
F
G
B
Yeah,
so
the
final
issue
is
is
context
propagation.
We
have
distributed
context
propagation,
but
right
now
it's
kind
of
glued
to
distributed
tracing
and
that's
a
little
bit.
There's
been
problematic
and
open
tracing
in
other
areas
because,
for
example,
in
tracing
you
might
do
sampling
right.
So
you
may
be
just
like
disabling
tracing,
depending
on
how
your
tracing
systems
working,
but
if
you're
also
trying
to
propagate
something
else,
like
baggage
so
you're,
just
trying
to
propagate
some
values.
To
do
things
like,
let's
say,
a
be
testing
in
a
distributed
system.
B
So
if
you're
you've
deployed
some
new
version
and
you're
saying
like
if
you're
in
the
a
path,
it
might
go
through
your
code
in
one
way
in
a
distributed
system
and
if
you're
in
the
B
path,
it
might
go
through
your
code
another
way
in
the
distributed
system.
That
can't
be
something
that
gets
sampled
right
like
that
has
to
always
be
propagated.
Even
if
you're
sampling,
your
observability
stuff,
so
having
like
baggage,
be
like
mixed
in
with
tracing
and
open
tracing
was
like
everyone's
just
like,
and
there
was
a
general
desire
to
be
like.
B
Can
we
just
have
a?
Can
we
separate
this
architectural
e
to
say,
we've
got
context
propagation
and
then,
on
top
of
that
we
have
a
couple
of
like
context,
propagation
applications
like
distributed,
tracing
and
baggage,
and
then
you
know
something
else
that
someone
might
want
to
create,
and
there
was
a
lot
of
interest
in
that.
But
it
is
like
this
totally
new
architecture.
B
So
that's
the
thing
I'm
championing
right
now
to
try
to
get
solved,
but
it
might
be
a
big
change
and
I
think
there
is
a
chance
if
it's
too
big
of
a
change,
we
may
just
drop
it
in
favor
of
you
know
just
getting
something
at
the
door.
That's
a
merger
of
these
two
projects
and
getting
up
to
parity
with
the
old
projects.
So
that's
sort
of
at
a
high
level
with
the
spec
work,
where
it's
sort
of
back
right
now.
F
B
Apparent,
but
especially
for
this
name,
tracer
and
contacts
propagation
separation
I
feel
like
why
I'm
going
to
all
the
SIG's
right
now
is
to
kind
of
solicit
feedback
for
those
two
issues,
because,
unlike
the
prior
spec
changes,
those
these
are
like
genuinely
new
features.
So
it's
not
sufficient
to
just
write
it
down
in
the
spec
and
say:
go:
do
it
everyone's
trying
to
implement
this
stuff
and
then
they're
like
having
questions
about
like
how
the
hell
this
is
actually
supposed
to
work,
or
how
do
you
handle
this
edge
case?
So
I?