►
From YouTube: 2021-03-18 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).
A
Yeah,
I
was
just
I
was
in
another.
There
was
a
demo
meeting
that
was
running
late.
Let's
see,
I
feel
like
we
had
something
that
we
want.
Oh
yeah,
that's
right,
record
exceptions.
B
A
Oh
yeah
is
there,
do
we
have
a
document?
Oh
the
the
signo
document.
B
A
B
B
Yeah
I
have
put
that
just
because
I
I
think
that
we
need
more
usage
from
people
just
for
for
the
code
to
be
tested.
Mainly,
we
wanted
to
be
proof,
so
I
think
that,
having
a
better
version
that
we
say
it's
a
1.0,
video
version
could
be
good
for
for
the
project,
so
I
don't
know
what
what
the
steps
to
to
take
for
this
or
how
to
work
it
out.
I
think
that
ted
talked
about
three
points
for
the
1.0
readiness.
B
So
one
of
the
things
is
end
user
feedback.
So
that's
the
thing
that
it's
more
important
to
have,
I
think,
from
what
we
have
now,
because
backwards,
compatibility
and
versioning
plan.
A
A
But
but
beyond
that
yeah
it
would
be
it's
kind
of
difficult
to
define
what
that
is,
ensuring
metrics
marked
as
experimental.
I
I
I
don't
really
know.
Does
that.
Let
me
just
like
add
a
document
in
there
that
says
that
the
metrics
yeah,
the
metrics.
B
I
I
can,
I
can
update
the
readme
okay
cool,
so
I
I
the
new
things,
the
exporter
and
the
metrics
thing.
B
A
Yeah
yeah
I'm
using
it
right
now,
yeah,
it's.
B
A
It's
yeah,
I
just
I
just
pushed
a
big
refactor
of
like
the
the
resource
kind
of
collection.
I
have
on
the
on
the
elastic
agent
that
I'm
that
I'm
working
on
and
it
all
you
know,
I'm
able
to
you
know,
merge
resources
together
and
all
that
sort
of
thing.
A
The
one
thing
that
is
kind
of
missing
are
a
complete,
a
complete
list
of
all
of
the
various
like
semantic
conventions
in
the
resource
and
and
the
actual
like
semantic
conventions
of
the
api
itself.
A
But
that's
really
the
only
thing
that
I've
seen
missing
so
far,
just
in
terms
of
functionality,
everything
else
seems
to
be
working
really
well,
like
I'm
able
to
add,
add
error
events
like
those
those
constants
that
I
just
added.
I
was
using
those
just
you
know
in
line
and
the
api
is
taking
here,
not
the
api,
but
the
the
protocol
is
actually
taking
those
and
and
and
using
them
properly
and
we're
able
to
ingest
them
on
our.
You
know,
server
side
data
as
actual
errors.
B
A
Too
we
can.
I
can
talk
with
him
on
monday,
at
the
maintainers
meeting
to
convince
them
to
make
sure
they
can.
They
can
show
up.
Oh
yeah.
B
There
is
that
maintenance
meeting
yeah
what
what
maybe
I
I
can
try
to.
I
think
I
it
was
not
an
easy
date
for
me,
but
I
can
try
to
to
go
there.
Yeah
next.
A
It's
I
think
it's
the
same
time
that
we're
holding
this
meeting
on
monday
or
but
but
on
monday.
So
I
know
it's
a
little
a
little
late
for
you
and
I
can.
I
I've
been
going
to
him,
but
it's
not
it's
just
more
for
my
own
knowledge
about
open
telemetry,
but
I
can.
I
can
bring
it
and
bring
this
up
and
ask
ted
to
come
to
this
meeting
next
week.
B
Yeah,
okay,
one
of
the
other
points
I
I
wrote
here
was
about
the
existing
issues
in
the
project.
Yeah.
Let's
take
a
look,
so
you
are
seeing
my
chrome
complete
right.
So
if
I
move
to.
B
Okay,
so
I
think
I
have
here
in
a
tab
yeah,
so
these
are
the
existing
issues.
I
think
it's
a
short
list,
so
we
we
can
talk
about
them
another
and
some
others.
If
we
think
something
should
go
here
for
short
term,
so
linux
support
and
neo
integration.
B
B
B
One
of
the
things
that
failure
is
that
when
you
have
a
network
response,
the
system
doesn't
get
that
doesn't
get
the
activity
from
the
network
request
or
did
you
I?
I
know
that
it
doesn't.
So
you
get
the
current
spam
on
the
response
and
it
doesn't
know
which
is
the
span
that
created
the
request,
at
least
in
my
experience,
so
I
had
to
use
another
table
for
for
storing
that,
but
for
the
rest
of
the
things
it
should
work
because
they
are
for
the
linux
thing.
Is
that
that's.
C
B
Of
context
provider
thing
it
also,
I
think,
unifies
with
having
a
we
have.
Oh,
maybe
not.
B
Hi,
hey:
hey:
hey,
hi!
How
are
you
good?
How
are
you
yeah
fine?
We
were
currently
reviewing
the
current
issues
in
the
project.
Do
you
have
any
point
that
you
would
like
to
talk
today?
No,
no!
I'm
just
going
to
listen!
So,
okay,
you!
I
I
think
you
asked
about
the
recording
and,
as
bryce
said,
I
I
think
it's
now
in
the
youtube
channel,
because
it
was.
A
Processing,
it
looks
like
it
was,
it
was
delayed,
it
just
takes
a
while
to
process.
B
Yeah,
it
should
be
automatically
yeah,
it's
automatically
recording
yeah,
so,
okay,
so
I'll
for
for
that
linux
support.
I
don't
know
if
it's
a
priority
for
the
project,
but
I
think
one
of
the
things
that
we
also
have
is.
We
have
differentiate
context
for
a
tracer
and
for
the
baggage
thing
it
was
because
I
wanted
to
support
baggage
fast
and
having
a
unified
thing.
B
There
involved
quite
a
lot
of
code,
but
it's
in
my
personal
plans
to
add
that,
and
I
think
that
in
the
same
time
that
writing
diet,
I
think
that
I
can
give
at
least
a
an
interface
for
linux
supporting
context
providing,
so
it
has
an
api
that
they
can
take
and
write
their
code.
I
don't
know
anything
about
linux,
so
I
don't.
A
Not
be,
it
might
not
be
too
much
work
to
get
working.
We
would
probably
we
probably
just
have
to
you
know.
Like
you
said
there
are
certain
libraries
that
we're
using
that
are
definitely
not
on
linux,
but
if
we
just
tried
to
spin
it
up
using
using
the
linux
toolset,
we
could
probably
figure
out
exactly
what
what
the
problems
are
and
at
least
document
them,
but
no
yeah.
I
don't
think
we
need
to
worry
about
it.
Yeah.
B
B
B
I
have
needed
that.
I
have
used
my
own
solution,
but
that
could
be
yeah.
That
could
be
another
nice
thing.
If
you
are
launching
processes,
maybe
not
for
mobile,
but
for
system
processes,
it
would
be
nice
to
have.
A
A
And
then
there's
also
a
couple
I
think
in
for
the
resource,
constants
and
yeah.
The
smart
conventions
are
in
the
I
api
and
then
the
resource
constants
from
the
sdk.
B
Yeah
so
resources
api-
I
opened
it
long
ago
because
I
didn't
think
it
was
functional,
but
you
have
said
that
it
is.
A
Yeah,
it
seemed
like
it.
We
just
had
that
one
problem
that
bug
fix
that
I
that
I
added
initially
and
that
got
it
working
as
it
is.
B
A
So
I
think
that
it
might
be
a
little
clunky
to
actually
inject
resources
into
into
into
trace
providers,
so
that
could
maybe
be
looked
at.
That's
just
a
better
way
to
to
get
those
available.
I'm
not
sure
I
can't
remember
if,
if
I
had
already
made
any
changes
to
that,
to
make
it
a
little
bit
better
if
you
did.
B
It's
fine,
okay,
so
it
it's
better.
After
some
changes.
A
I
think
I
think
it's
good
enough.
If
there's
yeah,
I
think
we
can.
B
A
B
That
that's
true,
I
will
add
here
also
sorry
for
this,
but
it's
something
that
we
have
never
done.
Oh
yeah!
No,
it's
good.
B
Either
okay
update
links
that
use
the
master
brands
for
community
specs
or
tips
to
use
main
branch.
I
don't
know
when
it
was
open.
I
think
that
we
don't
have
any
link
for
master.
We
already
changed
two
main
branch
in
the
project,
so
I
think
we
can
close
it
and
we
don't
have
references,
I
think,
maybe
in
the
rhythmic,
when
I
update
it
there
could
be
something
but.
B
Okay,
adjacent
collector
exporter
in
open
telemetry
protocol.
A
B
But
it's
not
json
yeah
it
could.
I
don't
know
if
someone
said
about
exporting
json
there.
I
don't
know
if
it's
there,
I
I
can
check
it
because
I
remember
someone.
B
I
don't
know
if
it
added
jason,
it
was
long
ago,
so
I
will.
I
will
check,
review
content
constant
and
determine
if
they
need
to
localized,
okay
yeah.
This
is
one
you
open
two
days
ago.
A
Yeah
yeah
sergey
brought
brought
this
up
in
the
last
pr
that
I
opened.
A
I
didn't
see
so
I
I
took
a
look
at
the
open,
telemetry
spec
and
it's
you
know,
hope
you
just
needed.
B
C
D
Yeah,
it's
been
going
good
yeah.
I
want
to
actually
pick
up
an
issue,
the
simplest
other
simple
issue,
so
I
can
get
get
familiar
with
this.
You
know
because
oh
yeah,
I
yeah
because
yeah
I
haven't
you
know
I
mean
I
looked
at
this
project.
Maybe
a
year
back-
or
maybe
you
know
earlier,
but
there's
not
a
lot
of
activity
on
this
right.
So
I
did
compile
build
and
you
know
I
think
there
was
like
a
sample
or
something
on
it,
but
yeah.
A
Yeah,
we
could
definitely
use
more
more
people
using
it
and
adding
to
it.
Nacho
vanad
worked
with
me
at
that
new
relic.
B
A
Right
yeah,
so
this
this
issue,
I
didn't
see
anything
regarding
localization
in
the
actual
semantic
convention
spec.
So
I'm
not
really
sure
if,
if
if
we
need
to
do
this
right
away
or
not
or
or
yeah,
but
it
probably
is
something
well
then
again,
the
whole
point
purpose
of
the
semantic
convention
is
to
like
have
a
standard
set
of
like
terminology
for
this
stuff.
So
I
don't,
I
don't
know
if
it
needs
to
be
all
english
or
yeah.
A
B
C
B
It
so
I
I
don't
know.
B
Yeah,
probably
so.
A
A
I
don't
know,
maybe
leave
it
open
for
now
until
until
we
can
get
a
good
like.
Maybe
we
could
ask
head
about
it
next
week
and
see
what
he
has
to
say.
B
A
B
A
A
Let's
see
all
right,
so
I
created
that
ticket.
Let
me
I'm
gonna
go
find
the
spec
info
for.
A
B
A
I
think
this
one
add
exceptions
would
also
be
a
good
first,
a
good
first
ticket
too.
A
Yeah,
I
didn't,
I
didn't,
have
okay
much
else.
B
So
we
also
talked
last
week
about
instrumentation
libraries.
Oh
right
yeah,
I
don't
know
if
we
we
talk
very
slightly
about
it.
B
I
don't
know
if
we
would
like
to
have
some
common
instrumentation.
A
Yeah,
I
think
I
think
it
would
be
good
to
so,
like
ted
was
saying
that
core
instrumentation
might
be
relevant
to
actually
keep
in
the
repo,
as
opposed
to
like
some
an
external
instrumentation
repo
and
that
that
data
dog
I've
I've
reused
that
data
dog
in
this
url
session,
instrumentation
that
you
wrote
to
be
more
generic
for
like
open
telemetry
protocol,
and
so
I
was
thinking
that
maybe
we
could
pull
that
in
as
some
as
some
initial
instrumentation.
A
A
I
think
that's
the
the
hard
part
is
making
it
plugable,
because
I
don't
know
if
you
have
this
experience
with
the
ios
instrumentation,
but
it's
a
little
it's
it's
a
little
like
dependent
on
when
it
gets
actually
run
right,
like
you
can
be
yeah.
B
Yeah,
for
for
my
mainly
I
wrote
it
so
it
it's
injected
in
the
load
of
a
framework,
but
for
swift,
you
need
to
create
a
objective-c
framework
that
calls
load
method,
so
it
was
just
using
a
objective-c
load
method
to
inject
libraries
in
to
inject
a
in
nsurl
with
method.
Switzerland.
B
So
that
that's
the
way
I
found
to
to
do
it
at
the
beginning
of
the
project,
it's
not
the
cleanest
way
and
you
need
to
to
have
a
objective-c
module
there,
but
it
can
work.
B
I
I
can
I
can.
I
don't
know
if
it
will
be
a
clean
enough
way.
I
don't
know
if
people
is
truly
want
method,
switzerland
for
instrumentation.
B
Yeah,
I
think
we
can
go
with
method,
switzerland,
and
maybe
people
can
find
it
useful,
maybe
other
prefer
to
write
their
own.
A
Yeah
yeah
there's
no
other
way
to
like
hook
into
network
networking
on
ios
right.
I
I've
not
found
a
way.
A
B
Have
your
own
url
session,
I
don't
know,
what's
the
name
like
an
observer
of
the
url
session
and
you
can
do
something
there,
but
you
must
actively
set
that
as
the
session
configuration
and
that
stuff
yeah,
so
it's
not
something
that
works
of
the
self
just
by
linking
with
it.
B
We
should
provide
something
that
just
works
out
out
of
the
itself.
Okay,
and
maybe
it's
reasonable.
It
doesn't
work
for
some
developers
they
can
just
or
we
can
provide
in
the
future
or
they
can
provide
another
version.
B
I
know
that
some
libraries
don't
like
the
load
method,
because
it
can
provide
crashes.
If
it's
not
well
done,
I
don't
know
yeah,
we
we
should
decide
in
how
we
I
how
we
go
there,
but
as
it's
not
something
that
must
run
in
all
people
that
so
they
can
decide
not
to
use
it.
So
it's
it's
not.
I
don't
think
it's
a
problem
having
something
that
everyone
will
like,
but.
D
Yes,
so
this
this
project
is
only
for
for
apps
using
swift,
or
is
it
what,
if
you
know,
if
app
is
returned
with
objective
c,
does
it
does
this
library
currently
a
lot.
B
B
Yeah,
currently,
the
the
library
cannot
work
for
objective-c,
because
it's
not
objective-c
compatible
classes.
B
It
was
done
on
purpose
because
you,
it
adds
a
lot
of
logic
and
performance
code.
It
adds
lots
of
code
of
to
your
classes.
If
you
use
that-
and
we
are
also
using
extracts
when
possible
and
also
enums
and
those
are
not
supported
well
in
objective
c,
but
it
can
be
used.
If
you
just
use
a
wrapper
in
objective
c
and
swift,
you
can
have
your.
You
can
write
your
I.
I
have
done
that
in
the
past.
B
I
use
this
project
from
objective
c,
having
a
an
intermediate
objective-c
class,
so
I
have
having
a
tracer
in
objective-c
that
just
have
so
having
a
yeah,
your
own
intermediate
swift
class
that
it's
an
objective
c
defined
swift
class,
so
that
one
can
talk
with
the
sdk
directly,
and
you
can
talk
with
that.
So
you
have
to
have
an
interface
written
in
swift,
but
that
has
the
objective-c
interface.
D
B
It's
the
same
library,
but
you
you
can
access
objective-c
from
swift,
but
you
cannot
access
swift
code
from
objective
c.
If
you
have
not
added
objective
c
compatibility
to
the
library-
and
we
have
not
added
that
to
the
open
telemetry
script
by
default,
because
it
adds
lots
of
code
and
also
some
memory
to
the
structures
and
also
limits
a
lot,
the
the
kind
of
objects
you
can
use
in
swift.
A
So
the
other,
the
other
possibility
for
instrumentation,
is.
A
If
that's
possible,
I
don't
know
if
you've
seen
anything
not
true.
A
Yeah,
like
like
a
big
code
through
the
llvm.
A
Yeah,
potentially,
I'm
not
sure
if
llvm
has
the
same
hooks
that
you
know
java
java
would-
and
I
I
haven't
done
any
research
on
that.
I
was
just
curious
if
you've
heard
anything
in
that
regard
not
sure
but
yeah.
B
I
have
never
done
that
because
yeah
you
should
do
it
at
the
compile
time.
A
A
B
I
can.
I
can.
D
D
Yeah,
we
need
some
kind
of
a
library
like
like
a
bite
body
or
asm,
or
something
on
swift
to
do
something
of
that
sort
or
or
someone
has
to
investigate
or
write.
Something
of
that
sort
is
that.
D
D
B
D
B
Interact
with
the
libraries
or
languages
at
the
high
level,
it
could
be
quite
a.
I
have
never
heard
about
that,
but.
A
B
Yeah,
I
think
llvm
is
very
low
level
from
a
library
perspective
and
method
names,
perspective.
A
Yeah-
and
I
guess
on
top
of
that,
like
the
the
swift
kind
of
cli
changes,
a
lot
too
so
or
no,
is
that
the
right
word.
A
D
B
Yeah,
the
problem
is
that
yeah,
it's
compiled
code,
so
you
you
need
to
inject
that
before
compiling
and
that
that's
not
easy.
Yeah.
D
Yeah,
I
think
that's
fine,
I
mean
you
know
the
compile
I
mean
java.
Does
it
run
time,
but
I
guess
you
could
do
it
in
in
compile
time
too,
like
with
java.
So
but
there
is,
there
is
no
mechanism
to
do
that
or
you
know,
there's
no
library
that
makes
it
easy.
Then
I
guess
it
is
something
that
is
yet
to
be
invented.
B
B
A
So
I'll
I'll
uphold
the
the
changes
that
I've
made
in
into
a
into
a
pr
for
open
telemetry,
and
then
we
could
maybe
take
a
look
at
that,
because
I've
already
done
the
work
to
kind
of
translate
it
over
to
a
hotel,
p
setup.
B
B
Okay,
I
will
yeah
yeah.
I
think
I
have
done
some
of
that,
but
maybe
not
in
the
with
the
telemetry
naming
and
that
stuff
and
yeah
we
can
check
about
yeah.
The
methods
with
lint
from
swift
can
also
be
tricky.
A
Wonder
how
you
do
this
in
swift
and
that
was
yeah.
I
was
marveling
at
your
instrumentation
there.
So
that's
good
stuff.
B
Yeah,
okay,
I
don't
know
if
there
is
something
more.
We.
B
So
beyond
this
new
relic
planning
to
use
open
telemetry
in
swift,
yes,
wherever
I
can
yes,
great
cool
yeah.
If
you
find
some
issues,
some
problems,
please
open
this
and
talk
about
it,
because
we.
D
Yeah
yeah,
I
I
was
telling
bryce.
I
actually
did
try
this
project
like
a
year
close
to
a
year
back.
You
know
at
that
time
there
was
not
a
lot
of
activity,
but
you
know
it
did
build
and
it
you
know.
So
it
was
a
working
project.
I
think
it
was
not.
I
mean
I
investigated
it
and-
and
you
know
I
kind
of
didn't
pay
attention
to
it,
because
I
think
there
was
not
a
lot
of
activity
at
that
point.
B
As
not
many
people
was
shown
interest
in
the
project
I
yeah.
It
has
been
mainly
myself
until
now,
adding
code.
I
had
a
project
working
with
a
pen
telemetry
I
before
being
a
data
doc,
because
we
I
was
working
on
a
startup
that
was
doing
some
stuff
using
open,
telemetry
on
swift
and
after
being
acquired
by
data
doc.
Also
we
are,
I
am
working
on
a
project
that
is
being
used.
B
International
data
doc
probably
will
get
to
ga
next
quarter,
so
I
I
will,
I
will
can
give
more
information
about
it,
but
yeah
we
are
using
it
for
tracing
stuff
and
also
event
handling
pricing
and
yeah.
It's
working.
I
think
it's
working
quite
quite
well,
but
some
other
areas
are
not
so
much
tested,
mainly
exporters
for
other
libraries.
I
wrote
them,
but
I
have
some
tests,
but
I
don't
know
if
they
are
really
useful,
yeah
and,
for
example,
metrics.
B
There
are
some
simple
exporters,
but
I
don't
know
if,
if
it's
really
useful
or
the
performance
is
good
enough,
yeah,
so
yeah
that
that's
that's
that
also
it's
so.
B
That
runs
with
a
docker
image,
with
hydra
in
a
docker
image
that
just
sends
there
the
traces,
okay,
so
yeah,
it's
a
very
simple
example,
but
it
shows
how
it
works,
but
it
only
supports
three
right
now
and
it
uses
a
three
flavor
library
that
I
forked
from
the
original
thrift
library,
because
it
didn't
support
the
version
of
sweet.
The
apache
3
library
is
a
bit
outdated.
B
B
B
B
It
was
like
steph
who
maintained
the
open
trades
in
objective
c
library.
So
maybe
he
can
talk
with
the
old
maintainer
yeah
and
not
that
so
we
cannot.
So
we
don't
depend
on
on
a
non-official
yeah.
D
Yeah
I
mean
if
they
should,
you
know
if
there's
not
a
lot
of
usage
of
the
open
tracing,
you
know
api
or
the
old
stuff,
then
I
guess
the
shim
is
probably
not,
and
you
know
it's
not
high
priority
right.
So
yeah.
B
D
I'm
saying
even
the
shim
right:
if
it
is
not,
you
know
if
it
doesn't
have
a
lot
of
pre-existing
customers
assume
it's
probably
not
that
yeah.
B
The
sims
should
be
working.
There
is
also
one
example
with
the
sim
working
I
think,
but
I
don't
know
if
it's
fully
working
with
all
the
stuffs
and
all
the
details.
B
Myself
just
examples
seen
that
it
worked
on
yeah.
D
Yep,
okay,
I
I
need
to
spend
some
time
on
this.
You
know
give
it
a
spin
and
then
maybe
pick
up
the
easiest
or
the
easiest
first
issues
and
and
get
more
familiar
with
it.
B
Okay,
I
think
there
are
some
examples
very
simple,
but
that
most
of
the
features
can
be
tested
there.
Just
adding
a
new
exporter
or
adding
some
spam
processor
to
them
should
be
quite
straightforward,
so
yeah.
B
Okay,
so
we
can.
We
can
edit
up
now,
let's
meet
next
week
and
see
how
we
have
advanced.
A
Yeah
I'll
ask
ted
to
come
to
our
next
meeting
and
help
out
with
with
the
beta
version,
1.0
stuff,
yeah.
Okay,
all
right
cool.
All.
D
B
D
F
F
Everyone
can
you
hear
me
yeah,
we
can
hear
you
cool
punya.
I
see
you
have
some
sweet
concrete
digs
just
like
I
do.
F
It's
yeah,
I
feel
really
post
industrial
in
my
place.
It's
it's
the
height
of
style,
sheep,
yeah.
Exactly
I'm
glad
you
get
that.
F
F
Yeah,
so
for
everyone
joining,
please
add
your
name
to
the
attendees
list.
Specification
is
very
light
today,
so
hopefully
some
people
have
some
things
to
go
over.
Otherwise,
I'm
gonna
give
you
a
lot
of
your
time
back.
We've
been
doing
a
lot
of
really
good
work,
so
I
kind
of
just
want
to
like
touch
on
that.
Let
me
see
what
kind
of
attendees
we
have
currently
on
the
participating
list.
Oh,
I
think
we
might
actually
have
everybody
cool.
F
So
I
think
one
of
the
first
things
I
wanted
to
follow
up
on
was
the
doodle
poll.
Let
me
share
my
screen.
F
Yeah
cool
so
last
week
anthony
had
suggested
the
fact
that
this
doesn't
work
for
a
large
number
of
people
on
the
call.
So
we
were
looking
for
some
different
times
to
gets
this
being
moved
to.
I
just
pinged.
F
Puny
was
one
of
them,
thanks
for
responding
by
the
way,
and
then
david
and
liz
and
josh
just
to
make
sure
that,
like
they
have
a
chance,
I'd
like
to
get
their
input
on
this,
I
think
one
of
the
people
we
were
trying
to
accommodate
it
was
liz,
so
it'd
be
really
cool
if
we
could
get
her
input
as
to
like
what
time
worked
for
her,
but
otherwise
I
think
that
we're
moving
on
this
one,
I'm
gonna,
leave
it
up
to
anthony
for
when
he
wants
to
close
the
poll
and
make
a
decision
steve.
H
F
H
Yeah,
no,
I
understand
I
totally
get
it
I'll
yield
if
need
be.
F
Yeah,
I
know
it's
turns
out.
Scheduling
is
not
the
easiest
thing:
okay,
cool!
That's
what
I
wanted
to
say
thanks
again
when
you
for
responding
and
anybody
else
on
the
call
who
I
didn't
mention
who
hasn't
responded.
Yet,
please
we,
your
opinion,
matters
to
us.
We
want
to
find
a
time
that
works
for
everyone
to
be
able
to
join.
F
These
calls
cool,
then
kind
of
jumping
in
kind
of
want
to
talk
a
little
bit
about
the
project
boards,
because
we've
had
a
lot
of
progress
in
the
past
week,
the
main
thing
being
the
1.0.1
spec
conformance
review,
we're
looking
really
good
here.
32
issues
done
five
in
progress.
I
know
the
one
that
I
still
have
in
progress
is
this
spam
exception
thing
which
is
going
to
take
a
little
bit
of
effort
which
I
was
hoping
to
resolve
today.
F
Anthony
still
has
one
in
progress
and
then
I
know
aaron
had
three,
but
I
kind
of
threw
those
atoms,
so
I
might
have
to
pick
those
up
because
I
don't
know
if
that's
really
fair,
but
hopefully
we
can
get.
Those
done
today.
Definitely
think
we
can
get
this
project
closed
this
week,
which
puts
us
in
a
really
good
position,
because
then
this
project
board
really
becomes
a
nice
accurate
depiction
of
when
we
can
expect
to
release
the
1.0
release
of
the
go
rc
so
cool.
F
I
just
want
to
say
thanks
everyone
for
contributing
to
this
or
put
up
with
me
asking
you
to
contribute
or
any
other
thing.
I
think
this
is
a
really
good
progress
and
we're
doing
really
good
on
this
one.
So
I
really
appreciate
it
cool
and
then
the
last
thing
that
I
had
that
I
wanted
to
talk
about
which
I
really
hope.
Somebody
else
has
something
to
talk
about,
because
it's
gonna
be
like
a
record
short
meeting.
Is
there
was
a
question
about
when
something
can
get
released?
F
We
had
a
fix,
come
in
for
a
memory
leak
and
looking
at
the
change
log
there's
a
considerable
amount
of
things
again.
It
turns
out
we've
been
pretty
active
in
the
unreleased
section,
so
we
may
need
to
do
another
patch
release
or
minor
release
before
we
actually
have
an
rc
for
the
stable
release
going
out.
I
didn't
know
if
anybody
else
had
any
opinions
on
this
or
if
they
wanted
to
get
their
pr
or
their
issue
resolved
to
get
in
here
I'll
shut
up
now
and
let
other
people
talk.
I
Yeah,
I
think
two
weeks
between
releases
is
a
pretty
good
cadence,
as
we've
talked
about
before.
So
it
seems
like
it's
probably
time
to
ship
another
one
with
a
roll-up
of
everything
we've
done.
I
don't
know
if
there's
anything
hugely
significant
in
terms
of
prs
that
are
close
to
merge
that
we
would
want
to
get
in
before.
I
But
maybe,
if
we
take
a
look
at
the
pr's
that
are
out
there,
you
know
the
removing
apply,
config
and
config
would
be
nice
to
land.
I
don't
know
where
we're
at
with
that.
F
We're
being
nitpicky
and
by
we
I
mean
me
with
this
test
coming
in,
I'm
really
hesitant
to
add
something
in
coming
in
with
timing
again,
given
the
fact
that
already
we
have
some
test
flights
that
need
to
be
readdressed,
but
I
think
that
we're
pretty
close,
I
think
that
the
responsiveness
from
a
song-
sorry,
I
don't
know
oh
inju,
which
I'm
probably
butchering
anyway,
so
I
probably
should
have
stuck
with
ij's
song
they've-
been
very
active,
so
yeah.
F
I
think
that
we
should
probably
be
able
to
get
this
done
today.
Hopefully
I
I
would
like
to
see
that
come
in,
because
I
agree,
I
think,
that's
a
nice
one
to
land
yeah,
but
other
than
that.
I
think
I
think
you're
right,
I
don't
know
if
there's
too
much
more,
it
needs
to
get
landed
here.
F
This
is
kind
of
something
that
probably
needs
to
get
addressed,
but
I'm
hoping
to
do
this
in
like
the
next
week,
especially
with
that
last
remaining
issue
that
I
have
for
spec
compliance
matrix
to
try
to.
I
don't
know
to
have
some
sort
of
plan.
That's
that's
the
goal
right
now,
I'm
still
very
confused
as
to
what
the
goal
here
is.
I
And
then,
over
on
the
contrib
side,
there's
the
aws
sdk
v2
that
I
would
like
to
get
landed.
I
think
I
had
one
last
request
on
that.
Pr
that
I
wanted
to
see
done
in
terms
of
the
the
timing
of
when
the
span
was
started
in
the
middleware,
but
if
other
people
could
take
a
look
at
that,
so
that
we
could
get
more
approvals
on
that.
That
would
be
great.
F
Yes,
I
will
try
to
get
a
review
of
this
before
I
do
the
release
and
we'll
try
to
get
this
landed.
I
know
rich
has
expressed
his
regret,
but
he
hasn't
done
a
review
yet,
but
I'm
extremely
confident
that
you
know
for
the
next
hour,
he's
gonna
have
it
in
yeah.
E
I
feel
awfully
late
to
the
party,
so
sorry
about
saying
I
do
something
and
then
not
doing
it.
It
feels
terrible.
F
Cool,
so
I
think
that's
good
for
the
release.
I
will,
I
think
I'll
see
about.
I
wonder
what
time
zone
injuries
in
oh
south
korea,
oof
yeah-
that
might
be
a
tough
one
to
get
done
today.
Then,
okay,
I'd
like
to
get
the
release
out
tomorrow
for
at
least
this
repo
and
then
maybe
next
early
next
week
at
the
control
repo
updated
as
well,
which
I
think
is
a
timeline
for
the
aws.
I
don't
know
if
it
gives
us
timeline
to
get
this
included
in
this
next
release.
F
But,
okay,
that's
up
to
me
I'll
figure
that
one
out
I'll
try
to
make
sure
this
is
communicated
in
slack
on
the
release
schedule.
But
this
kind
of
brings
down
to
another
open
question
and
I
think
I've
been
talking
with
punya
and
a
few
other
people
in
slack
about
something
that
this
introduces
because
we're
removing
the
apply
config
method
here.
F
The
thing
that
I
alluded
to
in
slack
about
us
being
able
to
associate
a
resource
and
that
resource
not
being
able
to
change
after
the
fact
will
not
be
the
case
in
the
go
implementation
anymore,
and
we
actually
would
be
conforming
with
this
idea.
That,
like
the
resource
is,
you
know
a
non-changing
thing.
However,
the
specification
doesn't
actually,
as
we
were
talking
about
in
this
life
channel,
have
a
normative
requirement
there,
and
it
has
a
cannot.
F
So
I
was
thinking
about
that
after
the
fact
that,
like
I
don't
know
what
to
do
there,
I'm
really
confused
on
that
one
espana
pointed
out.
I
haven't
actually
seen
like
the
response
to
my
response.
Sorry
about
that,
I
just
was
trying
to
get
ready
for
the
meeting,
but
yeah
it's
a
little
frustrating,
sometimes
working
through
the
specification,
because
some
of
these
things
are
written
in
a
little
bit
like
a
non
like
concrete
form
and
it's
hard
to
like
interpret
those
things.
F
So
I
don't
know
like
what
we
should
be
recommending
to
the
specification
at
that
point.
That
being
said,
it's
not
really
relevant
to
the
ghost.
I
don't.
I
don't
know
it's.
C
Okay,
I
do.
I
did
see
your
comments
punya
just
before
the
meeting
too.
This
is
steve.
I
have
a
vested
interest
in
being
able
to
mutate
the
resource,
but
I
also
think
that
the
spirit
of
the
spec
is
trying
to
tell
you
that
you
shouldn't
and
you
shouldn't
be
able
to-
I
think,
we've
just
kind
of
let
it
slide
accidentally
here.
C
Yeah
yeah,
I
think
it's
just
I
was.
I
was
using
resources
to
kind
of
emulate
this
feature
of
honeycombs
libraries,
which
I
think
you
know
anthony
where
they
have
like
a
add
field
to
trace
and
then
apply
the
same
field
and
appears
in
all
the
spans
in
the
trace
up
and
down,
and
you
could
maybe
you
could
do
that
with
a
span
processor
or
at
least
you
could
sort
of
emulate
resources
with
a
span
processor
yeah.
I.
I
I
C
Yeah
yeah,
so
it
I
think
it
makes
each
span
a
little
more
heavyweight
just
because
they're
not
sharing
structure
like
they
would
with
the
attributes
that
are
in
a
resource,
but
efficiency
aside,
I
think
the
net
effect
is
the
same,
and
then
the
author
of
that
spam
processor
could
decide
whether
it
wants
to
handle
the
possibility
of
changing
the
attributes.
You
know
after
it's
been
created,
it's
like
you
can
register
the
span
processor,
but
then
later
change
the
attributes
that
it
injects
into
each
span.
C
You
know
ideally
before
it
started
doing
its
work,
but
it
might
just
be
able
to
accommodate
this
initialization
order
that
I've
discussed.
I
H
C
I
You
would
you
would
register
the
spam
processor
with
the
trace
provider
so
that
all
of
the
traces
that
come
off,
that
provider
have
the
span
processor
and
then
have
some
other
way
to
get
a
handle
to
that
span.
Processor,
so
that
you
could
say
for
this
particular
trace,
given
a
trace
id
every
span
that
comes
along
on
this
trace.
Add
these
attributes
to
it.
C
Yeah,
I
wasn't
even
thinking
of
that
so
being
that
sophisticated,
but
that's
a
cool
idea.
I
was
thinking
of
just
you
know,
all
the
all
the
spans
that
go
through
it
and
add
the
attributes
that
it
knows
about,
but
allow,
maybe
even
after
the
fact
changing
that
set
of
attributes,
because
you
could
like
register
it,
but
not
quite
know
all
of
them
yet
and
then
you
know
shortly.
Afterward
tell
it.
Oh
sorry,
here's
the
attributes
I
meant
to
be
in
there.
C
You
know
we'd
have
to
handle
concurrency
and
all
that,
but
I
could
see
a
way
to
do
it.
Yeah.
F
Yeah,
I
think
that
makes
sense
right,
because,
if
like
you're
going
to
register
a
resource
with
the
trace
provider,
it
would
be
tall
in
many
ways
but
yeah.
I
think
that
that's
actually
a
really
cool
idea.
We've
had
onstart
here
and
I've
never
seen
a
spam.
Processor
really
use
it.
But
I
think
that's
a
really
cool
use
case
that
you
just
described.
C
Yeah
yeah,
I
actually
it's
funny
in
the
honeycomb
pollinator
slack
team.
I
shared
a
span
processor
that
I
wrote
that
was
trying
to
add
the
the
process
uptime
to
every
span
and
it's
dynamic.
So
you
know
it
has
to
be
f
spam
processor
and
I
I
had
missed
the
detail
that
in
onstart
you
can
read
the
end
time
of
the
span,
which
is
nonsense.
C
So
I
was
originally
using
that
trying
to
record
the
process
uptime
at
the
end
of
the
span,
forgetting
that
I
was
actually
doing
it
at
the
start
of
the
span
before
it
had
ended
and
it's
getting
lame
values
but
yeah.
So,
but
I,
if
I,
if
I
do
wind
up
writing
this
I'll,
be
happy
to
share
the
code.
Maybe
it'll
be
useful
in
general,.
F
Because
I
yeah,
I
think
it
would,
if,
at
the
very
least
as
like
an
example
test,
so
that
we
could
include
it
in
documentation,
if
not
like
you
know,
a
reference
to
an
implementation,
or
something
like
that.
Like.
F
Use
or
like
form,
kind
of
like
what
anthony's
talking
about,
I
think
it'd
be
really
cool
because
it's
just
like
you
know
really
quickly.
You
know,
even
even
like
a
non-current
current
safe,
like
example,
that
we
get
to
show
the
people
who
have
come
to
this
project
and
they're
like
how
do
I
update
the
resource
after
the
fact
we're
like
go
here,
you
do
this.
You
know
yeah.
C
F
Waiter,
so
are
you
saying
that
it's
not
representing
the
entity
producing
telemetry
for
shame.
F
Yeah,
it's
a
it's
a
common.
It's
a
common
point
of
contention.
I
think
in
the
project:
okay,
okay,
cool.
I
think
this
is
good.
I
think
punya.
Our
conversation
in
slack
may
just
end
with
a
slack
conversation.
I
don't
know
if
I'm
actually
gonna
file
an
issue
if
we
don't
actually
need
to
address
this,
but
if
anybody
else
thinks
that
I'm
just
skirting
around
actually
making
the
project
better.
You're
welcome
to
call
me
on
that
as
well.
F
Okay,
okay,
cool-
that
is
all
the
stuff
that
I
had
for
the
agenda.
Is
there
anything
else
that
anybody
would
like
to
talk
about
cool
customer
experiences?
Any
open
issues
that
you
have
you
want
us
to
pay
a
little
bit
more
attention
to
or
anything
like.
F
That
cool
david,
I
see
you
just
joined
the
call
a
little
late.
I
asked
I
didn't
slack
as
I'm
guessing
I'm
gonna
like
double
ping
you,
but
this
doodle
poll
here,
if
you
haven't
already
responded,
which
obviously
you
just
did
so
never
mind,
yeah
steve
man.
This
is
not
looking
good.
This
is
just
I
don't
know.
C
F
H
F
Well,
cool
this
sounds
good.
I
would
like
next
week
we
should
have
a
release
out.
We
should
have
potentially
a
new
time.
It
looks
like
likely
going
to
be
a
new
time
for
these
meetings
and,
ideally
the
spec
compliance
conformance
review
should
be
closed
and
we
should
just
be
working
off
this
up
until
elementary
go
rc.
F
I
would
like
it
if
next
week,
if
you
come
to
the
meeting-
and
you
have
some
allocated
time
from
just
like
the
week
in
front
of
you-
maybe
to
understand-
like
an
issue
that
you
could
pick
up
from
the
open
summary-
go
rc
project
board,
and
maybe
you
can
just
kind
of
like
identify
what
people
are
going
to
be
working
on
in
the
coming
week.
That
would
be
two
weeks
from
now.
So
just
kind
of
give
you
a
heads
up,
if
you
don't
have
any
time
I
understand.
F
Also,
this
is
an
open
source
project.
Still,
please
join.
I
love
having
contributions
in
the
project,
even
if
you
don't
have
time
to
allocate
to
working
on
a
particular
issue
in
a
week
but
yeah.
If.
B
F
Do
I
would
love
it
if
you
could
look
at
some
of
the
open
issues
and
pick
something
up
would
be
really
cool
but
other
than
that?
Actually,
I
feel
like
yeah,
just
in
general,
that's
always
a
good
option,
but
I'd
like
to
maybe
just
go
through
that
next
week.
That's
just
like
a
standard
practice
to
make
things
come
forward.
Other
than
that,
though,
I
think
that
that's
probably
probably
called
a
meeting
early,
I'm
going
to
hand
it
over
to
anthony
to
see
if
he
has
objections
to
that.
I
No
just
to
to
wrap
up
on
the
the
both
of
the
new
meeting
time
looks
like
most,
people
have
responded.
No
david
and
liz
have
responded
as
well.
I
think
I
pinged
ted
on
the
the
slack
as
well
to
try
to
get
his
input,
but
I
propose
we
wrap
that
up
and
make
a
decision
tomorrow
on
a
new
time
so
that
we
can
schedule
next
week
at
the
new
time.
F
Yeah,
I
agree:
let's,
let's
do
that.
We
can
just
do
that
in
the
slash
channel.
Anthony's,
sound,
good,
okay,
cool
yeah
and
I
have
access
to
that
calendar.
I
don't
know
if
you
do,
I
think
you
should
be
able
to
have
access
to
the
calendar,
but
if
you
don't,
I
could
definitely
adjust
the
discussion.
I.
I
I
don't
think
that
I
do
so
if
you
could
handle
that
that'd
be
right.
F
Okay,
cool
okay:
I
can
do
that
awesome.
Oh,
that
was,
I
totally
almost
forgot,
so
puny
and
david.
One
thing
that
I
did
address
in
the
past
week
was
the
go
vanity
urls.
They
were
originally
set
up
by
people
like
google.
I
I
think
this.
G
I
think
probably-
or
I
mean
I
I
think
I
know
who
they
likely
are.
Is
there
a
problem
with
them
right.
F
F
Yeah
there
we
go
so
on
the
deploy
we
have
this
going
out
and
it
looks
like
the
project
in
google
cloud
is
not
actually
yeah,
we
don't
have
permission
to
set
up
the
apis
and
it
looks
like
the
like.
The
permissions
on
this
repo
essentially
have
been
like
dropped.
I
don't
know
what
happened
in
this
particular
one,
but
I'd
love
it
if
we
can
get
somebody
who
has
permissions
or
has
access
to
this
to
take
a
look
at
this.
G
G
No,
he
pinged
us
or
he
pinged.
He
ended
up
pinging
me,
I'm
chasing
it
down
at
the
time
when
he
pinged
there
was
also
a
github
actions
outage
or
a
circle
ci
outage,
and
so
that
kind
of
shadowed.
This
other
problem,
so
I
said
well,
let
me
wait
for
the
circle
ci
outage
to
end.
I
I
didn't
come
back
to
this.
So
actually,
would
you
mind
re-triggering
that
build.
I
just
want
to
see
that
it's
if
it's
still
going
on,
are
you
able
to
trigger
that
book.
G
F
G
F
F
Would
you
need
admin
or
maintain
maintenance.
G
I
I
think
whatever
is
appropriate.
I
don't,
I
don't
know
exactly
what
I
need.
So,
if
okay
yeah,
I
think
the
key
part
here
is
the
sensitive
ones
right
like
the
project
id
and
stuff
like
that.
F
G
G
I
think
that
looks
so
that's
without
access
to
sensitive
x.
So
that's
going
to
be
things
like
project
management
in
the
github
issues.
Sense.
G
I
Confirm
with
bogdan
since
there
that.
G
F
That
sounds
good
and
then
should
I
put
you
as
the
the
point
person
punya
or
david
as
well.
F
All
right
david
drop,
the
hook
cool.
I
will
ping
bogdan
right
after
this
and
see
if
I
can
address
an
admin
to
this,
and
I
will
hit
you
up
in
slack
afterward
as
well,
so
you
can.
F
Did
you
want
me
to
rerun
this
as
well?
If
you
can,
yes,
that'd
be
great.
Okay,
I've
done
this
before,
and
I've
completely
failed,
so
yeah.
G
F
A
G
David,
I
can
add
you
to
the
thread
that
bogdan
started.
It's
from
last
week.
F
Thank
you,
okay,
cool
thanks
for
taking
a
look
at
this
by
the
way
I
appreciate
it,
and
I
will
pick
pogba
right
after
I
get
out
of
the
so
any
time.
F
Okay,
now,
okay,
all
right!
Everyone
thanks
for
joining-
and
I
will
see
you
all
next
week.