►
From YouTube: 2022-06-02 meeting
Description
cncf-opentelemetry@cncf.io's Personal Meeting Room
A
C
B
Could
tell
it
like
okay
for
every
occurrence
on
the
day
or
at
the
time
start
of
the
meeting
or
an
hour
before
or
whatever
like
look
here's
the
new
thing,
then
that
would
be
much
more
useful.
Just
as
a
complete
aside,
morgan,
you
use
git
pod
right.
A
A
It
was
good,
I
mean
the
core
problem.
I
often
have
developing
from
github
or
really
any
open
source
solution
which,
which
is
what
I
ran
into
this
time,
was
often
the
builds,
don't
work
because
the
environment
isn't
set
up
correctly
and
git.
Pod
does
not
solve
that
problem
because
it
just
creates
a
new
sort
of
blank
environment
on
an
ubuntu.
What
I
assume
is
ubuntu
vm
yeah,
but
in
terms
of
quickly
getting
started
and
not
having
to
run
a
bunch
of
crap
on
my
computer
for
like
a
five
minute
edit,
that
was
nice.
B
A
B
A
B
A
B
B
B
A
C
B
Just
being
here,
great
topic,
we'll
add
fix
blog
instructions.
B
B
Yeah,
maybe
throw
a
no,
maybe
ping
him
in
the.
If
someone
wants
to
ping
him
in
the
slack.
B
A
C
B
B
My
favorite
was
still
drop,
sonde
or
drop
sonde
d-r-o-p-s-o-n-d-e,
which
is
a
particular
type
of
weather,
beacon
and
it's
called
a
drop.
So
it's
sande
sonde,
which
is
a
like
weather,
telemetry
thing,
and
it's
a
drop
sunday,
because
it's
something
that
you
literally
drop
into
a
hurricane
or
whatever
have.
B
A
A
It
awesome
was
it
lightstep
the
two
years
ago
they
had
an
april
fools
thing
saying
they
bought
a
logging
company
and
they're
like
they
had
this
fake
website
of
like
a
forestry
like
like
a
timber
like
logs
chopping
company,
that's
saying,
they'd
acquired
them.
B
A
B
B
And
then
we
did
it
last
year.
I
think
so.
We
we'd
already
done
the
website.
We
already
we
had
a
ton
of
stuff
planned
for
this.
We
had
a
beard
growing
contest
to
see
who
would
be
the
mascot.
A
D
B
I
don't
so
hey,
philip
okay,
so
hello,
someone
is
here
that
has
a
gender
item,
so
let's
get
going
first
thing:
what
exactly?
What
for
the
blog
instructions?
What
is
the
problem
that
we
don't
need
to
tell
you
that
you
have
to
build
it.
B
A
Yeah
for
for
the
blog,
it's
yeah,
the
add
a
blog
post.
Instructions
today
are
like
go,
deploy
the
whole
website
and
build
it
locally
and
then
run
a
script
that
will
automatically
generate
a
new
blog
post.
D
B
B
C
C
B
Yeah,
I
think
that's
a
good
thing
to
check
out.
If
we
can,
I
don't
see
a
reason
why
we
couldn't
we'd
have
to
have
a
github
action
for
it
or
something
to
or
something
we.
You
know
what
there's
probably
a
way
to
do
that
to
there
had
to
be
some
level
of,
like,
I
guess,
netlify
and
or.
B
B
Do
it
on
demand?
The
one
thing
you
can't
do
is
with
hugo:
if
you
build
a
site
and
you
have
a
post
in
the
future,
then
it
won't
actually
build
it
in
it's
like
if
you
future
date,
your
blog,
when
you
write
it
and
then
you
hit
build,
it
won't
show
up
because
it
skips
it
because
it
doesn't
get
it's
happening
in
the
future.
But
because
hugo
is,
you
know,
static
site,
it
just
never
gets
built
yep,
so
yeah.
We
need
to
yeah.
That's
a
good,
patrice
question.
B
B
Some
personal
home
page
philip-
you
want
to
go
through
yours.
F
Yeah
was
a
thing.
Sorry,
if
there's
noise,
I
took
my
car
to
the
car
fixer
people.
Now,
I'm
outside
at
a
coffee
shop,
yeah
cubecon
was
the
thing
there
were
a
shitload
of
people
there.
I
think,
like
the
biggest
takeaway
that
I
had
like
when
I
was
working,
the
honeycomb
booth
was
like
the
top
question
that
I
got
was:
do
you
support
open
telemetry
and
I'm
like
okay?
F
So
you
know,
okay,
what's
the
reason
for
the
question
and
like
every
single
person
was
like
either
we're
already
using
it
or
it's
on
our
roadmap
to
use
it,
and
we
know
that
this
is
what
we're
going
to
do.
I
think
that's
pretty
huge
and
when
I
asked
them
like
continued
follow-up
questions,
they
were
like
yeah.
We
don't
know
like
necessarily
how
we're
going
to
use
it
or
they're.
Like
you
know,
how
do
I?
How
do
I
worry
about
like
sampling?
F
How
do
I
worry
about,
like
you,
know,
data
being
duplicated
and
deduplication?
If
I
send
it
to
you,
know
a
vendor
or
something
like
that,
is
it
possible
to
send
to
you
know
they
would
ask
honeycomb,
but
then
they're
like
you
know,
could
I
send
it
to
the
other
people
who
are
here
too
and
like
try
them
out
like?
F
Yes,
it's
by
design,
lots
and
lots
of
questions
that,
like,
I
think
we
would
perhaps
consider
basic
in
nature
that
I
think
a
lot
of
the
attendees
just
didn't
really
know
about,
but
I'm
curious
about
other
people's
thoughts.
A
We
like,
similarly,
I
was
at
the
splunk
booth,
the
decent
amount.
We
had
a
lot
of
people
coming
by
and
asking
about
open
telemetry
to
the
point
where
we
had
what
I
assume
are
desperate
sales
people
coming
by
from
other
from,
like
non-observability
vendors
being
like.
Oh,
I
hear
you
guys,
you
you
contribute
to
open
telemetry,
would
open
telemetry
be
interested
in
using
our
proprietary
database
like
using
it
like
paying
us
money?
A
I'm
like
I,
don't
what
it's
not
that
kind
of
project,
but
yeah
like
tons
and
tons
of
interest
more
than
I've
ever
seen
at
kubecon,
which
is
really
exciting.
A
D
I
had
a
conversation
with
like
a
very
smart
engineer,
who
was
like
yeah
yeah
we're
trying
to
adopt
open
telemetry,
but
we're
like
it's
like
really
confusing
like
it's
like
you
know,
I've
spread
all
the
like
hype
about
like
captures
all
these
signals
and
does
all
this
stuff,
but
it's
like.
I
can't
find
like
like
the
database,
you
know
or
like
like
how
do
I
run
it,
and
I
I
was
like?
Oh
no.
No,
it
doesn't
work
like
that.
It's
just
just
the
telemetry
system.
A
D
B
I
wanna
specifically
acknowledge
adriana
for
getting
that
like
docs.
Revamp
start
kicked
off
with
the
conceptual
stuff
because
that's
been
very
good
and
I
think
that's
like
super
needed
and
welcome.
B
B
Is
it
that
there's
not
enough
direction
on
like
what
docs
are,
but
basically,
where
is
the
actual
like?
What
is
the
sharpest
of
the
sharpie
edges
here?
Is
it
like
a
pm
thing
that
we
don't
have
someone
very
clearly
saying
here's
the
work,
industrial
chunks?
Is
it
a
sort
of
you
know
doing
the
stuff
part
where
there's
just
no
one
who
is
dedicated
to
sit
down
and
do
it?
Is
it
something
else.
B
D
D
I
I
think
we
need
more
people,
writing
docs
for
sure.
I
think
we
have
a
general
game
plan
about.
You
know
what
docs
need
to
be
written
right
like
we
know
we
need
getting
started
guides.
We
know
some
of
our
like.
D
Maybe
some
of
the
flow
of
the
site
needs
to
be
improved
to
get
people
like
straight
into
that
stuff,
but
yeah
you
just
gotta
do
it.
The
other
thing
is
like
you
know,
metrics
and
logs
are
are
kind
of
like
in
progress.
D
It,
like
you
know
like
well
one
way
to
organize:
it
is
just
by
language
right,
like
sort
of
a
as
a
group
or
or
a
team
of
people.
You
know
we
just
say
we're:
gonna
just
try
to
make
go,
super
pretty
and
then
java
super
pretty
and
then
dot
net
super
pretty.
D
B
B
B
Here
is
the
minimal
set
of
getting
started
stuff
and
that's
fairly
well
represented
it's
not
complete
by
any
means,
but
once
you
get
past
that,
then
it's
very
much
like
here
be
dragons
right,
like
I
think
someone
that's
also
complaining
about
like
how
do
I
use
the
stackdriver
exporter
for
rust,
I'm
like
and
I
go
and
look
and
it's
like
there's.
The
only
docks
are
like
literal
api,
auto
gen
docs,
and
it's
like
this
is
the
constructor
to
the
the
exporter.
You.
B
D
I
think
rust
is
like
this
is
kind
of
like
I
mean
we're
moving
past
this
a
lot
of
language,
but
I
think
rust
is
like
prototypical
here
you
you
come
here
and
it
says
like
yadda
yadda,
there's
some
pieces
whatever
they
are:
here's
a
bunch
of
packages,
okay,
here's
the
docs
and
then
you
go
the
docs
and
then
there's
getting
started,
but
getting
started
is
focused
on,
like
here's
like
a
basic
example
of
like
how
to
make
a
span-
and
I
would
say
like
this-
is
most
of
our
docs
kind
of
followed.
D
D
So
like
that
that
that's
the
the
like,
how
do
I
install
this
and
like
get
it
running
and
then,
like
you
know,
like
that's
the
piece
that
that
we've
got
a
lot
of
this
on
the
lightstepdoc
site
and
we
can
port
that
over.
But
we
definitely
don't
have
rust
stocks
for
that
yeah.
So.
B
What
lot
you
know,
what
are
you
using
for
automatic
instrumentation
right
like
it's?
Not
we
just
there's
the
surface
area
of
open
telemetry
is
extremely
large,
and
so,
if
we
were
gonna
say
like
okay,
here's
how
you
install
let's
pick
go
even
as
something
that
is
pretty
simple.
It's
like
you
still
have
to
have
like.
B
Okay,
here's
how
you
set
it
up
and
then
here's
how
you
use
it
for
net
http
or
here's
who's
it
for
gorilla
or
here's
how
you
use
it
for
sql,
client
or
here's,
how
you
use
it
for
foo
barb
as
because,
like.
B
There's
a
lot
of
stuff
that
ends
up
being
extremely
duplicative
in
terms
of
content
where
it's
like
you're
writing
the
same
exact
thing,
40
billion
times,
because
you
need
like
here's,
how
you
install
open
telemetry
for
spring
on
java.
Here's
how
you
install
open
telemetry
for
and
that
to
me
is
a
like
that's
a
hands
problem
where
we
just
don't
have
enough
people
to
go
and
create
the
40
pages.
We
need
per
language
for
every
single
integration.
C
I
mean
the
details
are
different
per
language,
but
there's
some
introduction
text
and
I
see
like
I
think
there
was
something
like
hey
here-
is
on.net
and
here
is
on
on
java
and
here
and
it's
like
the
same
thing
again
and
again
and
then
the
only
thing
that's
different
is
the
code.
So
the
question
is
like:
is
there
in
in
hugos
a
way
to
yeah
to
create
some
some
snippets
that
I
say
like
add
explanation
what
what
otlp
is
or
whatever
I
I
don't
know
so
this
yeah
already
so.
B
F
So
I
would
propose
that
there
is
also
a
bit
of
a
pme
kind
of
problem
here
which
is
going
to
this
this
route
of
so
I
talked
with
several
of
my
sales
engineers
pretty
frequently,
and
they
were
the
ones
who
raised
the
issue
back
in
like
october
with
me,
where
they're,
like
okay
people
literally,
don't
even
know
how
to
get
started
with
this
thing,
like
okay,
cool
I'll,
see
if
I
can
contribute
some
and
improve
that
and
then
they're
like
yeah.
That
problem
is
not
as
bad
now.
F
The
problem
now
is
like
people
are
like
cool.
I
can.
I
can
get
started,
but
I'm
using
this
specific
library
and
there's
zero
docs
for
like
even
how
to
like
wrap
a
handler
for
a
thing
in
a
given
language,
let
alone
like
detailed
considerations
about
like
oh
here's,
how
like
it
collects
instrumentation
for
this
library
in
this
way
and
that
kind
of
stuff-
and
I
mean
to
me-
that's
that's
totally
a
prioritization
problem
like
what
are
the
most
used.
Instrumentation
libraries
and
stack
income,
be
like
all
right.
F
B
B
B
B
Work
in
devrel,
so
I
already
know
the
answer.
Okay,
I
don't
want
to
while
away
on
this
topic
too
long.
Did
you
so
the
one
thing
I
do
think
from
a
recap
perspective.
B
That
would
be
good,
is
I
wasn't
there
obviously,
but
like
ted
was,
and
then
I
don't
know
if
shower
was
some
of
the
end
user
working
group,
people
morgan,
you
were,
it
would
be
really
good
to
get
like
the
people
that
were
there
to
maybe
coordinate
in
slack
to
do
sort
of
a
wrap
up
post
or
something
for
the
blog,
specifically
around
sort
of
the
learnings.
From
like
the
end
user
researchers.
I
think
that
was
very
generally
interesting.
So
morgan,
can
you
sort
of
take
point
on
that
cool
beans.
A
Yep
I'll
connect
with
hand
wreck
at
dinotrace
as
well,
because
I
know
he's
taking
those
doing
those
video
interviews
and
I
haven't
I
haven't
heard-
I
mean
yeah.
A
D
D
I
paid
we
were
using
a
contracting
agency
and
I
just
went
to
them
and
said:
like
hey,
you
give
us
people
to
interview.
Can
I
like
pay
your
people?
D
If
you
are
like
in
your
interview,
pool
to
like
you
know,
pay
them
for
an
hour
or
two
of
their
time
to
like
sit
with
me
and
like
walk
through.
You
know
some
stuff
and
they
said
yeah
sure,
and
so
I
walked
them
through.
This
is
trying
to
get
started
with
open,
telemetry
and
kind
of
rearrange
the
pieces
until
it
seemed
like
they
were
able
to
kind
of
move
through
it
without
getting
stuck
because
something
they
needed
to
know
was
like
down
the
bottom.
D
This
doesn't
cover
metrics,
obviously,
but
I
think
we
could
we.
We
have
some
of
these
on
the
light
step
site,
but
we
could
go
to
the
sigs
with
this
template
and
be
like
yo.
Can
you
fill
this
out.
B
D
B
Like
a
markdown
template
create
a
give
up
issue
with
this
in
it
and,
like
I
think,
the
languages
we're
kind
of
focusing
on
right
now
are
we're
trying
to
prototype
stuff
in
java
right
because
that's
sort
of
or
putting
the
improvements
into
java,
because
that's
remains
very
popular.
D
Yeah,
god,
you
zoom
get
out
of
my
way
so
getting
started.
Template.
D
B
Yeah
I'll
throw
this
in,
I
think,
but
also
I
actually
kind
of
want,
like
phillips
next
point
about
manual
context,
propagation.
I
think
that's
also
something
that
we
should
think
about
with
this.
F
I
would
agree:
it's
come
up
enough
times
with
customers
that
are
working
through.
That
should
probably
be
somewhat
moving.
B
B
F
Services
or
within
a
service
at
least
the
context
of
that
it's
come
up
in
the
most,
and
this
is
also
capturing.
The
github
issue
is
someone's
like,
oh
well.
I
can't
install
the
http
instrumentation
library
some
number
of
reasons,
but
I
want
to
propagate
a
trace
across
my
services.
F
Part
of
it-
I
I
I
I
don't
know
if
it's
necessarily
part
of
the
getting
started
flow,
I
mean
it.
It
certainly
will
be
for
like
some
people,
but
I
think
for
the
majority
for.
A
Very
advanced
people
yeah,
I
agree
like
I
would.
I
would
expect
that
when
people
are
getting
started,
the
just
tell
them
yeah.
We
propagate
your
context
and
it's
not
magic
like
here's,
a
link
to
further
information,
if
you
really
care,
but
for
a
lot
of
people
when
they're
getting
started,
they
probably
just
care
that
it
works
more
than
how
does
that
make
sense.
Yeah.
A
Yeah
because,
like
I
would
just
link
them
into
a
deep
dive
on
context,
in
fact,
and
the
challenge
with
that
is
of
course,
like
technically
the
context
format
we
use
the
default.
One
is
the
w3c
format
right,
which
is
technically
a
different
project
and
has
its
own
docs,
but
we
could
also
document
it
here.
I
don't
really
care
yeah.
We
definitely
can.
D
Need
to
tell
people,
though
so
people
need
to,
I
think
we
need
to
have
a
good
high
level
page
under
concepts
about
context
propagation
breaking
it
down
into
like
you
know
in
process
context
and
then
like
serialization
and
propagation,
because
this
is
like
the
one
big
new
concept
for
people.
D
Everyone
gets
all
the
rest
of
it
it's.
This
is
the
new
concept
and
when
it's
working,
it's
like
great,
but
when
it's
not
working,
which
usually
pops
up
right
at
installation,
because
you
have
some
homegrown
framework
or
you
know
something
something,
and
so
it
is
broken
in
there
somewhere.
That's
the
moment
that
people
need
to
now
kind
of
like
understand
some
of
these
concepts,
so
yeah
a
good
high
level
page,
and
then
you
know
some
docs
about
how
to
troubleshoot
your
your
spans
not
being
linked
together.
B
Keep
in
mind
we
do
have
so
yeah,
I'm
looking
now
so
baggage
is
in
baggage
is
under
sigma
is
the
same
under
signal
concepts
and
then
that
links
to
context
prop
in
the
tracing
thing
so
yeah.
Maybe
there
should
be
a
like
general
concept,
page
for
context
and
then
under
instrumentation
we
would
have
a
manual.
F
But
but
that
yeah,
I
think
we
don't
have
conceptual
detailed
language,
agnostic,
conceptual
docs
that
talk
about
context
and
its
uses
and
some
of
these
scenarios.
Oh.
F
A
given
language
will
have
it
be
implicit
versus
in
some
languages
like.
Oh,
you
gotta
pass
this
thing
around
all
the
time
and
like
if
you're,
in
this
scenario,
where
you
can't
use
an
instrumentation
library
which
propagates
it
across
services,
then
you
gotta
do
these
other
things.
Please
refer
to
the
language,
specific
docs
and
how
you
do
that
in
your
given
language,
like
that
stuff,
yeah.
B
Okay,
I
think
I've
captured
the
two
big
things,
which
is
a
conceptual
page,
and
then
individual
pages-
patrice
isn't
here
but
I'll
take.
This
is
a
thumbs
up
thumbs
down.
B
B
Okay,
the
one
thing
not
as
a
caution
but
as
a
like
as
just
a
thing.
This
is,
if
you
do
mark
down
linters,
that
enforce
line
length
is
very
good
for
making
epr's
easier,
but
very
bad
at
when
you
want
to
change
things.
B
C
The
thing
I
wanted
to
call
out
is,
I
think
we
had
like
four
or
five
pull
requests
over
the
last
week
where
people
just
had
typos
and
I
think
it
every
time
requires
a
maintainer
to
to
merge
this
back
into
into
the
project.
So
I'm
not
sure
if
there's
a
smarter
way
to
say
like
it
is
just
a
typo,
just
yeah
fix
that
automatically
or
whatever.
B
Yeah
I
mean
I
think
there
is
like
there
are
automatic,
merge
rules
that
could
be
set
up,
but
I
think
it
would
be
easier
also
to
just
enforce
spell
check
through
the
linter.
C
That
was
something
I
was
so
so
I
had
happily
encouraged
what
one
of
my
colleagues
to
to
post
something
on
the
blog.
But
of
course,
if
we
had
internally
some
discussions
like
are
we
going
to
put
this
on
the
company's
blog
or
the
open,
telemetry
blog?
C
So
I
was
wondering
if
it
would
be
good
for
for
people
to
have
at
least
like
a
link
to
the
linkedin
or
twitter
profile,
or
maybe
also
information
like
this,
and
that
person
working
for
this
and
that
company,
so
so
making
it
a
little
bit
more
easier
to
convince
people
internally
to
to
put
things
on
the
open,
telemetry
blog
versus
some,
some
company
blogs.
B
B
Yeah
I'll
put
it
on
here
and
ask
patrice
he's.
I
think
that
would
be
a
templating.
We
need
a
template
change
to
add
that
in,
but
we
could
probably
do
something
where,
in
the
front
matter
for
a
blog,
we
could
have
like
some
options
for,
like
social,
you
know
handles
or
emails
or
short
by
and
then
short
bio
and
then
throw
that
in
at
the
footer
and
then
make
like
the
author's
name,
clickable
or
whatever,
to
jump
to
it.
So
I
think
it's
a
good
idea.
B
Last
thing:
I've
got
is
a
quick
thumbs
up
thumbs
down,
but
I've
asked
our
designers
to
do
a
pride
month.
Treatment
for
the
hotel
logo.
B
So
they
said
they
should
have
by
friday.
If
that
the
only
question
I
had
was
like
should
we,
I
was
thinking
because
I
think
the
cncf
one
they
replaced
their
navbar
logo
with
the
so
the
variant,
and
so
I'm
thinking
do
the
same
thing
for
the
hotel
site
just
swap
out
the
navbar,
the
white
one,
with
the
pride
colors
rather
than
the
hero,
one.
F
Oh
actually,
one
one
thing:
there
was
an
issue
I
didn't
put
this
in
the
dock.
I
should
have
done
that
there
was
an
issue
filed
by
one
of
the
approvers
who
I
think
works
for
I
forget
his
name.
Works
for
spelling
is
involved
in
a
lot
of
the
projects,
as
he
said,
approvers
to
this
repo
should
not
have
merged
capabilities,
and
apparently
we
do
I've
been
not
merging
prs,
even
though
I
can
approve
them.
I
don't
know.
B
E
B
If
it's
something
that
is
kind
of
like
broad,
then
I
would
open
an
issue
and
sort
of
ping
people
in
the
comms
slack
channel.
Just
if
that's
something
that
needs
discussion,
if
it's
sort
of
more
like
I'm
fixing
the
readability
of
this
or
like
I'm,
adding
I'm
adding
more
examples
or
I'm
restructuring
content
in
a
way
that
doesn't.
B
It
doesn't
require
a
lot
of
cross-cutting
changes.
Then
I
think
it's
probably
fine.
Just
do
a
pr
we're
there
aren't
so
many
people
working
on
it
and
their
work
is
not
so
cool,
no,
so
unnecessary
of
being
coordinated.
B
B
We
should
change
how
we're
doing
x,
y
or
z,
then
that
I
think,
would
be
an
issue
first
to
kind
of
build,
not
consensus
necessarily
but
at
least
to
get
input
from
people.
B
F
B
B
45
feels
like
a
good
yeah
feels
like
we
aren't
super
rushed
and
we
got
through
everything
so
cool
thanks.
Everyone
for
showing
up
I'll,
see
y'all
on
the
interwebs
and
again
we're
always
in
slack
so.