►
From YouTube: 2021-04-08 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).
B
I'm
good,
I
don't
think,
we've
met
before,
or
maybe
we've
been
in
meetings
with
lots
of
people
but
nice
to
meet
you.
Thank
you.
A
D
C
D
Yes,
steve
is
indeed
speaking.
C
I've
been
out
of
the
loop
send
me,
send
me
more
details.
I'll
join.
D
Yeah
the
way
paul
bruce-
and
I
structured
it-
we're
doing
essentially
one
day
of
people
talking
about
observability
adoption
use
cases
and
then
the
second
day
is
kind
of
a
community
open
telemetry.
How
do
you
instrument?
How
do
you
integrate
with
other
technology
cool
yeah?
We
should
have
the
speaking
agenda
published
today
or
tomorrow,
but
we've
got
the
list
of
speakers.
So,
let's
just
not
who's.
Speaking
in
what
order
on
what
subject.
D
Yeah,
of
course,
and
yana
is
keynoting,
which
is
exciting.
That.
D
E
C
F
C
A
D
A
C
I
think
liz
you're
good
to
go.
D
All
right,
so
I
had
a
meeting
with
the
with
the
executive
director
of
the
continuous
delivery
foundation,
as
well
as
one
of
their
tc
members,
and
they
wanted
to
understand
how
to
better
tie
the
story
of
observability
and
telemetry
to
their
existing
set
of
projects
and
their
existing
remit.
D
So,
basically,
as
they
outlined
their
remit,
they
were
originally
created
when
google
and
netflix
donated
spinnaker
to
the
linux
foundation.
So
that's
where
the
continuous
delivery
foundation
came
from
and
now
also
jenkins
x
has
been
donated
to
the
continuous
delivery
foundation
and
basically
a
couple
of
interesting
areas
of
overlap
that
I
was
able
to
find
with
them.
First
of
all
that
you
know
they
should
be,
they
should
be
emitting
open,
telemetry,
formatted
data
for
anything
that's
kind
of
trace
formatted.
D
As
far
as
you
know,
hey
here
is
you
know,
a
process
that
is
happening
right,
like
you
know,
start
end
of
the
starting
end
of
the
process.
You
can
look
at
it
as
a
trace
kind
of
similar
for
metrics
similar
for
their
logs.
D
D
Can
you
use
it
to
generate
a
kind
of
you
know
procedural,
build
or
like
abort
the
release
signals,
which
are
you
know,
not
a
telemetry
thing,
but
instead
already
what
you
do
with
the
telemetry
after
analysis,
and
then
we
also
talked
about
the
subject
of
how
to
consider
the
signal
type
that
is,
a
change
has
happened
to
a
system
right,
a
deployment
has
rolled
and
there
is
a
new
host
right.
D
Like
events
like
that,
where
I
know
that
new
relic
has
been
calling
them,
events
honeycomb
calls
them
markers
right,
like
kind
of
there's.
This
need
for
a
short
understanding,
around
kind
of
what
is
a
a
change
in
your
system
and
right
now
we
don't
have
that
signal
in
otel,
but
it
sounds
like
the
continuous
delivery
foundation
is
working
on
that
problem,
so
kind
of
figure
once
they
standardize
something
it'll
be
interesting
to
make
sure
that
it
interoperates
with
kind
of
our
various
clinics,
propagation
and
tagging
systems.
D
A
Can
yeah
that
makes
sense,
I
got
to
say
for
the
hotel
stuff
I
just
given
the
conversations
we've
had
about
scope
creep
in
the
project
in
general.
I'm
not
super
excited
about
biting
off
the.
How
do
we
hook
observability
into
cd
like
the
full
thing?
I
absolutely
support
the
idea
of
honeycomb
doing
something
like
that,
but
that's
very
different
than
hotel,
because
it's
like
once
we.
A
A
D
A
Definitely
think
that
cd
right
now
is
hard
to
do
with
observability
because
of
a
lack
of
semantic
conventions.
Obviously
there's
some
stuff
around
just
locking
down
the
tagging
for
versions
and
stuff,
which
I
think
is
already
part
of
the
semantic
spec.
But
things
like
like
we've
done
a
lot
of
kind
of
painful
work
at
lightstep.
A
That
would
be
a
lot
less
painful
if
we
could
look
at
some
kind
of
tagging
for,
like
you,
know,
various
forms
of
cd,
whether
it's
like
canary,
deploys
and
percent
roll
out
information
and
blue
green
deploys,
and
things
like
that
would
just
be
amazing
if
that
stuff
can
get
standardized,
and
that
feels
like
something
that
could
easily
happen
in
this
sort
of
collaboration
and
like
it's
just
a
matter
of
writing
down,
I
don't
think
anyone
really
hopefully
has
super
strong
opinions
about.
A
G
Yeah
yeah,
I
mean
liz
in
your
discussions.
Did
you
also
discuss
you
know?
Perhaps
them
you
know
doing
the
work
on
their
end
to
in
the
initial
discussions,
but
also
kind
of
submitting
a
potential
otep
on
our
end,
so
that
it's
there?
You
know
as
a
discussion
issue
on
on
our
backlogs
and
then
you
know
if
there
is
ever
any
critical
mass
on
the
project
to
really
drive
some
of
that
work
forward.
D
I
think
they
have
something
that
they
need
to
do
on
their
end,
to
define
what
the
problem
is
from
their
perspective
right
and
then,
after
that
we
can
have
a
discussion
about
where
to
solve
it
so
yeah.
Basically
that
was
like
you
know.
It
was
a
very
preliminary
initial
discussion
and
you
know
I
don't
act
alone.
I
act
on
you
know
when
I'm
acting
on
behalf
of
of
hotel,
I'm
transparent
with
the
hotel
community,
so
yeah
so.
D
Let's
take
this,
to
you
know
an
otap,
or
at
least
like
a
a
feature
or
community
discussion,
but
for
now
you
know,
let's,
let's
see
what
they
come
up
with.
H
So
I
think
I've
popped
the
next
item
on
it,
but
I
don't
actually
know
what
I
don't
actually
know
what
question
I'm
asking.
So
let
me
read
you
the
question
I
was
asked
so
honorable
from
aws
mentioned
last
week
that
he
and
ted
were
chatting
about
possibly
hosting
vendor
java
agent
distros
in
hotel
repost.
H
But
that
was
just
a
theory
and
then
we
I
wanted
to
then
the
question
is
vendor
distro
pieces
and
so
on
within
the
hotel
repository.
So
I
mostly
want
to
understand
what
the
what
the
idea
is,
what
the
intent
was
and
then
what
our
thoughts
are.
I
haven't
before
seen
vendor
distros
inside
a
project
community
morgan.
Are
you
shaking
your
head
at
me
or
shaking
your
head
at
something
else?.
C
I'm
and
I'm
curious,
like
we
have
our
own
distro
splunk,
but
it's
just
because
we
throw
in
fluid
d
and
some
old
things.
We
haven't
converted
open
telemetry.
Yet,
but
in
like
six
months,
it's
just
gonna
be
basically
like
yeah,
it's
a
fork
of
ot
that
we
update
every
week
or
so
and
it's
the
one
we
support.
G
G
Mean
I'd
like
to
better
understand,
I
mean
sarah,
you
bring
up
a
good
point
and,
and
you
know
again,
I
think
there
have
been
many
discussions
which
ted
has
had
with
different
vendors.
You
know
who
are
working
on
the
project,
but
you
know
also
in
general,
where
again
ted
needs
to
provide
more
detail
on
what
his
thinking
is
here.
Right
he's,
usually,
you
know
quite
concerned
about
any
vendor.
G
G
H
We
table
this
for
ted
next
time
and
I'll
put
the
the
question
into
the
the
agenda
for
next
time
so
that
we
can
get
ted
to
address
it.
C
G
I
mean
ted,
and
I
have
I
mean
just
again
for
an
update
ted
and
I
have
had
multiple
conversations
on
you
know
what
we
would
need
to
do
in
order
to
add
conformance
testing
and
certification
for
the
project
itself
and
really
looking
at
you
know
some
of
the
groundwork
that
we've
been
able
to
do
with
the
aws
distro,
for
example,
with
performance
testing
or
conformance
testing
security
testing
that
we
can,
you
know,
contribute
and
happy
to
contribute
back
to
the
project.
It's
more
that
you
know.
G
E
Yeah,
I
think
one
item
like
I
have
information
about
challenge
footage
from
very
past
and
at
that
point
I
remember
conversation
was
that
open
telemetry
holds
some
common
pieces
for
instrumentation
and
actual
installation,
and
everything
happens
in
vendor
distress.
So
by
itself,
if
you
don't
want
to
enter
distra,
you
cannot
use
java
instrumentation
because
it's
just
common
pieces.
G
C
Yeah,
my
preference
would
be
the
the
latter
right
where
we
we
provide
like
for
for
a
lot
of
these
things
like
the
collector,
for
example,
we've
pre-packaged,
collector
binaries
from
the
pure,
open,
telemetry,
open
source
repo,
and
it
works
fantastically,
and
the
only
thing
it
might
be
missing
is
like
the
exporter
of
choice
that
some
people
want
to
use,
but
it's
it's
like
98
there
for
what
people
want.
At
least
personally,
that's
generally
my
my
preference.
G
For
most
things,
the
project,
yeah
and
and
again
I
think
that
the
again
there's
a
fair
bit
of
work
required
there.
You
know
we've
been
just
starting
the
process
of
starting
to
think
about
an
instrumentation,
spec
and-
and
you
know,
really
be
an
instrumentation
again
as
sergey
was
also
pointing
is
specific
to
each
language.
So
we
there
is
a
fair
bit
of
work
to
be
done
there.
There
are
no
shortcuts
here.
C
Yeah
and
there's
this
like
and
for
us-
and
I
think
this
is
pretty
consistent
across
all
the
companies
here
like
our
messaging
to
our
customers-
is
like
we
have
the
splunk
distribution
you
can
use
that
the
benefit
like
literally
the
benefit
of
using
it
is
includes
some
logging
stuff
out
of
the
box.
That's
eventually
going
to
be
replaced
by
native
ot
components
later
this
year
and
that
you
can
pick
up
the
phone
and
call
us
for
support,
and
someone
will
help
you.
C
C
Good
chunk
of
them
are
going
to
say,
like
hey,
we've
already
instrumented
everything
with
hotel
like
we're
going
to
send
you
today
to
make
it
work,
yeah
and
and
we're
supporting
both
yeah.
It's
just
one
of
them
has
a
much
easier
path
to
pick
up
the
phone
and
yell
at
someone
at
splunk
and
the
other
one
we're
like
well
you've
engineers
who
work
on
this.
You
know,
here's
excuse
what
you
should
do
can.
A
C
A
Post
this
into
the
doc-
I
I
mean,
I
know
ted
and
I
worked
together,
but
I
actually
have
no
idea
about
this
issue
until
today,
but
the
reading
it's
talking
about
java
agent
distributions,
which
is
not
the
same
thing
as
collector
distribution,
their.
B
H
A
I
still
I'm
frankly
confused
why
hotel
would
host
either
I
mean,
but
it's
not
the
same
thing
just
so
that
we're
clear
I
mean
yes,.
A
It
up
a
java
agent,
I
even
more
than
the
collector
does
seem
like
something
that
benefits
from
being
like
wrapped
in
a
vendor,
because
it's
very
difficult
to
do
like
a
one-line
post-compilation
integration
of
a
java
agent.
Unless
it
does.
A
So
I'm
sure
every
vendor
who's
trying
to
build
a
java
agent
type
of
thing
around
open,
telemetry
installers,
would
want
something
like
this,
but
but
I'm
I'm
not
clear
on
why
it
would
be
beneficial
really
for
anybody
for
those
to
be
hosted
in
open,
telemetry
repository,
and
you
know
no
one
has
that.
No
one
knows
what
the
argument
in
favor
of
that
is.
Do
we
not.
G
I
think
my
understanding
from
you
know
just
ted's
conversations
has
been
that
he
wants
to
consolidate.
You
know
all
activity
happening
in
the
you
know
with
different
vendors
under
the
open
telemetry
umbrella,
but
it
also
should
be
something
that
you
know.
We
look
at
it
as
an
a
more
scalable
strategy
so
that
you
know
there
is
that
push
and
pull
between
the
being
centralized
versus
being
taking
a
more
distributed
approach.
Where
you
know,
additional
innovation
is
added,
you
can
argue
for
either
approach,
but
you
should
have
a
cleaner
strategy
of
you
know.
G
H
C
Do
we
have
any
next
steps
there?
I
I
was
just
reviewing
it.
I
was
about
to
review
it
today.
I
think
I
I
had
an
action
item
to
paste
in
some
data
data
that
I
had
in
a
comment,
but
beyond
that,
what
what's
the
the
process
for
the
next
steps
for
this.
G
I
mean
morgan.
My
understanding
is
that
elena
and
cornelia
want
the
document
to
be
fully
addressed
for
the
questions
that
are
on
the
document
itself
and
then
they
are
in
the
process
of
doing
you,
know
other
interviews
and
reviews
and
building
us
for
more
information
information.
That
was
my
understanding.
Okay,.
G
E
Yeah,
I
think
I
pasted
next
link,
so
I
just
wanted.
We've
been
talking
about
signals,
interoperability
and
I
pasted
the
link
that
fluent
corn
it's
a
day,
zero
event
of
kubecon
accepted
this
talk
from
intern.
I
mean,
I
don't
think
he's
intern
he's
a
student
who
was
participating
in
our
community
bridge
program.
F
E
E
This
project
and
this
project
was
of
interest
for
fluent
community,
which
is
amazing.
G
Okay,
any
any
other
questions
from
the
community
folks
who
have
joined
in.
I
think
I
can
see
punya
weird.
B
No
questions
from
me
and
I'm
mainly
here
to
learn
how
this
forum
works
and
to
say,
say:
hi
to
everyone.
I
pulled
off
my
second.
F
B
F
Yeah,
so
I
I
raised
this
question
with
the
technical
committee
as
well,
but
this
is
sort
of
in
between
those.
So
the
the
question
is
about.
We
have
sort
of
like
country,
instrumentation
repositories
in
in
multiple
languages
whose
purpose
is
to
instrument
the
the
vast
majority
of
existing
open
source
frameworks
out
there
right
and
the
model
that
we
currently
adopted
is
very
different
from
what
open
tracing
used
to
do
in
open
tracing.
It
was
a
contrib
org
where
anyone
could
create
their
own
repo.
F
They
would
be
the
owners
of
that
repo,
and
so,
if
anyone
wants
to
adopt
a
specific
instrumentation
for
let's
say
express
framework,
then
they
would
make
a
standard
call
about.
Is
that
repository
maintained?
Is
it
like
what
what
well
license?
We
would
kind
of
enforce
the
apache
license,
but
everything
else
was
sort
of
as
a
normal
process
for
picking
an
open
source
project
with
open
telemetry.
We
went
in
upper
in
a
different
direction.
This
is
like
tightly
centralized.
Everything
is
in
one
one
repo.
F
There
is
a
small
number
of
maintainers
on
that
repo
and
my
concern
with
that
is
hey.
First
of
all,
these
maintainers,
I
I'm
hearing
complaints
that
they're
slowing
down
a
lot
of
the
development,
and
so
like
people
are
starting
to
put
essentially
just
instrumentations
in
their
own
works,
because
they
can't
sustain
the
speed
of
of
sort
of
the
the
the
friction
of
of
having
like
few
maintainers
they're
proving
all
this
stuff.
And
the
second
thing
is
like.
Theoretically,
why
do
you
even
need
that
sort
of
centralized
model?
F
Because
it's
I
mean
in
open
tracing
at
some
point?
We
even
I
I
started
recommending
people
saying
like
don't
even
bother
with
the
open
tracing
country,
because
we
have
a
registry.
That's
the
discovery
point
if
someone
needs
to
find
it,
whether
it's
in
countries
or
in
your
own
work
doesn't
really
make
any
difference,
because
a
contrib
doesn't
sort
of
convey
any
sort
of
guarantees
from
open
tracing
projects
itself,
and
I
think,
by
having
everything
right
now
in
the
monorepo.
F
G
Agreed
and
and
uri
just
to
you
know,
echo
your
concern.
I
have
raised
this
concern
several
times.
You
know
in
terms
of
maintainer
velocity,
not
keeping
up
with
the
kind
of
influx
of
you
know,
work
that
is
being
contributed
or
done
on
the
project
and-
and
you
know,
you'll
again,
we
really
need
to
take
a
look
at
you
know
a
decentralized
approach
on
the
project
in
terms
of
at
least
allowing
a
space
for
contributors
and
contributions
to
grow.
G
It
doesn't
have
to
be,
you
know,
tagged
as
broad
ready
or
one
dot
already,
but
you
know
still.
I
think
that
that's
something
that
you
know.
We
need
to
figure
out
what
the
good
right
balance
is
because
right
now
we're
we
are
really,
you
know
very,
very
thin
on
maintainer
maintainers,
if
you
will,
I
mean
we
have
two
maintainers
on
the
collector.
We
have
two
maintainers
for
javascript.
We
have
two
maintainers
for
go,
you
know,
and
that's
really
very
small
for
such
a
large
project,
where
we
have
more
than
a
thousand
contributors
yeah.
H
So
I
popped
a
thing
in
the
chat
like.
I
would
really
rather
deceit
rather
see
otel
optimized
for
discoverability
of
someone
else's
ways
to
integrate
with
hotel
than
for
us
to
try
to
centralize
it
like
centralizing
a
contrib
ends
up
with
you
know
with
having
to
put
up.
You
know
signs
of
therapy
dragons,
and
we
have
no
idea
where
this
is
because
somebody's
fallen
off
the
map
and
and
it
doesn't
offer
the
transitive
trust
that
we
want
it
to,
and
so
all
it
does
is
risk
you
know.
So
it
could
be
a
registry.
H
It
could
be
a
awesome
hotel
page
that
just
has
links
to
other
people's
connectors
or
or
implementations
for
frameworks.
You
know
whatever
it
ends
up
being,
but
I
think
yeah,
I
think
long
term.
We
should
not
have
a
big
contrib.
I
agree,
yuri.
A
So
I
mean
it
seems,
like
everyone
agrees
with
this:
do
do
we
want
to
just
change,
or
at
least
make
a
pr
to
change
whatever
documentation
we
have,
that
suggests
using
contrib,
or
maybe
we
can
like
do
we
wanted
to
try
to
define
like
where?
Should
we
be
using
contrib,
I
mean
like?
Is
there
any
scenario
where
we
think
can
trip
is
appropriate,
given
how
big
the
core
non-contrib
piece
is
becoming.
E
Just
from
like
maintenance
perspective,
it's
somehow
really
hard
to
distinguish
what
this
country
was.
Not
country
like
if
specific
instrumentation
is
interesting
for
many
people
who
will
be
setting
it
up
and
like
if
we
expect
maintainers
of
some
languages
to
set
it
up,
and
they
need
to
maintain
like
numerous
repositories
now,
which
makes
because,
like
main
maintenance
job,
even
harder,
just
to
release
all
those
repositories
and
like
if.
E
H
E
A
Like
the
idea,
I
mean,
I
know,
we're
gonna
run
out
of
time,
but
I
mean
I'm
just
saying:
look
if
you
want
to
use
contrib.
The
only
requirement
is
that
sort
of
like
the
one
liner
for
your
project
needs
to
be
about
open
telemetry
in
some
way
and
we'll
make
you
a
repository
and
that's
kind
of
it
and
like
there's,
no
guarantee
whatsoever
about
anything
in
terms
of
like
the
quality
of
the
software
or
fitness
or
purpose,
or
anything.
A
It's
just
a
place
for
people
who
are
doing
projects
for
open
telemetry
that
don't
want
to
be
attached.
Their
personal
github
handle
or
whatever,
which
I
can
understand
being
a
thing.
A
G
Mean
I
think
that
the
registry
approach
is
a
good.
You
know,
approach
to
begin
with,
but
one
of
the
issues
that
you
know
I've
seen
is
that
it's
just
not
integrated
with
the
development
workflows.
Today
it's
standalone
on
the
website
and
that
really
causes
a
great
disconnect
and
and
stability.
If.
A
We
I'm
sorry,
I
have
a
hard
stop
at
10
30.
if
we
use
github
tags,
which
is
a
thing
to
say.
I
want
to
tag
this
repository
as
being
open,
telemetry
registry,
something
something,
and
then
we
have
some
script
that
ran
once
a
week
or
something
to
pull
that
would
that
work,
because
I
I
hear
you
alita
on
the
on
the
process
front.
E
We'll
have
all
the
problems
with
cla
and
like
how
do.
H
E
It
up
and
how
like,
can
people
use
it
after
she
like
how
can.
E
H
G
I
mean,
I
think
sarah
would.
The
next
step
would
be
two
things.
One
is
the
registry
and
how
we
automate
that,
having
a
clear
you
know,
otep
or
an
enhancement
proposal
which
then
I
can
you
know,
maybe
add
attached
engineering
to
to
build
it
out
in
the
summer,
but
also
the
second
part
where
we
actually
have
a
formal
discussion
with
our
maintainers
and
tc.
G
Out
you
know
what
scheme
works
because
I
can,
I
can
tell
you,
I
have
filed
at
least
six
issues,
if
not
more
on
this
concern.
H
And
it
seems
that
we're
allowed
to
assign
things
to
ben
now
that
he's
left,
so
we
just
assigned
this
whole
problem
to
him.
Good.