►
From YouTube: 2021-05-05 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
B
C
D
B
C
B
D
A
A
All
right,
so,
if
no
one
has
a
specific
topic
to
add
there,
I
would
like
to
go
straight
to
the.
A
A
F
So
this
this
particular
item
I
didn't
when
I
created
it
ages
ago.
I
didn't
mean
it
as
it's,
not
something
it
wasn't
meant
as
a
work
item
on
the
board.
It
was
like
it's
not
something
that
can
be
done.
I
just
wanted
one
place
where
we
can
have
a
list
of
things
that
is
like
a
little
pins.
That's
why
I
put
the
pin
think
I
was
at
some
point.
F
I
thought
it
would
be
reasonable
to
maintain
a
list
of
libraries
that
are
instrumented
using
diagnostic
source
and
or
activities
so
that
we
know
that
those
are.
You
know
to
be
treated
differently
by
auto
instrumentation.
A
Yeah,
this
is
a
very
good
thing
for
us
to
have
an
idea
of
how
fast,
let's
say,
especially
open
source
projects,
are
moving
to
adopt
activity
source.
You
know,
so
I
I
think
actually
not,
that
we
are
going
to
really
list
these
things,
but
I
think
let
me
take
this.
I'm
just
going
to
create
a
readme
on
the
box
and
if
anyone
becomes
aware
we
try
to
put
stuff
there,
you
know
I
don't
have
any
list.
A
I
can
look
at
some
of
the
projects
around,
but
I
don't
have
any
lists
about
this,
but
then
we
have
a
place
and
then
people
can
just
originally.
B
That
this
ticket
is
about
something
else.
Do
we
have
even
at
least
what
libraries
are
we
instrumenting.
D
We
we
have,
we.
E
A
Yeah
so,
but
I
I
would
like
to
do
two
separate
things
in
this
case.
You
know
our
instrumentations,
the
things
that
are
instrumented
by
the
ripple
and
the
things
that
are
mainly
instruments
by
their
authors,
but
use
technology
like
diagnose,
source,
active
source.
F
You
know
I
was
just
thinking
about
it
and
actually
maybe
it's
better
as
an
issue
because
then
imagine
we
know
something.
People
can
go
just
comment
on
it,
which
means
oh
yeah.
You
guys
forgot
this
and
this
and
then
we
can
go
and
modify
the
main
issue.
Maybe
we
should
like.
Maybe
the
issue
should
be
a
pinned
issue.
That
is
like
here's
a
table
of
things.
F
We
know,
please
community,
add
a
comment
if
you
know
something
that
should
be
changed
there
and
then
one
of
the
maintainers
can
go
and
update
the
issue.
I
think
it's
in
that
sense.
It
would
be
easier
than
having
a
readme
and
opening
a
pr
on
this
every
time,
because
it's
just
it's
not
even
documentation
format.
It's
more
of
a
kind
of
forum
thing.
Let's
collect
this
information.
A
F
Might
be
easier
to
work
might
be
easier
to
work
with
and
it's
less
hidden
and
it's
not
something
that
really
requires
code
reviews.
But
I
don't
mind
too
much
just
as
an
idea.
A
I
I
think,
actually
I
I
kind
of
think
that,
as
things
mature
people
will
want
to
see
the
readme
you
know,
but
but
I
I
think
we
can
start
with
this
kind
of
have
a
at
least.
We
keep
the
the
issue
open
with
the
list
and
then,
when
we
transform
we
have
perhaps
I
don't
know
five
ten
entries
there
will
transform
your
rhythm
and
start
to
have
listen.
B
It
is
rasmus
because
before
it
was
not
pro
right
now,
it's
it
it
can
lead
to
arbitrary
code
execution
before
plugins
were
introduced.
It
was
not.
It
was
just
about
changing
the
configuration.
G
Way
that
you
want
to
access
that,
you
just
want
one
path:
a
full
file
path
to
a
plug-in
to
jason
is
that.
B
C
F
B
A
A
Okay,
so
I
think
I
think
you're
in
agreement
what
what
should
be
doing
about
that.
I'm
I'm
not
sure
if
this
is
something
that
you're
gonna
pick
up
to
work
now,
but
I
think
we
we
understand
that.
A
A
So
I
knew
that
last
time
the
part
that
you
guys
decide
about
the
committed
items.
Do
you
guys
wanna,
because
I
see,
for
instance,
that
oh.
B
G
A
D
A
Yeah,
I
think
we
need
to
to
clarify
this,
so
people
looking
from
outside
understand
why
that's
the
case
right
now
and
also
what
are
plans
for
the
future
in
this
regard.
You
know
this
is
a
kind
of
I
think
for
people.
Looking
from
the
perspective
of
open,
telemetry
is
hard
to
understand.
A
Yeah,
so
I
think
we
can
redirect
to
to
to
this
dark
and
close
the
issue.
So
now
we
have
a
explanation
of
why.
F
I
think
I
would
I
would
still
be
so
we
have
the
tribal
knowledge
I
would.
I
would
call
out
the
danger
of
reusing
things
as
a
library
as
libraries.
B
F
F
Yeah,
no
totally
it
can
be.
What
I
mean
is
essentially
when
we,
when
we
explain
why
it's
not
simply
a
library,
we
kind
of
help
the
need
where
they
need
to
explain
the
cherry
new
person
every
time.
A
Yeah,
I
I
think
we
wanna-
I
I
wanna
kind
of
consider
this
for
the
future,
also
because
we
have
the
the
trace
and
we
understand
the
current
situation,
but
going
forward
4.net,
5
and
6
in
principle.
We
could
have
the
activity
already
there.
You
know
it's
a
big
change,
but
I
think
it's
a
conversation
that
we
should
have
down
the
line.
You
know,
I
don't
think
it's
the
the
priority
for
now,
but
I
think
it's
a
conversation
that
we
should
plan
ourselves
for
down
the
line.
A
A
And
so
perhaps
I
I
that's
the
part
I
I
I
said
this
part
of
the
conversation
last
week,
so
you
guys
already
discussed,
can
you
so
it
perhaps
is
better
somebody
that
was
there.
He
was
in
the
meeting
last
week
right,
yeah.
B
You
are
correct,
so
basically,
I
think
you
can
go
on
our
own.
Basically,
we
try
to
sort
somehow
this
backlog
here.
So
you
have
several
bars
and
I
think
we
have
like
something
like
a
beta
bar
which
are
like
they
must
haves
before
we,
I
don't
know
tell
the
community
that
it's
ready
to
play
around
that
we
have
some
kind
of
release,
not
stable,
but
something
usable
below
the
better
release
bar
we
have.
We
are
having
like
additional
things
that
needs
to
be
done
before
ga.
B
A
E
Yeah,
the
other
thing
to
call
out
is
that
we
only
prioritize
the
things
that
are
above
the
the
beta
bar,
so
anything
that's
between
ga
and
beta.
It's
just
a
random
order.
A
I
see
so
so
I
I
don't
want
to
take
anybody's
time,
while
I
kind
of
learn
about
the
the
status
of
those,
so
I
can
catch
up
later,
looking
at
the
youtube
video,
so
I
I'd
like
to
move
to
things
that
you
guys
are
want
to
just
cause.
G
B
B
B
Just
started
so
I
will
try
to
basically
show
on
your
pull
request,
because
I
do
not
have
it
checked
out.
So
basically,
we
have
this
idea
of
having
the
possibility
extension
accessibility
prototype
that
we
can
define
in
a
plugin
json
file,
concrete
files
that
will
try
to
discover
assemblies
which
are
implementing
some
interfaces,
and
then
they
will
be
plugged
into
some
places.
B
And
and
right
now
we
see
one
big
problem
that
do
you
remember
the
name
of
the
interface
rasmus.
C
F
But
so
the
the
problem
with
that
is
typically
when
people
add
extensibility
points
for
things
that
might
come
up,
but
no
concrete
examples
have
been
seen.
Usually
those
are
yeah.
B
F
During
but
that's
not
exporters,
this
propagators
right.
B
And
so,
and
also
even
in
the
specification
of
hotel,
there
is
described
that
it
should
be
possible
to
add
custom
propagators,
so
we
decided
to
start
with
something
auto
specific,
so
this
will
be
one
extension
point
and
it
will
be
at
least
something
valuable
for
everyone
and
second,
one
that
we
will
need,
but
it
is
not
covered
by
this.
Pr
is
to
design
some
kind
of
http
middleware
that
will
can
modify
the
responses.
B
Use
cases
yep
correct.
B
C
Yeah,
if
you're,
looking
at
the
interface
now,
the
main
interface
is
just
a
marker.
Now
yeah.
C
B
D
G
To
the
tracer,
oh
sorry,
the
span
contest
shouldn't
have
a
reference
to
the
tracer.
It
should
be
standalone.
B
Yes,
so
basically,
the
problem
right
now
is
that
we
to
in
order
to
have
any
version
of
this
plug-in
mechanism,
we
will
need
to
leak
some
kind
of
types
which
are
like
not
auto
sdk.
You
know
implementation
and
probably
before
beta.
It
is
not
a
problem,
but
we
should
probably
have
some
plan
how
to
address
it
before
the
ga.
E
G
I'm
sorry,
I'm
not
sure
I
understand
the
issue
yet.
Can
you
well
the.
B
Problem
is
that
maybe
in
a
different
way,
in
order,
some
someone
needs
to
implement
these
interfaces,
so
we
will
need
somehow
to
publish
or
someone
will
need
to
reference
these
dlls
to
implement
the
interface
correct.
B
B
F
Here's
a
a
provocative
question
that
I
realized
that
I
keep
keep
kind
of
dwelling
on
this,
but
I'm
just
trying
to
avoid
technical
complexities
that
can
be
dealt
with
later.
So
is
it
really
a
scenario
that
people
just
implemented,
they
drop
it
and
we
ought
to
discover
it.
F
I
mean
we
could
just
as
well
say
no
guys
if
you
we
offer
the
sustainability
point,
you
have
to
build
the
entire
trace
for
yourself
so
and
then
ship
it
as
a
separate
like
basically
do
we
really
need
to
allow
people
to
say
we
take
the
default
open,
telemetry,
tracer,
install
it
and
then
take
this
library,
because
if
I
understand
correctly,
please
correct
me:
if
I'm
wrong,
that's
what
they're
trying
to
do.
Then
I
implement
an
assembly
that
implements
these
interfaces.
I
place
it
next
to
each
other.
F
F
You
would
really
like
to
have
a
tracer
that
does
some
custom
propagation
sure
take
the
tracer
implement
this
interface.
That
is
right
there
in
the
source
code,
rebuild
the
entire
tracer.
Now
you
have
your
own
version
of
the
tracer
extended
by
this
functionality,
because
I
have
a
feeling
that
at
least
initially,
like
five
people
in
the
world
will
be
doing
this
at
best
and
in
some
future,
when
there
is
a
need
for
105
people.
To
do
that,
then
we
can
have
a
real
extensibility
mechanism,
like
with
the
one
that
you're
planning
for.
A
So
the
fear
when
we
have
this
kind
of
solution
is
kind
of
requires
basically
discipline
to
kind
of
not
let
the
project
diverge
and
then
having
a
bunch
of
projects
that
start
from
open
telemetry,
but
they
are
not
any
more
collaborating
or
doing
anything
in
common.
You
know,
so
I
think
the
the
main
the
main
concern
is
about
really
the
the
forks
diverging
you
know,
so
the
propagators
that
we're.
D
B
Correct
but
the
the
specification
says
that
it
they
you
can
implement
custom,
auto
custom
propagators.
It
is
an
extensibility
point
by
design.
F
Right
right,
is
it
right?
It's
an.
F
I
I
think,
I'm
not
saying
that,
like
at
the
interface,
I
think
you
were
on
the
right
track
there.
So
no
no
concern
there.
I'm
saying
that
we
could
choose
to
not
tackle
the
problem
about
nugets
and
versions
and
all
of
this
stuff.
For
now.
I
think
that
we
are
trying
to
make
it
perfect,
which
is
the
enemy
of
good.
Yes,
in
an
ideal
case,
some
of
this
will
not
live
in
the
in
in
this
repo.
F
But
what
I'm
trying
to
like
make
us
consider
is
create
a
separate
special
directory
in
this
repo
for
stuff
that,
in
an
ideal
case,
doesn't
belong
in
this
vehicle.
F
Put
the
stuff
there
and
say:
okay
for
now,
for
now
we
have,
we
have,
like-
I
don't
know
five
propagators
so
make
it
seven
whatever
be
there
implement.
This
interface
have
some
sort
of
repo
like
we
don't
create
the
problem
with
with
the
assembly's
versioning
yet
and
then,
if
some
person
right
now,
there
is
very
few
participants
here
and
some
other
people
join
and
say.
Well,
you
know
what
we
would
like
one
more
propagator,
which
is
technically
not
open,
telemetry
compliant,
but
added
there
edit
in
that
directory.
F
At
some
point,
when
we
have
enough
examples,
then
we
say
you
know
now
we
have
to
refactor
it
out.
Then
now
we
have
to
go
solve
this,
this
problem
with
with
versions
and
and
assemblies
and
then.
E
So
here's
an
idea
that
might
balance
things
out
a
bit
so
greg's
point
about
not
needing
to
solve
it
right
now,
perhaps
in
the
short
term,
there's
a
fork,
but
long
term.
We
talked
about
how
we
want
to
use
the
hotel.net
sdk,
which
already
has
the
ability
to
plug
in
these.
These
extension
points
correct,
and
so
maybe
that
will
be
an
appropriate
point
to
be
able
to
migrate
away
from
a
fork
and
be
able
to
auto
load.
Some
of
these
things
just
a
thought.
F
Yeah,
at
the
end
of
the
day,
I
definitely
don't
mind
too
much.
All
I'm
saying
is,
I
think
I
agree
with
chris,
like
all
I'm
saying
is,
it
seems
to
me
like
a.
F
If,
if
just
as
a
thought
experiment,
imagine
we,
because
in
the
long
term
I
think
it's
not
only
bureaucracy,
it's
the
right
thing
to
do
to
have
such
extension
points
right
so
when,
if
if
there
was
no
bureaucracy,
when
would
we
do
this?
Would
you
do
it
now
great?
Let's
do
it
now?
Would
we
do
it
later,
then?
Is
it
an
option
to
reflect.
B
Basically,
maybe
I
should
rephrase
my
question
because
do
you
think
that
maintaining
a
fork
is
before
even
better
release
or
for
better?
But
why
is
there
fork?
I
didn't
quite
understand
it
because
you
suggest
to
make
a
fork
and
just
place
some
library,
some
dedicated
folder
when
we
plug
some
additional
things
right.
G
F
A
I
think
that
it
you
need
to
be
when
we
claim
any
release
you
know
kind
of.
If
you
want
to
claim
a
open,
telemetry
release,
I'm
not
saying
ga,
but
I
I'm
saying
anything
perhaps
like
beta
or
something,
then
we
need
to
to
clean
up
this
stuff.
That's
my
understanding
can
be
wrong,
but
that's
my
understanding.
You
know.
F
Really
so
just
just-
and
I
no
position
just
to
make
sure
I
really
understand
this
so
says
there
is
open.
Telemetry
requires
a
bunch
of
propagations,
so
say
we
have
them,
but
then,
in
addition,
we
have
a
possibility
to
say
okay,
so
we
know
that
some
customers,
some
industry
partners,
also
use
other
ones
that
are
just
propagating
in
a
different
way
and
you're
not
allowed
to
have
them
there.
D
G
A
Put
something
somebody
has
a
different
race
id
format
and
we
know
some
some
cases
that
that
that's
true
and
I'm
not
talking
about
kind
of
64
120.
A
So
I
think
the
idea
is
because
this
is
vendor
tired.
They
want
to
avoid
open
telemetry
to
have
anything
that
sounds
like
it's
favoring
one
vendor,
because
then
you'll
be
okay.
The
vendors
participate
on
the
project,
but
then
you'll
be
like
oh,
this
distribution
favors
vendor
x,
because
it
includes
an
open
telemetry.
There
are
other
participants
of
the
open,
telemetry
project
that
are
not
involved
on
this
specific
project
and
their
specific
propagation
is
not
there.
You
know,
I
think,
that's
what
they
are
trying
to
avoid.
B
A
Yeah,
so
I
I
think
it's
in
the
end.
It's
just
trying
to
avoid
kind
of
having
some
case
of
let's
say,
for
instance,
the
collector.
A
There
are
a
lot
of
contributors
there,
aws
splunk,
a
lot
of
people
create
their
exporters,
on
the
other
hand,
but
the
idea
is,
I
think,
to
the
core
to
not
sound
like
because
aws
and
explain
contributed
a
lot
to
this
to
the
poor.
The
core
should
not
have
anything,
that's
related
to
any
of
those
you
know
it
may
it
may
provide,
let's
say
facilities
that
these
or
some
other
companies
kind
of
asked
to
be
put
on
the
project,
but
that
goes
through
the
the
same
process
and
it's
extension,
that's
sloping.
F
Another
idea,
just
like
brainstorming
here
right
so
say
if
there
is
a
okay,
so
if
we
can't
have
it
in
this
repo
in
the
main
repo
can
we
have
a
another
report
that
has
the
extensions,
but
rather
than
being
a
fork,
that
adds
something
it
has
only
essentially
it's
like
only
the
extra
stuff
and
then
to
build
it.
There
is
some
kind
of
script
that
essentially
clones
the
main
repo
and
adds
like
into
a
sub
directory,
or
something
like
that.
E
So
idea
behind
the
contrib
repos
that
are
out
there,
so
the
sdk
has
a
contrib
repo.
The
collector
has
a
contrib
repo
and
that's
where
the
vendor
specific
things
can
go,
but
then
that's
a
decision
for
the
sig
to
make,
because
there's
a
maintenance
burden
with
managing
these
contrib
repos.
In
addition
to
the
main
repo.
F
Well,
if
we
were
about
to
implement
those
non-auto
compliant
propagators
ourselves,
then
we
could
set
up
that
repo
implemented
ourselves.
Since
we
were
about
to
maintain
it
anyway,
we
would
be
just
maintaining
the
same
code
in
a
different
repo
and
we
would
avoid
these
extension
point
complexities.
F
Like
all
I'm
saying
is
that
the
versioning
problem
is
a
really
bad
one
and
we
spend
we.
We
have
a
lot
of
experience
having
a
lot
of
pain
with
that,
so
I'm
just
having
having
had
so
much
pain
with
this
just
brainstorming
possibilities.
B
I
was
muted,
sorry,
I'm
sure
I'm
sharing
my
screen,
yes
yeah.
So
I
was
asking
explicitly
because
I
was
reading
this
specification
about
this
country
packages,
because
it
was
mentioning
only
that
they
have
to
be
packages
like
for
me.
A
package-
and
I
asked
it
of
course
to
be
correct-
is,
for
example,
a
dll
so,
and
I
was
asking
if
he
didn't
need
to
be
a
separate
repository,
and
the
answer
is
it's
up
to
the
sick.
If
they
want
to
have
this
contrib
stuff,
it
just
need
to
be
somehow
separated.
A
This
one
at
first
at
first
it
sounds
me
to
me
as
compliant.
We
should
do
do
diligence
and
be
sure
about
that.
Exactly
yeah.
Of
course.
Definitely.
B
B
E
Assuming
we
do,
which
would
prevent
the
splunk
from
needing
to
maintain
a
fork
in
the
short
term,
then
at
some
point,
if
we
do
switch
to
supporting
the
sdk,
the
sdk
already
has
a
contrib
repo,
and
now
I
I
think
that's
going
to
be
a
point
of
confusion
as
far
as
what
can
support
what,
and
so
that's
just
something
I
I
want
to
be
careful
of.
A
I
I
I
think
I
understand
your
concern
is
in
the
sense
that
it's
going
to
be
very
confused
to
people
about
which
ripple
does
what
and
it's
even
confusing.
B
E
A
F
So
would
it
be
potentially
an
option
to
follow
up
with
the
red
hotel
folks
to
check
whether
it's
okay
for
us
to
have
cointrip
in
a
separate,
clearly
distinct
subdirectory
inside
of
our
repo
and
all
the
contrib
features
from
there
being
off
by
default
and
an
explicit
opt-in?
If
that
is
compliant,
then
we
could.
F
We
would
at
least
for
now
avoid
the
whole
versioning
problem
and
we
could
take
that
route
and
if
it's
not
compliant,
then
hopefully
we
know
by
next
week
and
we
can
either
decide
to
proceed
with
the
plan
that
you
guys
came
with
or
like
brainstorm
a
little
more.
And
so
how
do
you
guys
feel
about
that.
A
I
think
if
it's
best
monster
by
hotel.
A
It
sounds
very
tempting
you
know,
but
I
I
just
want
to
be
careful
and
not
be
too
excited
about
that.
If,
if
kind
of
hotel
governance
board
doesn't
like
that,
you
know,
but
if
it,
if
you
pass
that
monster,
I
think
it's
very
tempting,
because
then
the
few
of
us
that
are
participating
on
the
project
can
have
their
builds.
We
deal
with
one
project
makes
our
life
much
easier.
You
know
yeah,
so
I
I'm
firm
very
tempted
for
that.
F
Nugget
right
I
mean
interfaces
cannot
be
versioned
anyway,
once
they're
shipped
right.
So
maybe
we
strictly
only
include
interfaces.
F
It
found
that
nougat
at
least
for
now,
and
that
means
that
some
code,
some
tooling
code,
may
need
to
be
duplicated
and
maybe
that's
okay,
some
sort
of
strict
policy
sure
would
like
because
then
we
know
at
least
we
have
only.
F
A
I
I
think
you
are
done.
I
think
the
action
item
for
me
is
to
follow
up
with
opinem's
governance
board
to
be
sure
that
for
us
to
have
this
specific
vendor
inside
at
least
for
the
time
being,
it's
okay
and
each
vendor
that
may
want
to
do
a
build,
put
their
changes
there.
If
that's
acceptable,
it's
it's
good
for
us.
You
know,
I
think
it
simplifies
our
and
and
that's
the
follow-up.
So
that's
that's
what
I
want
to
do
kind
of
as
soon
as
possible.
E
That,
okay,
so
the
the
topic
I
have
is
about
otlp.
Are
you
all
familiar
with
what
that
is
for
open
telemetry?
E
It's
basically
the
one
of
the
standard
export
protocol
protocols
from
open
telemetry,
its
primary
communication
method
is
grpc
and
so
for
net.
There's
two
libraries
out
there
that
you
would
use
for
this.
One
works
one
supports.net
framework,
the
other
only
supports
net
standard,
2.1
and
so
the
net
framework
one.
E
B
E
Yeah,
so
I'm
mostly
asking
the
question
here,
especially
for
the
the
data
dog
engineers
to
see
if
you've
had
to
do
anything
with
grpc
on.net
framework.
Yet.
E
D
A
So
just
for
me
to
be
sure
I
understand
it
brings
a
native
component
on
that.net
framework
to
work.
I
mean
on
the
dot
net
when
it's
one
with
against
dotnet
framework.
It
needs
a
native
dll
tool
to
be
able.
E
H
E
Usually,
and
so
I've
had
some
experience
using
it
at
new
relic
for
for
some
of
our
things
and
so
far
we
haven't
run
into
any
problems
with
our.net
framework
customers,
but
where
we've
run
into
problems
with
it
is
with
some
of
our
linux
customers,
specifically
there's
certain
versions
of
debian,
where
debian
was
built
with
a
different
version
of
a
native
library
that
the
grpc
library
also
depends
on,
and
it
causes
problems.
H
E
Yeah
there's
a
few
and
then
there's
other
gotchas
where
so
with
net
core
3.1.
The
windows
performance
of
the
the
fully
managed
version
is
super
efficient,
but
when
you
run
it
on
linux,
it
wasn't
as
efficient.
I
believe
that
changed
with
dot
net
five.
A
E
A
E
A
A
All
right
sounds
good
to
me.
Anyone
else
wants
to
raise
their
hand
and
bring
issues.
G
And
then
the
ci
work
that's
just
taken
a
long
time,
so
I
don't
really
have
any
updates
on
any
of
those
right
now.