►
From YouTube: 2022-10-17 meeting
Description
cncf-opentelemetry meeting-2's Personal Meeting Room
A
B
Ers,
as
you
all
are
filling
in
your
report,
packs,
if
you
could
just
mention
in
that
section,
the
state
of
the
metric
stocks
or
ideally
a
link
to
the
current
metric
stocks,
rear
implementation.
That'd
be
super
helpful.
We're
trying
to
compile
all
of
that
for
kubecon
next
week.
D
E
C
Better
excellent,
okay,
just
give
me
one.
Second,
we
can
get
rolling.
C
C
Great
to
be
back
great
to
be
back
wonderful
to
have
you
back,
okay,
let's
get
started
we'll
work
through
the
agenda.
As
for
extra
announcements,
there's
a
fair
amount.
Let
me
just
check
the
date
might
be
on
the
wrong
one.
A
lot
of
topics
today,
as
for
extra
announcements,
I
anticipate,
there's
quite
a
few
just
with
the
election
coming
up
this
week,
as
well
as
Hotel
unplugged
coming
up
next
week,
along
with
kubecon,
so
just
I'll
quickly
read
through
these
and
we'll
get
to
the
the
meteor
topics.
C
A
Mostly
focused
on
metrics
yeah
in
general,
all
of
them
but
metrics
first
yeah
there
were
some
issue
regarding
cardinality
anyway,
yeah.
B
Yeah
one
major
major
thing
people
are
interested
in
doing
is
identity.
So
far,
we've
just
been
like
Samantha
conventions.
Here
you
go,
but
when
it
comes
to
which
of
these
conventions,
you
should
use
to
identify
particular
Services,
for
example,
or
use
yeah
in
your
default
metrics,
dashboards
and
stuff,
like
that
cool,
we
haven't
actually
identified
those,
and
that
makes
it
harder
to
to
change
or
adjust
other
semantic
conventions,
because
you
don't
know
if
how
they're
being
relied
upon.
C
C
C
The
annual
I
take
it.
That
means
it's
like
very,
very
close,
Greg
yeah.
E
Yeah,
it's
done,
there's
a
there's,
a
PR
for
the
actual
release,
but
we're
waiting
to
make
sure
that
we
have
things
like
documentation
buttoned
up
before
we
actually
release
it.
Well,.
E
We're
hoping
to
get
it
before
kubecon,
but
if,
like
normal
kubecon
Target
release
dates
the
closer
they
get,
the
the
more
intimidating
that
is.
C
If
that's
with
any
fixed
date,
moving
on
to
go,
we
have.
A
One
point
go
ahead:
yeah,
sorry,
Daniel
is
that
sorry
I
feel
like
this
is
but
I
think
that
somebody
may
ask
whether
it's
API
only
for
now.
E
C
It
well
congrats
moving
on
to
go.
We've
released,
1.11.0
0.32.3
with
metric
Trace,
SDK
bug,
fixes
and
features
0.33.0
of
the
SD
of
the
traces
metric
SDK
is
also
planned.
For
next
week
we
have
a
the
work
towards
the
beta
metrics
xdk
is
22
complete
for
C
plus
plus.
We
have
a
lot
of
metrics
work
ongoing
with
a
ga
Milestone.
That's
currently
targeted
for
October
29th
wow.
That
is
very
fast
progress.
C
Metric
socks.
We
also
7,
has
released
an
initial
getting
started
guide
at
this
link
that
everyone
can
go
review.
The
API
and
SDK
reference
dock
is
already
in
read
the
docs.in
The
Collector
0.62.0
went
out
with
a
bug.
0.62.1
went
up
the
following
day
to
address
that
bug:
no
updates
for
erlang
no
docsi
for
metrics,
okay
for
comms
metric
stocks
sketches
for
the
docs
layouts
are
here.
C
So
if
you
have
any
interest
in
how
the
docs
will
be
laid
out
on
the
website,
you
can
do
that
here
and
I'm
sure,
Austin
and
others
would
appreciate
feedback.
The
end
user
working
group,
the
first
survey
results
have
been
published.
C
I
think
they
chose
to
link
them
in
slack,
so
they
just
can't
be
picked
up
immediately,
but
I'm,
not
the
person
who
made
that
call
I
guess
you'll
have
to
go.
Look
through
slack
to
find
the
link
I'll
get
more
details
next
week,
but
why
those
were
only
in
slack
I'm,
not
actually
certain
for
for
everyone
throughout
the
diet.
I
also
have
just
purged
through
these
last
week,
when
they
shared
over
slack.
It's
interesting,
there's
there's
some
good
data
there.
C
It's
mostly
from
actual
Hotel
end
users,
not
just
pure,
like
vendor
contributors,
and
it
yeah
a
lot
of
like
good,
useful
information
about
how
people
are
using
open
Telemetry
how
they
like
it.
What
works,
what
doesn't?
Unsurprisingly
documentation,
is
probably
the
least
like
thing,
but
it
was
still
in
the
positives.
C
So
just
like
a
lot
of
great
data,
there,
that's
useful
for
guiding
the
project
forward
and
also
for
contributors,
I
think
a
lot
of
great
data
that
shows
the
impact
of
all
the
work
that
we've
put
in,
which
is
also
just
very
endearing
updates
on
Oto
unplugged
from
Dan,
which
maintainers
will
be
in
attendance
Dan.
Do
you
want
to
lead
this
discussion.
E
Yeah
I
mean
obviously,
which
maintainers
will
be
in
attendance.
Is
the
important
question
for
now
we're
trying
to
get
an
idea
of
how
many
people
may
want
to
participate
in
the
maintainers
panel
and
some
things
like
that,
similar
to
what
we
did
in
Austin.
If
people
remember
that
maintainers
here
do
are
eligible
for
a
comp
ticket,
so
if
you're
a
maintainer
contact
myself
or
Austin
to
get
the
discount
code
for
that
anybody,
that's
already
purchased
a
ticket
without
going
through
that
process.
E
I
apologize
if
you're
worried
about
it
contact
me
and
we
can
maybe
figure
something
out.
I.
C
Think
I
just
for
reference,
Dan
I,
think
a
bunch
of
the
Splunk
maintainers
will
be
there,
but
I
I
forgot
to
tell
them
about
the
registration
link.
Don't
worry
about
the
money!
It's
fine,
but
I'll
get
the
list
of
of
the
Splunk
maintainers
who
are
attending
I
can
add
them
to
this.
There's
like
three
or
four
okay.
E
C
E
C
Yeah
I-
wouldn't
yes,
I,
wouldn't
necessarily
travel
for
it,
but
there's
always
a
chance.
Someone
lives
nearby,
like
I,
mean
yourself,
Daniel,
you're,
of
course,
going
to
kubecon,
but
you
do
live
in
Detroit.
So
if
you
weren't,
you
could
still
go
to
this.
E
Yeah
Philip
put
in
the
chat
I
would
say
definitely
advertise
this
event
with
any
end
users
that
yeah
that
will
be
gone,
because
even
if
you're
not
attending,
if
you
know
end
users
who
are
attending
it
would
be
great
to
have
them
select
event
end
users,
definitely
I
would
say,
are
our
most
important
demographic
for
attendance.
If
we
can
get
them.
C
At
a
vendor,
where
you
have
customers
who
are
going
to
kubecon
like
please,
please
send
us
their
way
or
if
you
are
an
end
user
yourself
or
you
know,
end
users
and
no
end
users
are
going
like.
Please,
please
advertise
this
all
right.
Next,
we
have
Tristan
best
practices,
Docs.
D
Came
up
in
the
in
as
a
question
in
the
erlang
Sig
again
I
get
similar
questions
like
this
all
the
time
and
realize
we
have
this
meeting
today,
so
I
figured
I'd,
throw
it
on
the
agenda
and
that
I'm
not
sure
which
Sig,
maybe
Comes
This
would
I
should
bring
this
up
with,
but
I
think
it
would
be
something
very
useful
to
have
a
common
patterns
and
best
practices
type
doc
for
at
for
at
least
tracing
and
the
others
eventually
I
know
vendors
have
I
can
find
Pages
if
I
Google
best
practices
tracing
you
know,
I
can
find
Pages
for
vendors,
but
nothing
Consolidated
and
under
just
open,
telemetry,
I
guess.
C
Probably
the
commsig
like
they
don't
they
don't
like
strictly
just
handle
documentation,
but
right
structural
stuff
like
this,
would
almost
certainly
go
to
that.
B
B
D
And
I
guess:
yeah
I
can
to
bring
it
to
comms
and
try
to
somewhat
drive
it,
even
though,
even
if
I'm
not
writing
it
but
try
to
make
it
happen
and
then
I
guess
yeah.
Another
reason
to
bring
it
up
here
is
there's
so
many
different
people
from
different
vendors.
If
vendors
who
have
documentation
like
this
would
be
great
to
get
that
information
to
use
as
a
you
know,
starting
point
to
consolidate.
B
Yeah,
actually,
if
you
want
to
anyone
who
knows
about
that
information,
if
they
and
it's
they
feel
like,
it
would
be
fine
for
somebody,
anybody
to
Port
it
over
from
their
vendor's
website
or
wherever
it
currently
lives,
or
even
potentially
just
linking
out
to
that
file.
An
issue
on
the
hotel,
IO
GitHub
repo,
like
we
have
a
bit
of
Staffing
on
that
comsig.
E
C
Next
up
is
Philip
and
Severin.
Metric
stocks
are
missing.
F
Yeah
actually
I
don't
know
if
Severn
is
on
the
call
or
not.
Yes,
you
are
yeah
I'm
here,
yeah
yeah.
A
H
Can
I
I
can
quickly
say
something
I
mean
we
started
with
this
already
right,
so
so
most
of
you
have
now
filled
in
where
the
Elvis
metrics
and
if
there
are
docs
right.
So
this
is
actually
just
what
we
also
wanted
to
ask
like
hey.
We
want
to
improve
that
until
kubecon
starts
next
week,
so
let
us
know
where
you're
standing,
let
us
know
if
you
already
have
some
talks
or
I,
saw
that
some
languages
have
working
examples
already.
H
So
maybe,
if
you
can
point
us
there,
then
Philip
and
I
and
the
rest
of
the
team,
we
know
some
languages
a
little
bit.
A
lot
of
dangerous
half-knowledge
so
cannot
help
with
everything.
But
just
let
us
know,
and
then
we
can
see
what
I
I
say
you
can
Cobble
it
together.
I
mean
that's
the
preferred
way.
But
but
if
you
say
okay
we're
out
of
resources,
then
we
can
at
least
try
to
to
figure
something
out.
H
That's
at
least
good
enough,
then
for
kubecon
to
to
have
something,
and
then
we
can
work
from
there.
F
Yeah
I'll
I'll,
add
to
that
that
I
mean
I.
Think
I
mentioned
this
a
while
ago
and
then
sketched
out
some
stuff
in
the
the
issue.
That's
linked
up
there,
oh,
which
some
folks
have
engaged
on
I,
know
that
Tyler
I
believe
you
added
it
to
like
the
the
milestone
for
metrics
and
go
already.
F
So
that's
great
to
see
basically
like
the
the
general
pattern
with
Hotel
seems
to
be
that,
like
it's
kind
of
hard
to
figure
out
how
to
initialize
stuff
correctly
like
what's
like
the
most
correct
way,
to
sort
of
initialize
things,
and
then
from
that
point
it's
pretty
easy
to
like
go
ahead
and
use
apis
for
stuff
and
just
kind
of
pre-code
samples
from
there.
We
don't
need
a
whole
lot
of
content
like
we
can.
C
So
would
it
be
helpful
for
you
guys,
like
this,
is
an
outstanding
offer
to
help
with
docs,
because
the
project
always
needs
more
Docs?
Would
it
help
like
for
the
next
few
minutes?
If
we
just
go
through,
like
I,
mean
I
think
we
know
the
statuses
of
each
language,
but
but
specifically
like
where
they
are
with
their
metrics
docs?
So
you
can
see
where
you
need
to
help
that'd
be
valuable.
Is
that
a
waste
of
time
I.
C
So
Java
is
metrics
or
ga.
I
They're
available,
you
know,
I'm
not
sure
whether
there's
standards
that
are
kind
of
cross-language
but
I
took
a
stab
at
creating
tax
for
metrics.
Both
configuring,
the
SDK
and
using
the
API
to
initialize
some
of
the
common
instruments
and
and
use
them.
I
took
a
stab
at
that
a
while
back
and
so
they're
available
in
some
in
some
form.
B
There's
a
link
up
higher
in
the
the
meeting
notes,
but
it
looks
like
we
might
need
some
docs
for
configuring
if
you're
using
the
agent.
I
Well,
maybe
like
so
the
agent
the
the
way,
the
the
typical
way
to
configure
it
is
just
to
use
all
the
environment
variables.
So
unless
we
were
to
kind
of
duplicate
that
configuration
there's
special
ways
where
you
can
kind
of
have
hooks
to
do.
Programmatic
configuration
for
the
agent
but
I'm,
not
sure
where
I
mean
that's
documented.
In
the
the
agent
repository
the
open,
Telemetry
Java
instrumentation
repository
okay.
I
C
All
right
did
you
have
a
script.
Matrix
is
painting
GA.
We
already
talked
about
that.
Docs
are
spread
out
all
over
the
place
in
the
repo.
So
maybe
we
could
use
some
assistance
here,
because
we
know
that
JS
is
looking
to
do
the
JGA
next
week
or
very
shortly
afterwards.
E
Yeah
I
have
been
docs
has
been
something
that's
been
on
my
mind
for
a
while
I
know,
I
promised
Philip
that
I
would
do
it
weeks
ago
at
this
point
and
have
not
quite
gotten
there
I'm
sure
he's
getting
tired
of
waiting
for
me.
C
Like
Phil
might
be
offering
to
lend
a
hand
too
or
Severn,
so
that
might
be
might
be
a
great
one
to
Target,
because
I
think
it
sounds
like
for
Java.
Things
are
reasonable
for
go.
Metrics
are,
of
course,
still
in
Alpha.
Doc's
exists
in
thought.
Only
I
Isis,
please
someone
speak
up.
You
disagree,
I
suspect,
go
it's
probably
so
early
that
we
don't
want
to
invest
too
deeply
in
docks.
At
this
point.
J
J
B
Yeah
I
would
say
as
long
as
you
know,
even
when
things
are
done,
if
there's
like
some
assume
during
kubecon
everyone's
going
to
be
going
to
like
to
GitHub,
read
me
or
whatever,
hunting
around
for
docs.
So
as
long
as
there's
something
prominent
there,
that
just
says
like
hey,
it's
an
alpha
beta
coming
soon.
You
know,
that's
probably
good
enough.
J
Yeah
we
have
that
it
also,
we
also
have
like
go
docs
themselves
are
hosted
on
its
own
package
website
like
those
are,
those
are
accurate
and
I
think
there's
even
examples
there,
but
they
just
haven't
been
translated
into
open,
Telemetry
docs,
specifically
so.
A
F
I
think
what
what
I'd
say
to
that
is
so
long
as
there's
a
reasonable
degree
of
confidence
that
the
apis
you're
going
to
be
interacting
with
are
stable.
Then
it's
okay.
If
we
think
there's
going
to
be
some
churn,
then
it
might
not
actually
be
worth
the
trouble
to
get
it
on
the
site
because
then
like
they
might
turn
by
the
time.
F
J
I
think
that's
our
understanding
as
well,
but
I
think
also
by
the
end
of
the
beta
Milestone,
which
is
why
we
included
in
this
like
that,
should
not
be
the
case
or
I
mean
our
idea
is
that
beta
minus
zone
is
one
off
from
a
release
candidate.
So
that's
the
plan
but
yeah.
Currently
we
can't
make
that
guarantee.
C
A
I
think
we
have
a
pretty
decent
state
right
now.
Awesome
great.
C
Sweet
C,
plus
plus
metrics
or
RC,
because
we
saw
status
earlier
doc.
Severn
has
initial
getting
started.
Pr,
API
and
SDK
reference
dock
is
already
in
read
the
dot.
Yet
we
already
saw
that
earlier
today,
some
point:
we
need
to
move
it
to
open
telemetry.io,
it
might
be
worth
looking
at.
C
We
missed.net
I,
don't
know
if
we
have
any
of
the.net
people
on
the
call.
Does
anyone
happen
to
know
like
I
know
the.net,
metrics
or
GA,
but
does
anyone
happen
to
know
the
state
of
the.net
Docs
I
might
just
quickly
Google.
It
give
me
a.
F
Second,
they
tend
to
be
spread
out
a
little
bit
on
the
site.
There's
no
metrics
there,
but
I
know
that
metrics
are
pretty
damn
stable
in.net,
so
it's
probably
just
a
matter
of
getting
them
at
it.
They
have
examples
in
some
of
the
repos
that
are
pretty
well
fleshed
out,
but
not
really
dockified
much.
C
Okay,
so
maybe
we
can
look
at
that
one.
Let
me
just
put
a
note
in
here:
while
we
go
to
the
next,
so
I'll
put
a
note
and
then
whoever
wrote
Swift
do
you
want
to
speak
to
the
status
of
the
docs
there.
K
B
Anyone
know
the
status
for
Ruby
Visa,
V,
metrics,
I.
C
D
C
A
Speak
to
this
a
little
bit:
oh
hey
man,
there
there's
ongoing
work.
The
the
SDK
is
partially
complete,
so
yeah,
no,
no
firm
timelines
around
where,
when
okay
everything
will
be
finished,
but
work
is
in
progress,
cool,
but
not
something
people
should
be
trying,
probably
not
yet
tribal.
Okay
cool.
Unless
you
are
super
adventurous,
I
guess.
A
There
there
is
a
metrics
project
on
on
the
Ruby
repo.
If
people
want
to
kind
of
track
issues,
not
cool,
that's
all
I
have.
C
K
So
I'm
sure
everybody's,
you
know,
privy
to
the
changes
in
the
OTL
P
like
I,
think
the
the
recent
one
was
a
name
change
from
the
instrumentation
library
to
like
instrumentation
scope,
and
this
is
the
second
time
we've
seen
some
breaking
changes
that
have
gone
into
the
collector
and
thus
like
rendered
the
data
coming
out
of
otlp
or
exporter
and
the
Swift
SDK.
It
just
gets
kind
of
you
know
vanished
and
there's
no
response,
which
is
another
kind
of
side
topic.
K
But
the
big
problem
here
is
when
there's
breaking
changes
in
the
spec.
You
know
our
current
SDK
stops
working
and
this
isn't
a
big
deal
for
a
lot
of
the
other
language
agents.
K
It
really
easy
you
have
control
over
the
servers
they're
running
on,
but
in
Mobile
it's
a
little
different
because
our
you
know
people
using
this
to
monitor
their
apps.
Their
apps
are
running
on
third-party
hardware
and
getting
those
getting
those
users
to
update
their
app.
Even
if
we,
when
there's
like
an
update
like
this,
you
know
you
need
to
push
a
new
agent
out.
K
That
doesn't
happen
right,
so
we
need
to
have
a
better
solution
for
backwards.
Compatibility
on
this
stuff,
the
the
the
specification
actually
says
that
this
should
be
a
a
graceful
handling
of
of
the
changes,
but
that
didn't
seem
to
be
the
case
and
yeah
go.
L
Ahead
can
I
stop
you
for
a
second
I
think
this
change
started
a
year
ago,
so
I'm
not
sure
what
you're
talking
about
about
not
being
a
graceful
change
and
everything
so
started.
One
year
ago,
when
we
renamed
that
we
put
this
into
a
different
number
for
Json
compatibility,
so
I
I'm
not
sure
what
you
are
referring
to
here.
K
Okay,
well
maybe
my
my
expectation
of
a
graceful
change
is
that
it
would
be
handled
gracefully
throughout
the
stack
as
it's
being
adopted
and.
G
K
I
think
you
just
mean
by
like
like
gracefully
as
in
it
was
everybody
was
aware
of
it
I,
like
so
I
misunderstand
what
what
graceful
means
in
this
context.
So
to
me
what
my
big
problem
is
that
this
you
know
the
there's
not
backwards
compatibility
with
with
this
change.
I.
L
K
This
is
grpc,
so.
L
You
use
Proto,
then
you
sh,
you
shouldn't,
have
any
breaking
change
unless,
unless,
unless
when
you
upgrade
it
to
version
I,
think
15
you
you
ignore
the
deprecation
warning.
You
shouldn't
have
any
problem.
K
Oh
I
see
so
the
I
see
the
issue.
Is
this
wasn't
a
well
I?
Guess:
here's
the
problem
so,
regardless
of
the
transition
period,
the
the
issue
is
is
if
there
are
people
who
so
do
we
want
like,
like
the
question
is
like?
K
Do
we
want
open
Telemetry
to
be
adopted
in
production
because
that's
like
you
know,
we
want
our
customers
to
use
open,
Telemetry
and
and
if
we
make
these
changes
like
the
problem,
is
you
know
you
have
a
very,
very
long
tail
in
Mobile
of
adoption
for
new
versions
of
things
right,
and
so
we
might
have
like
years
years
going
back
and
so,
but.
L
You
know
this
change
shouldn't
break
that
so
so
let
me
explain.
The
protocol
encoding
works
with
IDs.
This
change
broke,
broke
goal
for
their
own
reasons,
broke
Json
intentionally,
and
we
announced
that
the
Proto,
the
Proto
problem
for
you
should
have
been
so
so
we
rename
a
field
okay
with
id5.
If
every
time
when
you
upgraded
the
proto-dependency,
you
follow
the
deprecation
messages.
You
will
never
have
a
breaking
change
because
on
The
Wire
this
was
not
a
breaking
change.
K
G
K
K
G
So
so
yeah,
so
bargain
there's
a
combination
of
situations
under
which
this
can
still
be
problematic.
I
think
the
the
important
thing
to
talk
about,
though,
is
we
should
be
putting
a
1.0
stamp
on
otlp
I've
got
this
on
the
agenda
for
the
spec
meeting
tomorrow.
I
think
we've
covered
all
of
the
bases
we
need
to
cover
and
we
need
to
get
a
1.0
stamp
on
it.
So
we
can
have
some
guarantees
about
stability
and
some
reasonable
assurances
that
you
know
five
years
from
now.
Things
will
continue
to
work.
Yeah.
K
Okay,
so
Define
making
a
mistake,
because
if
I
don't
do
anything
right,
it.
L
K
Quite
possibly
not
that
might
yeah
I,
don't
think
so.
No,
it
seems
to
be.
We
use
the
proto-dependencies
inside
the
exporter
itself,
referencing
their
their
classes
generated.
Okay,.
L
L
Let's
have
that
because
I
I
still
believe
that
it
shouldn't
happen,
because
it
was
just
a
rename
somewhere
and
again
Anthony.
Has
it
its
own
reasons
why
in
goal
broke
it
because
we
go
allow
people
to
upgrade
the
the
two
things
independently
and
then.
Hence
we
made
a
mistake
when
we
did
this
and
allow
sorry
allowed
that
to
happen
and.
G
It
also
happened
in
Python,
so
it
wasn't
a
go
only
thing.
There
was
a
problem
with
this
transition.
We.
E
G
G
C
L
L
Yeah,
so
that's
that's
the
point
and
if
it's
not
the
case,
we
need
to
learn.
I
mean
exactly.
L
That's
why
that's
why
I
want
to
learn
from
you
guys,
because
otherwise
we
we
not
only
be
able
to
do
any
changes
or
anything
and
just
the
the
think
of
let's
make
it
one
zero
and
never
upgrade
anything
or
never
improve.
Things
is
not
working
to
me
and
like.
G
C
So,
okay,
we'll
get
a
follow-up
on
that
and
then
bugged
in
Anthony
sounds
like
you
guys
need
to
chat
about
about
otlp
versioning
with
the
other
people
who
are
involved
in
that
moving
on.
We
have
the
GC
election
this
week.
Really
two
things
of
note
here.
If
you
are
a
heavy
contributor
to
open
Telemetry,
but
you
don't
contribute
a
huge
amount
via
GitHub.
C
There
are
ways
that
you
can
register
to
vote.
Those
are
available
in
the
link
to
this
blog
post
here.
Otherwise,
if
you
do
make
a
lot
of
contributions
on
GitHub,
which
probably
everyone
does
on
this
call,
you
are
automatically
registered
to
vote
to
vote
and
you
need
to
do
it
this
week.
You
can
click
here
to
find
out
more
about
the
voting
process
and
you
can
actually
enter
your
vote.
I
think
the
vote
is
on
the
eight
yeah
okay
on
the
18th
perfect,
which
is
tomorrow
through
the
20th
which
is
Thursday.
C
So
please,
please,
vote
I,
think
there's
four
GC
seats
that
are
opening
up
and
there's
a
big
number
of
candidates,
including
a
number
of
end
users
and
everyone
else
who
are
running
for
those.
So
we
have
a
very
diverse
set
of
people
running
from
it's
less
of
a
vendor
Fest
than
usual,
which
I
think
is
very
healthy
and
very
great
for
the
community.
So
it's
an
opportunity
for
people
to
select
who
they
want
to
govern
the
project.