►
From YouTube: 2021-06-24 Governance Committee private 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
A
It's
like
right,
depending
on
okay
cool
well
in
base
10.
Should
we
start
since
it's
1004
and
we
technically
have
more
yuri.
You
wrote
that
right,
the
hotel
community.
I
think.
C
We've
been
having
a
look
at
doing
this
initially
with
the
cncf
that
turned
out
to
be
a
real
pain
in
the
butt,
but
we
have
been
able
to
secure
space
at
nearby
sheraton
hotel,
so
we're
in
negotiation
with
them
to
get
a
price
on
on
the
room
and
the
event,
and
all
of
that,
and
once
we've
got
the
budget
together,
we'll
set
up
some
kind
of
sponsorship
tiers
to
go
around
and
ask
member
organizations
to
fund
it,
but
the
the
basic
idea
is
it'll,
be
an
unconference
capped
at
400
people
that
doesn't
require
a
kubecon
ticket
and
will
probably
be
cheap
to
enter
like
10
or
20
bucks,
which
we
could
waive
if
people
need
it
waived.
C
C
B
I
still
have
no
updates
on
if
they
will
extend
the
use
of
their
covid
like
verification
vaccine
after
us,
because
priyanka
wasn't
a
call
last
week
and
they're
out
today.
So
I
might
have
an
update
next
week.
C
Yeah,
I
I
was
poking
them
about
whether
they,
if
they
just
had
extra
space,
could
they
just
like
sublease
it
to
us
and
have
the
conference
just
be
like
you
still
needed
a
kubecon
ticket,
but
otherwise
we
would
just
run
it
ourselves
and
not
go
through
all
the
hoopla,
but
that
you
know
it's
not
as
painful
to
them.
So
we're
just
going
to
stretch
it
separate.
C
Great
so
that
could
be
a
potential
backup
plan,
but
it
looks
like
this
thing
with
sheraton's
working
out.
We
just
got
to
get
a
price
quote
from
him,
so
hopefully,
next
week
we'll
be
able
to
bring
that
around
to
people.
D
Which
day
are
you
thinking
one
of
the
conference
days
or
monday
or
tuesday
we're
thinking
tuesday
just
putting
it
on
my
calendar.
C
One
thing
we
we
don't
have
in
the
bag
is
any
kind
of
like
after
party
thing.
If
we
wanted
to
do
that,
so
if
people
have
any
ideas
or
leads
on
how
how
we
should
do
that,
it'd
be
helpful.
It's
not
clear
to
me
what
kind
of
emcf
stuff
is
happening
that
tuesday
night.
So
so
that's
that
that
one
part
we
haven't
worked
out
yet,
but
we're
just
gonna
try
to
get
the
conference
together
so
next
week,
more
info.
C
A
I
had
an
item:
that's
open,
histogram
stuff
it.
I
think
that
there's
like
a
technical
discussion
happening
that
I
believe,
is
moving
away
from
the
open,
histogram
implementation.
A
As
I
understand
it,
trying
to
follow
the
otep,
but
I
also
have
I
mean
awareness
being
a
recorded
call,
so
I
don't
want
to
like
over
share
whatever,
but
it
sounds
like
there's
some
kind
of
murky
in
my
mind,
flimsy
kind
of
legal
concern
about
it
from
some
parties
involved.
In
this
whole
thing
I
mean
the
the
software
is
now
cleanly
apache
license,
and
I
wasn't
sure
if
the
technical
arguments
that
are
being
made
are
just
sort
of
a
front
for
some
kind
of
I
mean
I
think,
the
concern
and
I'm
speculating
genuinely.
A
So
I
want
to
make
sure
that
that's
super
clear,
but
I
think
the
concern
is
that
there
may
have
been
some
code
that
doesn't
look
like
a
clean
room
version
of
open
histogram
that
didn't
like
properly
license
open
histogram
code
before
it
was
turned
into
an
apache
project,
and
so
there's
probably
some
desire
just
to
keep
it
as
far
away
as
possible,
and
I
honestly
just
think
that's
a
stupid
reason
to
make
a
decision
open
telemetry
which
didn't
have
any
such
ip
issues.
And
now
the
code
is
very
cleanly
licensed.
A
A
So
this
is
all
like
second
or
third
hand,
but
it
does
seem
to
me
that
there's
both
a
seemingly
legitimate
technical
thing
happening,
but
then
why
shouldn't
it
illegitimate,
but
I
would
say
irrelevant
to
open
telemetry
ip
concern
that
that
I
don't
want
to
to
trump
making
the
right
technical
decision,
because
people
who
understand
this
stuff
and
have
gone
way
deep
on.
A
It
believe
that
that
code
is
the
the
sort
of
the
or
that
approach
is
the
best
thing
out
there
in
terms
of
just
being
the
right
technical
solution
and
it's
going
to
be
difficult
to
change.
So
I
just
so
the
question
I
the
specific
question
is:
does
anyone
know
what's
going
on
with
that
stuff?
If
not
then
I'll
just
sort
of
take
this
offline,
but
I'm
just
trying
to
investigate
that
a
bit.
A
F
Okay
and
also
always
bring
lawyers
to
bear
to
ask
to
to
give
us
sort
of
an
outside
opinion
as
to
whether
or
not
it's
risky.
A
B
A
Not
our
problem
so,
but
it's
not
risky.
It's
a
very
clean
story
at
this
point
for
open
telemetry,
but
I
think
it's
sort
of
like
you
get
that
vibe,
where
people
are
like.
I'm
not
allowed
to
talk
about
that
word
or
anything
related
to
that
word.
I'm
like
okay!
Well,
that's
not
my
problem.
You
know
so
like
that's,
that's
that's
the
thing
I'm
trying
to
understand
better,
but
I'm
kind
of
out
of
the
loop
on
the
details
of
it.
A
So
I
think
what
I'll
do
I'll
I'll
share
the
otep
and
the
notes
here.
That's
like
all
pretty
you
know
technical
and
boring
frankly,
or
not
not
boring,
technical
and
and
undramatic,
and
then
I'll
do
a
little
bit
of
digging
to
figure
out
what's
actually
going
on.
If
this
is
just
a
technical
decision
to
move
away
from
that
than
soviet
and
not
like
trying
to
fight
that
fight,
I
just
want
to
make
sure
it's
on
tactical
merits.
So.
A
G
Yeah
so
trusts
and
email
a
couple
of
weeks
back
and
I
lost
my
inbox
saying
that
he
sees
a
delay
in
github
actions
execution.
So
maybe
it
was
just
a
hot.
G
F
I
can
reach
out
to
trask
and
see
where
that's
going
see
what
his
concern
was
and
if
he
has
followed
up
already
with
github,
because
he
may
have
done
that.
E
Yeah,
this
is
probably
just
more
of
a
fyi,
so
jaeger
documentation
all
over
the
place,
says,
use
open
tracing
and
we
discussed
like
changing
that
recommendation
to
to
recommend
an
open,
telemetry,
but
open
telemetry
doesn't
support
the
sort
of
one
of
the
popular
features
of
jager,
which
is
the
remote
sampling
configuration
and,
as
a
result
like.
If
we
were
to
say,
recommend
as
open
telemetry
is
the
case,
then
we're
losing
this
big
chunk
of
functionality.
E
That
jager
can
provide
today
right
and
we
even
like
rolling
out
and
you
like,
truly
adapt
to
something
on
the
back
end,
which
also
relies
on
that
same
configuration
mechanism.
So
the
the
one
thing
that
we
are
trying
to
do
so
above
all
kind
of
volunteer
to
build
the
prototype
of
essentially
implement
and
open
telemetry
sampler
using
the
jaeger
remote
configured
remotely
configured
sampler
and
in
java
just
to
see
how
that
works
and
like
what
sort
of
like
workout
and
then
maybe.
E
If
people
want
to
use
that
across
other
languages,
then
just
help
out
implementing
that
same
thing,
because
implementation
technically
already
exists
in
jaeger
just
needs
migration
into
open
telemetry,
but
it
is
work
that,
like
the
project
currently
can't
stop
really
across
all
languages
so
and
the
I
guess
the
the
one
open
question
there
would
be
is
whether
this
kind
of
capabilities
should
be
part
of
the
spec,
because
in
the
prototype
we
will
just
do
it
as
an
extension.
E
But
if
it's
an
extension,
then
it
creates
all
kinds
of
interesting
questions
like
if
you
have
a
default
sdk
distribution.
Do
you
get
that
extension
or
not?
How
can
you
configure
it
without
actually
making
code
changes?
Things
like
that
which
is
not
ideal
situation
so
like?
Ideally,
it
would
be
part
of
the
sdk
natively,
but
it
is
currently
jager
specific,
like
no
other
back-end
supports
it.
E
As
far
as
I
know-
and
it's
also
like
if
I
were
to
design
it
today,
I
would
probably
design
it
slightly
differently
in
terms
of
like
that
how
the
configuration
is
expressed.
So
it's
not
sort
of
like
obvious
to
me
that
this
is
the
best
way
to
say:
oh,
that's
exactly
how
it
should
go
into
the
spec
and
it's
not
even
obvious
whether
it
should
go
into
the
spec
or
not,
because
there
hasn't
been
agreement
in
that.
C
So
there
is
a
lot
of
interest
in
in
sampling,
but
I've
wanted
to
find
people
who,
who
have
concrete
use
cases
and
a
good
understanding
of
this
right
now.
Something
interesting
interest
seems
very
exploratory.
C
Could
people
show
up
to
help
lead
that
and
craft
a
spec,
because
I
do
think
this
is
something
that
should
be
put
into
an
otep
and
and
and
put
into
the
spec
having
like
a
fractured
landscape
around
the
control
mechanisms
for
stuff,
like
sam
sampling,
seems
like
an
area
where
open
telemetry
will
kind
of
fray,
and
since
there
isn't
like
a
huge
mortgage
board
of
different
ways,
people
do
this
right
now
it
seems
like
a
good
opportunity
to
to
maybe
modify
what
jaeger
is
doing,
like
you
said
into
something:
that's
maybe
slightly
better.
E
Yeah
I
mean
I
I
like-
I
know
that
joe
elliot
from
grafana.
He
is
actually
working
on
the
on
the
adaptive
sampling
in
the
jaeger
back
end
right
now.
It's
like
sort
of
the
code
existed
there,
but
never
was
production
lice
to
actually
use.
It
was
only
used
at
uber
internally,
so
he's
interested,
and
he
said
he
built
the
sort
of
the
proxy
for
the
sampler
configuration
in
hotel
collector.
E
Already
he
worked
on
that
part,
so
yeah
he
would
be
interested
in
that
there
and
I
think,
they're
actually
using
it
in
grafana
this
mechanism,
because
it
does
allow
you
sort
of
like
not
worrying
about
configuring
all
the
multiple
applications
in
in
the
organization.
Because
everything
comes
from
the
back
end
yeah,
I
will
be
interested
to
to
participate
as
well.
I
don't
know
if
I
can
drive
it.
I
don't
have
time
for
driving
it.
C
Yeah,
I
think
that's
just
the
thing
we're
looking
for
some
someone
with
like
a
concrete
enough
proposal
that
other
community
members
who
have
a
lot
of
interest
in
sampling
could
review
it
and
verify
that
it's
it's
something
that
could
eventually
see
broad
adoption
but
yeah.
If,
if
this
proposal
is
this
sounds
like
a
good
candidate
to
get
that
discussion
rolling
my
concern
around
sampling
is
people
have
just
a
lot
of
vague
ideas
about
what
they
want,
but
not
like
a
concrete
proposal.
This
sounds
like
a
good
one.
E
Okay,
well
I'll
think
with
jordan,
see
if
he
wants
to
take
the
lead
on
that
we're
actually
doing
a
very
similar
project
at
facebook
as
well.
So
the
there's
like
there's
additional
learnings
that
they
have
from
that
so
again,
yeah.
E
If
we
were
redoing
it
natively
in
hotel,
I
would
do
it
slightly
more
flexible
than
what
jager
has
today,
which
fallbacks
like
it
degrades
into
like
the
jager
model,
without
any
fancy
features,
but
it
like
it
does
allow
you
more
fancy
features
if
you
want,
because
we've
we've
had
users
asking
for
it
like.
Oh
what,
if
I
don't
want
to
sample
just
by
operation,
name
right.
C
Yeah,
no,
I
think
I
think
that
would
be
that
be
a
great
starting
point.
So
yeah
hit
me
up
on
on
slack
and
there
is
an
issue
that
I'll
send
you
a
link
to
on
restarting
the
sampling
sig.
So.
B
C
F
G
Yeah
I
missed
it,
but
do
we
get
maintenance
track
or
not.
G
B
Haven't
been
able
to
bring
that
up
yet
because
priyanka
wasn't
there
we're
not
making
the
deadline,
because
the
email
was
already
sent
out
for
maintenance
track
a
week
ago
and
sign
up
was,
is
dupe
by
july
7th,
it's
something
I
need
to
bring
up,
because
we
like
once
we
know
how
much
space
we
have
I'll
be
able
to
bring
it
up
and,
like
I
know,
I'm
gonna
I'll
have
to
do
a
poll
like
everyone
else.
That
makes
like
an
incubation
deadline
too.
F
My
guess
is,
it
went
to
the
cncf
technical
oversight
committee
and
the
public
mailing
list
there
so
that
they
can
get
comments
on
them
because
there's
there's
a
public
mailing
list
for
that,
and
there
are
lots
of
comments
and
opinions
there.
D
I
have
a
quick
question
for
ben.
I
was
just
looking
at
the
otep
that
you
linked
and
I
couldn't
find
any.
I
didn't
find
anyone
complaining
about
like
a
legal
argument.
Oh.
A
Yeah,
no,
it's
not
it's
not
there
that
that's
the
thing!
It's
like
that's
what
my
concern
is,
so
that's
coming
through
side
channels
and
then
and
then
I
look
at
the
discussion
and
it's
totally
technical,
I'm
like
okay.
Well,
I
just
want
to
I
I
mean
I'm
completely
comfortable
with
not
trying
to
opine
about
what
I
think
the
right
decision
is.
A
I
did
deal
with
this
exact
issue
in
monarch
and
I
have
an
opinion
but
like
there
are
enough
hooks
in
that
kitchen,
I
just
want
to
make
sure
it's
a
technical
decision
and
that
that's
exactly
the
issue
it's
like
if
it
was
happening
in
the
open,
I'd
be
like
cool
like
we're
talking
about
it,
but
I
think
that
it's
possible
that
some
people
are
like
well,
actually,
I'm
not
allowed
to
go
in
this
direction.
So
I'm
going
to
make
up
this
typical
argument
about
base
10
encoding
or
whatever,
but
it's
actually
not
about
that.
A
So
that's
what
I'm
trying
to
figure
out
does
that
make
sense
sort
of
yeah
I
mean
I,
I
don't
I
actually
my
my
opinion
like
based
on
what
I
can
see
is.
I
think
it
is
a
totally
legitimate
typical
decision,
that's
being
made
and
that's
fine.
I
just
want
to
verify
that.
A
That's
all
because
I
actually
thought
the
open,
histogram
donation
was
really
exciting
and
I
was
like
oh,
this
is
great
like
this
is
the
best
thing
out
there
and
then
we're
for
whatever
reason
we're
not
adopting
it,
and
I
mean
that
that
might
be
the
right
decision.
I
just
want
to
make
sure
it's
it's
happening
for
technical
reasons,.
F
A
We
didn't
say
yes,
I
mean
you
know
the
donations
that
are
being
you
know,
offered
or
whatever
to
open
telemetry.
I
think
it's
we've
talked
about
this
a
number
of
times,
but
it's
it's
complicated
right,
it's
like!
Even
if
it's
good,
I
don't
think
anyone's
arguing
whether
open
instagram
is
it's
like
a
good
piece
of
code
but
like
there
are
lots
of
other
concerns
about
whatever
depends
on
the
product,
but
I
I
think
it
was.
It
was
sort
of
like
offered
up
before,
and
there
was
this
concern
that
it
was.
A
It
was
kind
of
open
source,
but
with
an
asterisk
around
like
a
patent
that
it
was
like
there
was
sort
of
a
handshake
that
there
is
not
gonna,
be
any
prosecution
of
that
patent.
But
who
knows
right?
So
I
think
that
the
concern
was
raised
like
oh.
This
is
great,
but
it
has
this
this
like
thing
that
we
can.
A
Exactly
and
so
so
I
think
they
decided
okay,
fine
well
after
some
sleeping
on
it
for
a
few
months
like
we'll
remove
that
restriction
and
just
say
it's
apache
plain
and
simple,
and
and
and
I
I
don't
think
that
there
is
any
explicit
like
if
you
do
this,
we'll
accept
it.
But
I
think
maybe
there
was
an
implicit
version
of
that
that
that
was
communicated.
C
And
just
for
some
clarity,
my
very
vague
understanding
of
the
technical
issue
is
we
want
to
offer
a
variety
we
don't
want.
We
can't
pick
one
histogram
type,
so
open
telemetry's
protocols
designed
to
handle
multiple
types
and
then
open
histogram
would
be
a
great
type
to
offer.
C
However,
open
histogram
as
it
stands
would
further
complicate
the
protocol
beyond
where
it
may
need
to
be,
and
so
the
question
is:
there
are
open,
histogram
people
in
that
conversation
having
a
discussion
about
whether
the
current
protocol
would
give
us
everything
we
would
effectively
want
from
open
histogram
without
having
to
add
whatever
the
heck.
It
is
extra
that
they
would
need.
So
I
my
understanding,
that's
my
vague
understanding
where
the
technical
argument
has
been.
You
might
have
a
different
understanding.
I
think
that's
that's.
I
think
that's
thought
on
yeah.