►
From YouTube: 2022-03-02 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
Just
filling
in
the
agenda
here
with
some
notes
for
spec
pr
heaven
process,
progress.
A
B
B
A
A
At
the
moment,
we're
doing
fine
yeah,
just
looking
at
2022
and
trying
to
decide
if
this
is
a
year
where
live
events
are
gonna,
be
a
real
thing
again.
That's
very.
B
Interesting
yeah,
I
I
know
splunk-
is
having
a
splunk
conf
yeah
likely
going
to
so.
A
A
So
I'm
trying
to
decide
about
going
through
with
it
again.
I
I
think
I'm
hoping
that
by
the
time
kubecon
in
a
rolls
around
it'll
be
fine,
but
I'm
gonna
pass
on
trying
to
organize
one
for
kubecon
eu
this
time,
because
it's
like
too
much
of
a
pain
in
the
butt.
Basically.
B
No
that'd
be
that'd,
be
awesome,
though
okay
kind
of
event.
What
what
dates
are
you
looking
at.
A
Well,
we've
always
discussed
having
it
as
like
just
before
kubecon
like
essentially
like
a
zero
day,
kubecon
event,
just
because
lots
of
people
are
coming
into
town
anyways
for
kubecon,
and
so
that
would
be
maybe
nicer
to
do.
It
then
then
kind
of
on
its
own.
You
know
and
make
make
everybody
buy
a
plane
ticket,
but
at
the
same
time
having
it
as
actually
separate
from
from
kubecon.
A
But
you
know,
the
format
is
just
like
a
lecture
format
is
kind
of
what
they
require
and,
and
you
can't
just
go
to
the
one
day
event
you
have
to
buy
a
whole
cube
con
ticket,
and
none
of
that
really
like
sat
right
with
me
for
doing
like
a
community
day
like
I
really
want
to
do
like
a
community
day
where,
like
anyone,
can
come
and
it's
more
like
an
unconference
where
we
just
kind
of
like
you
know
like
basically
like
set
set
an
agenda
live
and
then
just
do.
A
You
know
maybe
have
some
state
of
the
union,
but
mostly
do
like.
You
know.
Small
group
breakouts
and
just
just
have
the
whole
point
of
it,
be
to
let
everybody
kind
of
like
hang
out
and
talk
with
each
other
and
get.
C
A
B
A
Oh
really,
oh
yeah.
I
it's
just
a
way
of
describing
a
conference
where
there
isn't.
There
isn't
like
a
speaker
audience
separation
so,
rather
than
showing
up
and
there's
like
a
pre-existing
agenda
of
talks
that
you
go,
listen
to
it's
more
just
about
organizing
it
as
a
way
to
to
create
breakout
groups,
so
everybody
can
kind
of
like
talk
with
each
other,
and
I
don't
know
I
always
get
way
more
out
of
events
like
that.
A
I
feel
like
I
can
just
like
watch
watch
talks
on
youtube,
but
I
don't
really
need
to
like
fly
somewhere
to
have
that
experience.
So
so
I
really
wanted
to
do
one
of
those
for
open,
telemetry
and
also
start
organizing,
some
just
like
local
meetups,
like
sf
portland,
new
york,
austin.
A
B
All
kind
of
in
sydney,
it's
all
kind
of
opening
up
right
now
they
just
got
rid
of
like
mask
requirements
and
stuff,
like
that,
it's
kind
of
just
we're
just
kind
of
on
the
tail
end
of
all
the
covered
stuff,
and
you
know
the
the
bars
and
the
I
feel
like
dance
floors
are
the
last
thing
to
open.
B
So
that's
like
that's
like
an
indication
of
like
what's
allowed,
what's
the
dot
once
the
dance
floor
is
open,
so
yeah,
I
think
I
think
I
mean
I
haven't,
got
plans
to
go
any
of
them,
but
I
think
there's
something
planned.
Yeah.
A
B
B
A
I
mean,
I
think
you
know
definitely
portland
sf
and
new
york,
oh
for
sure,
have
like
enough
people.
If,
for
no
other
reason
than
like
you
know,
like
their
observability,
I
mean
well
everyone's
there,
but
there's
also
like
observability
vendors
anchoring
those
those
towns.
So
you
won't.
You
won't
have
a
shortage
of
people
being
willing
to
come
like
give
a
talk
or
something.
A
B
A
A
Getting
getting
to
to
this
this
group
meeting,
I
have
a
proposal
that
I
wanted
to
walk
by
you
guys.
I
had
a
talk
with
with
bogdan
and
some
other
tc
members
about
how
how
to
like
further
clarify
the
specification
process
both
like
you
know
how
to
have
like
a
clear,
clearly
written
down
order
of
operations
for
like
when
something's
like
not
getting
attention
or
it
seems
stalled
for,
like
you
know,
project
internal
reasons.
C
A
You
know
in
the
case
of
the
spec,
like
you
know,
tc
or
approvers,
just
not
not
responding
or
something
like
that
or
the
other
case
being
like
there's.
No
there's
no
like
it's
not
clear
what
the
next
step
is
because
there's
like
not
agreement
forming,
you
know
that,
like
should
400s
be
errors,
is
like
an
example
there,
where
it's
like.
A
So,
like
you
know,
we're
not
like
rejecting
the
the
concept
of
doing
it,
but
because
we
aren't
gonna
be
able
to
give
it
the
attention
right
now
we're
gonna.
Like
close
the
issue,
you
know
and
just
ask
you
know
that
people
come
back
later
or
something
to
just
anything.
We
can
do
to
avoid
the
situation
where
it's
just
kind
of
like
open
issues
sitting
there
with
with
no
resolution
like
like
just
bringing
them
to
resolution
one
way
or
the
other.
So
thinking
about
how
to
do
that.
A
One
thing
that
came
out
of
it
was
just
clarifying
that
the
person
assigned
to
the
issue
is
actually
the
the
point
person
for
resolving
anything.
That's
that's
stuck
in
the
issue,
so
I
think
in
our
particular
case
I
know
dennis
reached
out
to
I
think
tigran
who
was
like
assigned
to
his
issue
around
you
know.
Bogdan's
comments
about
links
and
tigran
was
just
like
you
gotta
you
gotta
sort
out
bogdan.
A
If
they
can't
like
solve
it
themselves,
they
can
at
least
be
the
person
to
like
poke
the
people
or
something
and
and
if
like
github's,
not
working
making
it
clear
that,
like
going
to
the
spec
channel
on
slack
and
requesting
assistance
there-
and
you
know
at
mentioning
the
assignee
is
like
the
better
way
to
do
it
rather
than
like
private
private
messaging
people,
because
that
just
makes
it
a
little
more
public.
So
you
know
more
people
can
see
it
and
take
it
up.
A
A
This
whole,
I
was
thinking
of
putting
this
into
spec
repo,
because
it's
specific
specific
to
the
spec,
so
that
it'll
be
an
extension
of
the
there's,
a
contributing
document,
and
this
will
kind
of
expand
the
contributing
document
to
describe
this
and
then
like
add
a
blurb
in
the
at
a
blurb
in
the
readme.
A
To
kind
of
point
to
that,
the
the
other
piece
was
to
establish
a
concept
of
sponsorship
so
like
for
any
meaningful
proposal,
that's
gonna
need
like
a
no
tap
or
if
it's
like
ongoing
work.
It's
gonna
need
like
a
spec
sig
like
this
one
or
the
rum
sig
and
suggested
for
proposals,
but
definitely
for
sigs
when
they
spin
up,
there
needs
to
be
a
tc
or
equivalent
sponsor.
A
In
other
words,
the
tc's
members
are
the
maintainers
of
the
specification
repo.
So
if
this,
if
it's
meaningful
change,
they're
they're
gonna
have
opinions
about
it
and
just
I
think
the
way
some
of
these
new
spec
groups
have
had
a
little
bit
of
trouble
is
that
they
got
rolling
without
a
tc
member
kind
of
like
being
in
attendance
to
them
as
they
were
going.
A
And
so
we
want
to
change
that
and
say
like
going
forwards
if
we're
forming
a
spec
sig
part
of
doing
that
is
one
to
two
tc
are
interested
enough
in
it
that
that
they're
willing
to
to
attend
the
sig
if
they
can't
make
all
the
sig
meetings.
Okay,
but
they're,
willing
to
be
available
and
like
following
the
work
and
kind
of
be
part
of
it,
so
that
we're
not
getting
in
a
situation
where
it's
like
way.
All
the
way
to
the
point
where
we're
making,
in
other
words,.
C
A
Want
it
to
be
a
situation
where
prs
have
like
a
very
high
chance
of
getting
approved
quickly
because,
like
they've
been
worked
through
with,
you
know,
close
community
members
right.
So
you
know
in
this.
In
this
case,
for
example,
if
bogdan
or
somebody
had
been
a
tc
member
had
been
sent
to
this,
we
would
have
like
hashed
out
things
about
like
links.
A
You
know
ideally
like
in
the
sig
before
before
the
pr,
not
that
you
can't
have
a
discussion
in
the
pr,
but
that
that
should
not
be
the
point
where
a
tc
member
is
being
like.
Oh
whoa.
This
is
like
kind
of
like
a
weird
interpretation
of
the
specs
that
you
got
going
on
there,
like.
I
don't
know
about
it
right,
like
that's
that
that's
like
too
late
in
the
process
for
that
that
stuff
to
be
happening,
and
so
we
think
having
a
sponsor
like
basically
ensuring
there's
like
a
tc
member
involved.
A
So
I
think
that,
combined
with
the
having
establishing
that
the
ticket
assignee
is
like
the
point
person
for
resolving
any
weird
things
like
that,
combo
should
hopefully
avoid
us
getting
into
situations
in
the
future
where
it
just
gets
like
kind
of
confusing
as
to
like
what
the
next
step
is
supposed
to
be
and
we're
gonna
recommend
that,
like
for
even
for
non-sig
work,
but
for
like
individual
oteps,
people
are
creating
that
they
start
that
process
by
creating
an
issue
and
otherwise
getting
involved
in
the
specification
sig
community
and
trying
to
get
to
get
a
sponsor
first
before
digging
in
and
doing
a
lot
of,
like
significant
work,
just
to
make
sure
that
that
again,
like
when
oteps
are
proposed,
they're
they're
kind
of
like
flushed
out,
and
they
have
like
a
reasonable
chance
of
success.
A
And
if
it's
like,
actually
something.
That's
like
gonna
have
a
low
chance
of
success
either
because,
like
we
don't
want
the
change
or
because
it's
like
just
like
out
of
scope
of
like
what
we're
capable
of
doing
at
this
time
like
that
can
get
established
like
ahead
of
time
rather
than
than
later.
So.
A
Over
the
past
couple
of
days
with
like
gc
and
tc
people,
so,
but
I
wouldn't
yeah,
I
wanted
to
bring
it
by
here
and
just
kind
of
like
check
in
with
you
guys
whether
that
that
sounds
like
reasonable
shift
and
give
you
guys
kind
of
like
first
crack
on
any
suggestions
before
I
I
submit
some
pr.
B
Yeah,
it
definitely
sounds
reasonable
to
me.
I
think
that
is
you've
kind
of
addressed
the
problems
that
we
have
right
like
it's
it's
just
there
are.
You
know
a
lot
of
cases
where
appeals
just
become
stagnant
and
there
needs
to
be
a
process
to
address
that,
and
it
seems,
like
your
proposal,
will
do
that.
C
C
Process
so
you
mentioned
like
a
week,
we
can
have
an
assignee
and
a
sponsor
can
be
the
same
person.
A
I
think
so
yeah,
it's
just
you
know.
Every
ticket
has
to
get
like
someone
assigned
to
it
in
github,
so
the
assignees
can
rotate.
A
A
The
the
job
of
the
assignee
is
not
to
not
necessarily
to
like
review
like
they
think
they
should
review
the
proposal,
but
they're
not
like
the
finer
final
gatekeeper
as
to
whether,
like
a
pr,
gets
accepted
right
like
the
person,
the
purpose
of
the
assignees,
to
make
sure
that
pr
is
like
going
through
its
process
and
like
gets
the
merge
button
clicked
when
you
know
it's,
it's
got
enough
approvals
basically
and
to
make
sure
the
merge
button
you
know
doesn't
get
clicked
before
then,
and
to
kind
of
like
respond
to
stuff
in
the
thread
and
in
addition
to
that,
the
person
assigned
can
also,
you
know,
have
an
opinion
about
the
issue,
because
they're
gonna
be
like
a
tc
member
or
or
an
approver.
C
B
A
You
know
we
don't
have
any
state
violence
at
our
direct
disposal
to
enforce
our
decrees.
So,
at
the
end
of
the
day,
when
something
is
not
happening,
the
recommendation
I
want
to
put
is
to
to
raise
it.
A
Raise
it
in
the
specification
slack
channel
at
mentioning
people
there
and
if
that
is
not
working
directly
contacting
like
the
rest
of
the
tce
members,
I
think
that's,
I
don't
think
we
have
like
a
public
talk
to
the
tce
channel
or
something
like
that,
but
so
that
might
be
be
private,
just
pming
dming
them,
but
basically
that's
that
would
be
the
chain.
Is
you
know
it
should
be
resolvable
by
stuff
just
happening
in
that
git
issue
or
pr?
A
If
that's
not
working,
the
spec
channel
is
like
slack
channel
is
just
like
another
venue,
for
you
know
where
people
can
see
that
there's
a
stale
issue
and
you
can
like
send
someone
a
message
that
might
cause
a
different
notification
to
appear
on
their
computer,
that
they
might
see
and
then,
if
that's
not
working,
then
taking
it
to
the
tc
and
and
if
that
is
like
really
not
working,
then
taking
it
to
the
gc.
Because
then
it's
it's
like
a
governance
issue.
Essentially,
so
that's
that's.
A
Yeah
as
far
as
timing,
I
feel
like
we
currently
have
like
an
auto
stale
feature.
That's
I
think,
set
to
seven
days
of
like
no
activity.
A
I
I
kind
of
feel
like
like
a
week
is
like
the
maximum
amount
of
time.
You
know.
If
something
has
has
not
gotten
a
response
in
a
week,
then
it
is
definitely
stale
and
needs
to
get
like
like
immediate
direct
action
of
some
kind.
A
I
don't
know
if
we
want
to
say
you
know
something
sooner
than
that
than
like
you
know.
You
should
expect
responses
within
three
days.
I
think
that
might
be
be
reasonable.
A
Maybe
a
nuance
here:
another
thing
we
discussed
was
being
more
intentionable
about
using
the
request
changes
feature
of
github
reviews
to
indicate
that
something
is
is
a
blocking
issue
and
I
think
it's
like
maybe
reasonable
to
say.
If,
if
you
do
re,
if
you,
if
you're
initiating
that,
then
you
need
to
respond
faster
right.
You
need
to
have
like
a
response
time
in
like
under
three
days.
Maybe
for
that
specific
situation,
because
because.
B
A
Not
there,
like
that's,
that's
just
like
pure
pure
sand
in
the
gears
so
and
it's
also
we
want
to
maybe
use
that
to
make
it
clearer
that
saying
conversations
which
are
not
requested
changes
like
like
they
should
be
resolved,
but
that's
actually
not
a
blocker.
A
You
know
to
to
not
read
that
as
like
a
blocking
of
of
the
issue
and
that
actually,
yes,
sorry,
go
ahead.
Yeah.
C
A
Yeah,
because
I'm
a
little
nervous
about
saying,
like
all
communication,
must
be
like
happen
within
like
two
days
or
where
it's
a
problem,
because
it's
like
a
big
project
right,
but
it
is
reasonable
to
say
like
if
you've
clarified
that
you
are,
you
are
blocking
something
intentionally
with
demands.
You
know
or
requests
of
some
kind
like
you,
then
are
on
the
hook
for
for
being
being
fast,
yeah.
C
Dad
actually
sorry
go
ahead.
Yeah.
I
just
want
to
ask
one
more
question
about
the
sponsor,
so
you
mentioned
it
can
be.
We
can.
We
can
leverage
some
another
kind
of
sig
meeting
to
to
basically
find
the
sponsor.
So,
for
example,
if
I
don't
know
how
I
can
contact
about
some
particular
question
or
some
particular
proposal,
I
have.
A
You
can
do
that
by
going
to
the
spec
meeting
or
going
to
the
the
the
specification
slack
channel.
I
think
we
want
to
just
start
using
that
channel
as,
like
our
main
point
of
asynchronous
communication,
because.
A
If
you're,
if
it's
just
like
a
question
like
I'm,
not
sure
what
to
do
next
or
something
that's
fine,
as
far
as
like,
like
formally
trying
to
like
get
a
sponsor
or
something
like
that,
that
should
happen.
That
can
happen
by
by
opening
an
issue
saying
you
know,
I
have
this
proposal,
I
want
you
know,
here's
my
proposal.
A
I
would
like
this
to
be
sponsored,
so
I
can
like
write
it
and
go
through
with
it
and
so
that
that
would
be
part
just
trying
to
change
it,
so
that,
like
proposals
start
with
an
issue
saying
like
I'm,
proposing
the
spec
change
and
I'm
seeking
sponsorship,
if
it's
a
big
big
change,
you
know
and
maybe
for
changes
that
just
like
roll
in
off
the
internet
as
like
a
pr
or
something
you
know
kind
of
like
directing
them
to
that
process
of
being
like
you're
asking
for
something
that's
gonna
take
some
time.
A
A
We're
not
going
to
just
like
be
like
yeah
cool,
we're
gonna,
add
whatever
to
the
spec,
unless
it
is
literally
like
a
tiny
change,
but
for
ongoing
work
like
what
we're
doing
where
it
would
be
really
annoying
to
do
that
for
like
every
single
pr.
A
C
A
Yeah,
it
was
actually
funny
in
this
particular
case,
of
the
the
http
retries
issue.
Bogdan
actually
wasn't
trying
to
plock
it.
When
I,
when
I
talked
to
him,
he
was
like
he
was
saying,
like.
A
A
Oh
sorry,
let
me
go
poke
people
for
you
when
you
poked
him,
like
none
of
that
worked
the
way
it
should,
but
it
was
also
funny
that
when
I,
when
I
got
him
on
the
phone
was
like
we
started
talking
about
the
specific
issue
he's
like
oh
yeah,
I'm
not
actually
trying
to
block
that
one.
Like
I
said
my
piece
like.
C
It
will
be
good
to
have
this
mentor
but
sorry,
but
to
have
a
sponsor,
but
it
might
happen
that
there
is
no
sponsor
right.
For
example,
people
like
to
see
people
just
available
and
they
might
be
interested
in
this,
but
they
have
no
time
and
have
no
bandwidth
to
you
know
sponsor
this
stuff.
A
A
This
is
like,
like
cigs,
like
ongoing
development,
not
like
we
can't
control
someone
coming
in
just
from
the
internet
with
some
proposal
out
of
nowhere,
but
for
like
work
that
we've
kind
of
like
approved
right
like
like
when
we
spin
up
a
sig
where,
like
we're
approving
this
work,
to
be
done
right
and
now
the
point
of
the
sig
is
to
to
figure
out
how
to
do
it.
But
if
there's
like
no
bandwidth
to
pay
attention
to
like
that
work,
that's
being
done
that!
That's!
C
And
that's
that's
exactly
the
issue
right
now,
probably
just
because
we
have
sick.
We
already
have
this
disgusting.
Actually
we
discussed
this
many
times
right
and
it
was
a
full
consensus,
at
least
in
my
mind,
that
links
are
something
which
allow
us
to
do
like
you
know
just
to
cover
all
all
the
things.
C
But
you
know
probably
you
are
only
like
a
dc
representative
here
and
you
know
it's.
We
have
this
kind
of
yeah
the
same
problem
right
so.
A
A
Like
I'm
helpful,
I
think
I'm
hopeful
because
you
know
like
I
know
how
all
this
stuff
works,
I'm
on
the
gc,
so
I
see
like
yeah,
like
my
form
of
solving
these
issues,
is
doing
this
thing
that
I'm
doing
right
now,
which
is
like
to
solve
the
process,
but
I
try
to
I.
I
try
actively
to
avoid
solving
these
individual
issues
by
just
like
doing
some
back
channeling,
because
then
we're
like
we're
never
going
to
actually
like
fix
our
process.
If,
like.
C
Yeah,
it
will
be
probably
good
to
have
some
dc
power
folks
participation
here.
Yes,
so
if
you
require
sponsorship
or
you
know,
recommend
sponsorship,
it
will
be
good
to
people
paying
attention
to
this
particular
sick
and
everything
which
is
happening
here,
so
they
they
can
say
up
front.
Do
they
have
bandwidth
or
not
yeah.
B
B
C
Want
to
do
something,
but
we
are
blocked
because
of
some
another
thing
happening
in
some.
A
A
If
it
feels
like
it's
suddenly
getting
like
hashed
out
all
over
again
from
scratch,
then
it
feels
like
there
was
some
context
missing
right.
There
were
some
stakeholders
who,
like
should
have
been
loaded
up
with
that
context
along
the
way
and
and
again
for
work
people
do
where
they've
done.
The
work
like
outside
of
the
sig
process
like
like
when
the
rum
stuff
first
showed
up.
A
It
was
some
open,
telemetry
members
right,
but
they
just
kind
of
like
here
is
like
our
fully
fledged
like
rum,
otep
proposal
and
of
course
it
like
didn't
it
didn't
fly
because
you
know
like,
like
a
bunch
of
people,
were
then
like.
Well
what
about
this?
What
about
that?
I
don't
know
I'm
like
super
skeptical
about
this
approach
or
whatever,
and
so
then
we
formed
the
sig
out
of
it,
but
that
sig
doesn't
didn't,
have
a
tc
person
assigned
to
it.
A
A
Yeah,
no
power
and,
and
and
and
just
some
missing
context
like
so
these,
basically
the
sig's
spec
specification
is,
I
mean
it's
just
like
any
other
project
right
like
if
it
was
java.
Work
like
the
java
maintainers
have
to
be
involved
on
some
level
if
people
are
gonna
like
design
some
java,
otep
open,
telemetry
stuff,
like
you
know,
for
the
spec,
the
tc
members
are
the
maintainers
of
the
spec,
so
you
know
they.
They
need
to
be
involved
in
the
ongoing
work
on
some
level.
A
So
so
I
think
that
that
should
should
solve
it,
and
you
know
that's
that's
basically,
this
right
now.
That's
this
sig
rum
and
ebpf
is,
I
think,
the
other
one
that
that
are
like
kind
of
like
spec
groups
that
don't
necessarily
have
a
tc
member
like
officially
attached
to
them,
as
opposed
to
like.
B
B
All
the
team
members
yeah
so
do
we
should
we
go
and
find
a
tasty
member?
Do
you
have
any
suggestions
of
someone
that
would
be
interested.
A
I
don't
have
a
direct
suggestion
at
this
time,
but
but
I'm
gonna
make
this
pr
and
then
yeah
as
part
of
that
you
know
talk
to
the
tc
and
like
hey
so
like
y'all,
just
figure
out
who
it's
gonna
be
for,
for
you
know,
instrumentation
ram
and
ebpf
like
who's
who's,
the
right
person
or
people
who
are
interested
enough
in
this
to
to
follow
it,
and
you
know
we
may
also
just
like
want
to
have
some
direct
conversations
with
bogdan.
A
It
would
be
easier
to
do
that
being
like,
let's
just
like
have
a
meeting
with
some
of
the
right
people
to
talk
about
links
specifically
and
get
those
hashed
out
like
bogdan,
doesn't
have
to
be
the
sponsor.
I
think
he's
like
pretty
stretched
pretty
thin.
So
I
don't.
I
don't
think
I
would
recommend
him,
I'm
looking
more
at
like,
like
maybe
armin
or
carlos,
except
they
are
they're,
both
eu
time
zone
people.
A
So,
just
among
the
the
people
who
are
who
are
currently
on
the
on
the
tc,
who
are
in
the
right
time
zone
to
come
to
an
eight
pack
meeting
but
but
that'll
be
on
them,
so
so
hopefully
we'll
get
that
resolved
going
forwards.
I
do
think
like
we
should
hopefully
be
kind
of
unblocked
in
our
work.
A
I
don't
know
dennis.
If
you
saw,
I
think
bogdan
commented
more
on
that
issue
yeah,
so.
C
Basically
yeah
he
he's
proposing
to
remove
spams
spam
links
right
now,
yeah,
the
only
problem.
Actually,
I
didn't
respond
because
I
wanted
to
wanted
to
discuss
it
with
you
guys
and
looks
like
we
had
this
consensus
last
time
and
for
now,
if
we
remove
these
links,
basically
we
are
removing
glue
out
of
it
and
all
these
pieces.
C
We
we
just
basically
if
we
have
no
links
there,
we're
just
stating
that
we
want
to
do
physical
kind
of
yeah
for
each
retry
and
redirect
that's
it,
but
how
you
are
not
answering
to
the
question
how
to
how
to
deal
with
this
data.
So
we
are
setting
all
these
data
about
how
to
correlate
it.
There
is
no.
There
is
no
way.
C
Right
it
looks
like
it's
like
a
you
know:
it's
not
fully,
I
would
say
solid
proposal
then
so
I
I
would,
I
would
be
rather
like
you
know,
holding
them
for
now
and
maybe
or
like
putting
them
for
now
and
then
to
figure
out
how
what
what
can
be
done,
because
if
we
remove
it,
it
will
be
half
done
this
way
and
I'm
not
sure
that
we
can
do
this.
You
know
we
can
call
stable
the
specifications
table
with
this.
C
C
And
it's
also
like
a
related
to
the
overall
kind
of
road
map
that
we
want
to
have
for,
for
we
want
right,
because
we
really
want
to
have
to
make
to
to
call
this
specification
stable,
because
we
have
a
lot
of
dependency
on
this.
So,
for
example,
here
they're
here
in
microsoft,
people
are
really
really
depend
on
it
because
they
can
know
they
cannot
move
forward
until
it's
stable,
yeah.
B
That
that's
just
to
really
like
that
is
the
most
important
thing
for
us,
like
is
the
stability
of
this
manic,
invention
that
actually
like
once
that
is
stable.
It
solves
so
many
problems
for
us,
and
it
like
makes
the
developers
atlassian's
lives
like
a
lot
easier,
and
we
can
do
all
this
cool
stuff
yeah.
The
most
important
thing
for
us
like
pretty
much
in
anything
with
a
hotel,
yeah.
A
And
I'm
hoping
like
it
has
been
slow,
but
I'm
hoping
it
will
go
a
lot
faster
at
this
point
one
when
we
have
like
a
tc
member
here.
I
think
that
will
resolve
some
of
that.
I
do
think
it
is
worthwhile
to
have
a
discussion
about
links.
I
don't
think
it
should
take
six
months,
but
that's
like
like
a
lot
of
the
work
we're
doing
in
these
instrumentation
groups
like
are
leveraging
links
so
like
in
some
sense,
whether
we
put
it
in
now
or
later,
like
it,
won't,
really
be
stable.
A
If
there's
like
some
big
unresolved
design
question
around
links
right
like
like
we
put
in
now,
but
then
like
we
decide,
this
is
like
bad
and
we're
gonna
like
not
use
links
for
this
because
it
it's
it
spreads
them
too
thin
right,
like
maybe
it's
bad,
to
use
links
both
as
like
the
thing
that
are
gluing
together.
These
big
trace
graphs
and
these
things
doing
these
tiny
little
reference
links
within
it
within
a
trace.
Maybe
it's
bad
to
use
the
same
thing
for
both.
C
Right,
yeah,
that's
a
valid
point
and
we
don't
know
at
the
moment
like.
Is
it
bad
or
not,
but
looks
like
the
overall
thing
that
we
discussed
and
all
the
other
proposals
for
actually
for
this
particular
problem.
Yeah.
B
C
For
any
kind
of
other
things
that
we
want
to
do
within
a
like
this
messaging
group
right,
so
it
can
be
also
the
same
trace,
but
they
want
to
use
links
to
indicate
kind
of
different
phases
of
this
of
this
process
yeah.
So
they
they
definitely
will
will
use
links
for
the
same
exact
purpose,
but
just
to
reflect
different
different
aspects
of
it.
B
C
So
and
yeah
I
mean
at
this
point
I
I
look.
It
looks
like
that.
We
already
know
that
links
are
really
useful
here
and
not
only
here
but
in
many
different
places,
but
they
feel
like
a
later.
For
example,
in
six
months,
one
year
figure
out.
That
lynx
is
something
that's
you
know
just
the
wrong
tool
for
for
these.
For
these
purposes,
definitely
we
need
to
make
a
break
and
change,
but
we
in
this
case
we
do
have
this
major
major
numbers
right.
So
we
can
go.
C
A
Yeah
we
do
want
to
use
schema
versioning.
This
is
like
a
practical
matter
that
I
brought
up
before
around
being
able
to
do
schema
translations
and
any
change
that
we're
gonna
make
after
we
declare
something
stable,
like
has
to
be
within
the
boundaries
of
a
schema
translation,
to
make
the
at
minimum
the
new
thing
backwards
compatible
with
the
old
thing.
A
So
it
is
as
like
a
a
basic
concept,
but
what
makes
me
concerned
is
we
don't
we
don't
have
any
implementation
of
that
right
like
there's
not
like
you
know.
In
theory,
you,
the
collector,
has,
like
a
bunch
of
you,
know
a
rules
engine
for
for
making
translations
and
changes,
but
we
don't
have
like,
like
we've,
never
actually
like
tested
this,
and
I
feel
like
there's.
A
A
You
know
like
in
some
places
you
will
want
them
in
other
places
you
may
not
like
do
there
does
there
need
to
be
like
a
predicate
language
for
knowing
when
to
trigger
them,
and
where,
like
I
don't,
I
just
I
suspect,
there's
there's
stuff
around
trying
to
do
it
for
real
in
the
sense
of
like
open
telemetry
like
shipping
schema
translation
modules.
It's
like
part
of
the
collector.
C
A
D
B
A
B
A
Think
I
think
you're
right.
I
think
it
was
sean.
They
were
thought
it
would
be
interesting
to
take
on
take
on
doing
some
of
that
work
so
that
that
that's
the
thing
we
need
to
do
you
know
as
part
of
getting
this
stuff
declared
stable.
A
Just
just
I
mean
like
we
don't
have
to
do
it
before.
We
declare
things
stable,
but
that's
going
to
make
a
hell
of
a
log
jam
to
the
first
post,
stable
change.
We
want.
A
Yeah,
I
just
I
feel
a
little
uncomfortable
calling
things
stable
without
well.
That's
still
a
little
hand
wavy,
but
I
I
really
want
to
avoid
a
2.0
with
with
these
things
you
know
like
in
general,
like
we
try
to
bat
aside
any
concept
of
a
2.0
with
open
telemetry
like
like
anywhere
and
like
so
far,
we've
been
successful.
A
You
know
so
yeah.
You
know
we
could
majorly
screw
up
with
something
like
links,
but
I
I
just
feel
like
it
should
be
totally
possible
to
have
done
our
due
diligence
before
calling
it
stable,
meaning,
like
you
know,
we
wrote
it
in
the
spec.
We
changed
some
instrumentation
to
use
it
asher
or
somebody
has
a
back
end
that
like
actually
processes
these
things,
you
know,
and
so
we
know
it's
and
and
then
we
hit
hit
stable
on
it.
C
Yeah
so
look
looks
like
looks
like
that's
exactly
what
we
already
confirmed
apart
from
this
later
part
of
the
platform,
because
they're
still
waiting
to
specification
to
be
stable
to
you
know,
plan
this
work.
It's
not
really
easy
thing
to
do
on
a
platform
level
to
you
know,
let's
say
process
all
these
links
and
re-re-build
building
the
overall,
like
you,
know,
different
structures
out
of
it
so
yeah
but
yeah
for
now.
Like
I
still
like,
I
don't
don't
know
what
what
will
can
be
the
next
step.
C
So
it
looks
like
background
really
don't
want
us
to
to
use
links
at
the
moment,
but
I
see
no,
no
one
like
no
like
a
better
alternative
yeah.
So
maybe
as
a
consensus.
C
Maybe
we
can
do
it
in
a
way
that
we
can.
You
know
change
shoot
to
me
when
it
comes
to
links
and
it
will
come
and
allow
us.
You
know
to
relax
this
this
kind
of
requirements
in
in
turn
that
it
can
be.
It
will
be
always
backward
compatible
right
so
because
it
will
be
basically
up
to
you
if
you
want
to
links
or
not
yeah,
isn't
it
sorry
yeah?
C
A
Yeah,
so
just
for
some
clarity,
what
bogdan's
suggesting
is
like
a
process
we
do
with
spec
work,
which
is
if,
if
90
of
a
spec
proposal
is
like
uncontroversial
at
that
point,
and
there
just
needs
to
be
further
discussion
on
on
one
part
like
not
blocking
it.
But
just
saying
like
like
we
want
to
just
talk
a
little
bit
more
about
this
one
park
that
the
process
we
tend
to
do
is
then
say
like
okay.
A
Let's
just
accept
this
90
here
and
immediately
open
another
pr
with
the
last
10
percent
and
continue
the
discussion
in
that
pr,
just
just
to
to
keep
the
discussion
clear
and
not
hold
the
train
up,
and
so
I
guess
my
question,
for
you
is
if
we,
if
we
did,
that
not
just
as
a
way
to
like
literally
hopefully
have
a
very
quick
conversation
about
links
and
then
approve
that
bit.
Would
that,
like
completely
gut
this
proposal
like?
C
C
And
whatnot,
so
basically
links
is
basically
a
glue
right,
it's
something
that
makes
it
possible
to
reconstruct
this.
These
scenarios,
like
on
the
platform
level,
going
forward
once
you
receive
this
stuff,
without
least
it's
not
really
possible
right.
But
but
overall
thing
is
just
it's
just
you
know
it's
still.
It's
still
valid
because
currently
there
is,
there
is
no
mention
of
this.
What
exactly
should
be.
C
Every
implementation
just
do
it,
you
know
yeah
or
wherever
they
want.
So
it's
it's
still
valuable,
but
I
I
feel
that
we
cannot
make
this.
So
definitely
it
will
be.
You
know
our.
It
will
allow
us
to
complete
this
pr,
but
it
will
not
allow
us
to
probably
make
the
the
specification
stable.
Yes,.
A
Yeah,
absolutely
so,
I
think
just
the
the
request
going
forward
is
like,
if
you,
if
you
just
pull
that
out,
we
can
immediately
accept
this
stuff
like
pull
it
out
and
just
like
immediately
put
it
into
a
separate
pr,
so
that
we
can
make
sure
we're
getting
like
quick,
final
final
discussion
and
sign
off
on
on
the
link
stuff,
specifically,
rather
than
having
this
like
endlessly
growing
set
of
pr
comments.
That's
because
we
don't
think
like.
A
Even
if
we
decide
okay,
we're
not
going
to
use
links,
we're
going
to
use
attributes
or
something
for
some
reason.
I
don't
think
we
would.
But
let's
say
we
do
like
it,
wouldn't
really
like
undo
any
of
the
the
rest
of
the
work
we're
putting
in
here
we're
definitely
not
going
back
to
having
a
logical
span
and
like
all
of
those
things.
A
So
it's
just
my
hope
is
if
we
can
get
this
like
cleared
up
in
a
in
a
week
and
and
be
done,
but
it
might
be
a
little
bit
easier
to
to
have
the
links
since
we're
like
the
part.
We're
trying
to
specifically
discuss
is
the
link
part
to
just
have
that.
Be
pulled
out
into
a
smaller
pr,
so
it's
just
it's
just
like
clearer.
C
Yeah,
it
makes
I
will
definitely
do
it
just
just
right
after
this
meeting,
and
hopefully
we
can
just
make
this
completed.
Yeah
and,
like
I
have
this
focused
discussion
in
another
one,
I
will
explicitly
mention
bogdan
there
yeah.
Hopefully
he
will.
It
will
be
possible
for
him
to
join
as
soon
as
possible
to
join
as
soon
as
possible,
right.
C
A
Yeah,
I
think,
going
forward,
hopefully
starting
next
week.
You
know
we'll
have
like
a
tc
member
like
assigned
to
this
group
and
and
like
attending
meetings
and
stuff
and
so
yeah.
That's.
My
hope
is
like
that.
This
has
been
like
like
long
and
convoluted,
but
but
once
we've
done
all
of
that
work,
my
hope
is
that
it
will
go.
A
Everything
will
be
able
to
go
a
lot
faster
at
that
point,
because
we'll
have
established
establish
some
ground
rules
about
links.
We'll
have
people
paying
attention
to
us
and
the
rest
of
the
http
stuff
should
hopefully
go
quickly
and.
C
A
Yeah
and
see
sql
and
database
stuff
will
have
their
own
shenanigans,
I'm
sure
for
defining
them,
but
you
know,
but
we
won't
hope
we
don't
need
to
be
revisiting
this
kind
of
stuff
like
be
able
to
like
reuse
what
we're
doing
here
and
then
like
the
messaging
sick.
So
so,
hopefully
for
the
remaining
conventions.
A
C
So
yeah
well
some
subject
matter
like
our
experts
was
with
dc
label.
Yes,.
B
A
C
A
A
Let's
try
strongly
to
have
like
a
way
that
we're
gonna
do
it
and
you
know
if
it
turns
out
to
be
like
in
practice
like
really
terrible.
A
C
A
Yeah,
exactly
that's!
That's
where,
like
having
a
a
sponsor,
is
helpful,
I'm
not
sure
who's
assigned
to
that
issue
james,
but
that
would
be
the
other
thing.
Is
you
just.
A
Sorry
who
carlos
carlos
yeah
yeah
so
like
a
next
step,
there
would
be
yeah
in
the
spec,
like
slack
channel,
just
like
pinging,
carlos
and
being
like
hey
you're
assigned
to
this
issue.
Can
we
get
it
it's
this
issue
is
getting
stale.
Can
we
get
a
some
next
steps
laid
out
for
how
to
how
to
how
we
should
resolve
this
or
something
you
know
something
to
that
effect
and
and
yeah?
It
should
be
his
his
job
at
that
point
to
to
sort
it
out.
Okay,
I'll!
Do
that.
B
Really
quickly,
I
know
we
just
ran
out
of
time,
but
if,
if
this
is
something
that
the
technical
committee
is,
it
looks
like
it
has.
Eight
people
is
there
like
what
is
the
pro
like?
What
is
the
is
that
the
magic
number
that's
just
chosen,
or
can
there
be
more
people
added,
because
if
there's
like
more
scope
to
what
they're
doing
now,
there.
A
There
can
be
more
people
added,
they're,
very
conservative,
about
about
adding
new
people,
because
we
want
new
people
to
have
been
around.
A
A
You
know
kind
of
co-founded
the
project,
but
he's
since,
like
moved
on,
like
he
moved
over
to
google
and
then
like
he's,
moved
over
to
like
working
on
kubernetes
stuff,
so
he's
like
like
effectively
he's
like
still
on
the
tc,
but
he's
like
effectively,
not
not
a
member
and
we've
been
trying
to
like
work
out
a
way
to
to
just
have
or
sorry
yeah
he's
he's
moved
off
the
gc
without
the
tc
yet
and
we're
kind
of
trying
to
push
for
a
way
to
have
like
emergency
status
for
maintainers
and
tc
people.
A
This
is
like
a
problem
in
like
every
big
open
source
project
like
kubernetes.
Has
this
too?
It's
just
people
like
if
it
weirdly
feels
bad
to
to
like
give
up
a
role
like
like
you're,
letting
people
down
or
you're
being
booted
or
like
like
somehow
something
bad
is
happening
because
you're
you're
being
removed
from
being
a
maintainer
of
a
thing
or
something,
but
we
kind
of
want
to
get
get
over
that,
but
anyways
long
story
short
like
in
practice.
A
I
think
the
tc
is
more
like
six
or
seven
people,
I
think
yuri
is
the
other
person
who's
like
around
and
like
will
respond
to
things
and
like
has
good
insights
about
stuff,
but
is
not
like
does
I
have
the
bandwidth
to
really
necessarily
be
called
like
a
full
tc
member
in
the
idea
of
like
being
able
to
do
all
this
day-to-day
work?
A
Fair
enough,
yeah
and
those
people
are
all
are
all
very
active,
like
the
the
the
only
issue
I
think
with
some
of
them
is
they're
also
like
write
a
lot
of
code
for
the
project
right,
and
so
it's
like
actually
like
they're
kind
of
just
a
little
bit
overclocked
on
open
telemetry.
B
A
Yeah,
I
was
a
fan
of
pair
programming
for
a
while
because
it
it
eliminated
code
review.
B
D
B
Programming
session
with
bogdan
dennis
and
you
can
just
sort
it
out-
yeah
yeah.
D
A
Really
yeah
yeah
you'll
see
that
pr
coming
soon
and
yeah
just
ping
me
in
that
spec
channel.
If,
if
you
run
into
any
more
roblox.