►
From YouTube: 20191008 sig arch conformance
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
A
C
B
B
A
Typical
modus
operandi
for
large
sweeping
changes
that
are
API
driven
things
that
people
might
be
depend
upon
is
to
send
a
detailed
description
that
outlines
the
issue
to
the
devel
list
and
point
them
at
the
issue
so
that
you
can
get
feedback
I.
Don't
think
that
that
would
be
a
bad
idea
in
this
case,
I'm
sweeping
me
a
broad
consensus
on
the
problem
statement.
C
B
C
Yeah
so
I
just
wanted
to
have
the
discussion
around
you
know,
you'd
you'd
said,
looks
good
enough
and
put
out
time
out.
I
did
just
push
a
small
tweak
to
it,
because
it
was
confusion
here
about
sort
of
level
of
expectations
of
what's
supposed
to
come
out
of
this
tooling,
which
is
the
cap.
Maybe
it
was
certainly
setting
the
expectations
higher
than
then
I
think
is
reasonable.
C
C
Those
behaviors
so
that
we
can
prove
them
independently
and
we
have
a
punch
list
of
the
expectation
here-
is
that
you
can
list
the
behaviors
with
much
less
effort
and
you
can
build
tests
for
those
behaviors.
So
you
can
essentially
list
the
behaviors
have
a
punch
list
of
things
that
you
want
to
cover
and
you
can
get
a
better
better
measure
of
your
coverage
as
well.
As
you
know,
your
task
list
and
you
can
independently
approve
those.
C
A
A
D
A
C
Okay,
yeah
Jeffrey
here
is
sort
of
done
a
little
bit
around
that
tooling
and
and
generating
the
list
of
behaviors.
So
I'm
gonna
go
through
and
kind
of
curate
one
and
we'll
see
what
it
looks
like.
So
we'll
get
something
for
people
who
kind
of
look
at
and
and
some
I
don't
know,
knowledge
start
to
start
to
iterate.
You
said.
B
A
C
C
C
Separate
breakout
sessions
a
lot
more
sessions,
so
there's
multiple
tracks
along
the
contributor
there's
multiple
rooms,
so
this
is
in
sort
of
the
there's
a
room,
the
main
main
track
kind
of
thing.
So
we
have
in
there
three
things
for
cigars.
We
have
a
25-minute
vision
session.
We
have
a
25-minute
conformant
session
and
I'm
gonna.
Do
a
prod
readiness
thing
for
25
minutes
and
I
can
show
you
that's
scheduled
if
we
want
to
want
to
see
what.
A
I
can
help
with
that
leading
up
to
the
contributor
something
I'm
gonna,
probably
offload
a
lot
of
my
current
work
items
to
just
focus
on
doing
slides,
cuz
I
have
many
things.
I
have
to
do
it
for
cuca,
so
I
probably
be
able
to
help
there
so
long,
as
doesn't
overlap
with
my
other
cyclists
or
motorcycle
responsibilities
in.
A
Next
up
was
an
accident
for
me
to
broaden
the
river
pools
and
start
to
do
the
public
admonishment,
slash
called
arms.
I
did
an
initial
call
as
Matt
and
other
people
we're
stepping
down,
and
other
people
were
trying
to
step
up
into
the
cigar
space,
but
I,
don't
necessarily
know.
If
there's
a
lot
of
people,
we're
actually
gonna
be
able
to
do
the
hard
work.
A
Not
explicit
admonishment,
but
like
a
little
bit
of
a
you,
know:
hey
we
haven't
been
doing
the
best
job
here
with
keeping
up
with
our
test
automation
apparatus
and
then
the
second
part
is
going
to
be.
This
is
really
important.
We
should
really
start
to
focus
our
some
of
our
effort
and
energy
there
and
what
I'm
going
to
try
to
also
do
is
figure
out.
What's
the
what's
the
stick
philosophy,
like
you
know,
it's
one
thing
to
have
make
public
statements.
It's
another
thing
to
have
some
type
of
policy
enforcement,
yeah.
C
So
I
have
a
few
comments
on
that,
one
in
general,
part
of
the
reason
for
getting
that
time
in
the
summit
was
France
where
there's
recruitment
thing,
but
we
can
also
I
think
talk
about
the
stick.
Another
thing
is
the
part
of
the
prod
readiness
is
a
part
of
what
will
come
out
of
that
is
more
tooling
around
enforcing
certain
criteria
be
met
when
you're
going
to
GA,
so
one
of
those
criteria
can
be
the
conformance
tests
and
so
more
of
a
you
know,
tooling,
that's
gonna,
say
hey.
C
A
So
that's
my
10-day
plan,
I
was
sort
of
I
was
gonna.
Try
to
run
this
around
cook
on
time.
In
fact,
I
originally
wrote
my
my
calls
my
talks
originally
about
the
realm
of
the
transmat
habits.
I
avoid
those
pitfalls,
so
I
got
plenty
of
fodder
there.
So
if
you're
interested
in
helping
on
that
effort,
please
let
me
know
so.
Tim.
C
Do
you
think
that
that
should
be
what
we
talk
about
on
a
summit
in
that,
so
we
have
two
new
times
for
conformance
right.
We've
got
one
in
the
summit
when
the
maintainer
is
track,
the
main
conference-
that's
really
our
face-to-face
thing.
So
during
that
contributors
summit
I
mean
that's
the
audience
we
want
to
reach
it.
A
B
That
was
something
that
we
noticed
as
I
think
it
was
Jeffrey
was
trying
to
promote
or
create
a
test,
and
he
noted
that
there
were
already
existing
tests
and
we
weren't
seeing
them
because
they
were
no
Danny
D
tests,
so
we
needed
to
copy
those
over
and
I
there
was.
There
is
a
current
promotion
in
place,
I'm,
not
sure
if
it's
so.
C
What
well
could
you
be
the
follow-up
on
that?
Maybe
the
maybe
did
not
even
aware
of
this-
could
welcome
a
testing
lab.
The
the
no
tests
are
split
into
two.
The
know.
Dentists
are
split
into
sort
of
two
categories:
one
are
the
ones
that
run
against
a
cluster
and
one
are
ones
that
don't
need
a
whole
cluster
and
just
need
a
node
and
and
there's
a
set
of
tests
that
are
shared
between
them.
B
A
B
That's
done
this
again
we're
going
through
our
action
items.
Lava
lamp
was
the
one
with
some
feedback
that
we
already
have
something
sort
of
going
and
I
want
exactly
how
we
operated
it.
Liggett
give
us
the
info
and
I
kind
of
set
it
down
for
now
not
been
high
priority,
but
the
action
wasn't
so
in
which
we
followed
up
everything
else.
At
the
end
of
the
action
items
that
I
captured
from
last
meeting.
B
Sure
the
first
thing
is
where
we
do
have
API
snoop
CN
CF.
That
IO
is
up
with
current
data
from
this
morning.
The
one
thing
that
made
me
sad
was
that
the
numbers
went
down.
Then
you
can
see
the
dates
over
the
past,
roughly
two-week
periods
other
than
this
last
ones
about
a
week
and
what
we
may
have
not
have
noticed
or
explained
from
last
time,
because
usually
I
find
there
increase.
B
There's
not
enough
time
for
me
to
figure
out
why
before
the
meeting,
but
we
had
promotion
of
admission,
webhook
and
customer
resource
definitions
was
the
largest
part
of
the
increase,
and
you
can
see
that
the
number
of
conformance
hits
is
about
the
thing.
The
conference
performance
hits
actually
goes
up
from
100
all
the
way
to
124.
B
But
one
thing
I
haven't
explained
for
this
week
is
we
have
a
drop
and
there's
no
changes
in
like
we
didn't
have
any
promotions
or
new
conformance
tests
in
the
last
seven
days
or
eight
days,
but
we
did
have
these
endpoints
that
are
no
longer
hit
and
there's
some
links
down
here
under
as
the
test
didn't
change.
That
made
clear,
maybe
something
the
framework
changed,
but
there's
the
two
links
one
from
a
week
ago
for
it
specifically
focusing
on
app
controller
revision
and
one
at
the
same
endpoint.
B
A
A
C
C
E
C
B
B
A
Think
with
that
list,
it's
all
list
operations
right.
So
it's
basically
it
looks
like
a
clean
up
portion
and
if
you
notice
that
they're
all
named
spaced
I'm
guessing
this
has
to
do
with
Clayton's
change
to
to
remove
a
section
portions
of
code
that
we
had
with
namespace
cleanup.
That
was
like
super
legacy.
Oh
listing.
A
Stuff
like
this
like,
if
we
wanted
to
add
just
basic
operations,
we
shouldn't
be
able
to
just
use
the
dynamic
client.
If
you
want
to
get
like
raw
larger
coverage,
you
could
have
a
set
of
operations
which
is
basically
just
listed,
watch
generic
set
of
tests
across
the
entire
state
space,
the
dynamic
client,
and
that
would
increase
your
coverage
significantly
and
hit
freaking
every
resource
time.
So
it
would
increase
the
the
current.
B
B
What's
going
to
get
us
to
the
point
of
automation
for
dating
and
PRS,
we
were
taking
30
minutes
on
a
bunch,
of
course,
to
use
regular
expressions
to
map
in
you
know
all
of
our
usually
three
hundred
thousand
or
more
audit
events
per
job
to
the
right
and
operation
ID,
and
we
went
through
and
look
at
some
code
pointed
to
us
by
lava
lamp
and
slid
well
on
how
the
API
server
itself
does
the
mapping
on
incoming
URLs
requests
and
yeah
that
got
us
to
ten
seconds.
So
that
was
a
major
increase
in
speed.
B
This
will
allow
us
to
look
into
at
least
commenting
and
do
those
initial
steps
on
automation
for
requesting
that
people
get
approval
from
the
right
folks
or
blocking
it
or
or
something
I,
don't
know
what
it.
What
exactly
one
do
with
it's
part
of
the
the
next
around
the
cap
for
a
non
required
prowl
job
later.
B
A
C
So
we
like,
they
think,
a
couple
of
little
tests
around
watching,
maybe
config
Maps
or
something
as
I
understand
it.
The
there
are
things
that
each
resource
has
to
do
in
order
to
properly
implement
all
of
this
stuff.
So
it
would
be
useful
to
have
a
suite
that
we
can
point
and
and
iterate
over
all
the
resources
and
validate.
B
B
That's
been
just
a
little
bit
of
time
what
we
do
with
the
dynamic
line
as
far
as
creating
an
object
of
like
for
each
of
these
underneath
these
operations,
that
will
be
a
particular
kind,
kubernetes
kind,
and
if
we're
talking
about
dynamically,
looking
through
the
API
server
for
the
available
kinds
and
then
instantiating
a
kind
and
then
creating
a
watch
on
it,
I
wasn't
really
clear
on
there
on
the
float
through.
For
that
I
mean
I.
A
Create
the
resource
successfully,
so
you
can
use
the
dynamic
client
to
do
this
by
by
basically
instantiating
a
single
object
and
because
every
resource
object
has
an
object
meta,
you
can
mutate
the
object
meta
without
knowing
the
full
details
of
the
kind.
So
using
the
dynamic
client,
you
can
basically
update
the
name
or
update
any
of
the
object,
meta
fields
of
the
resource
type.
So
this
allows
you
to
basically
walk
the
entire
kind
list
that
exists,
create
an
informer
on
that
kind,
create
one
with
an
object.
A
C
B
Okay,
thanks
for
that
clarification,
Timothy
will
hopefully
come
back
next
time
and
see
the
total
at
least
tested
parameters
pop
up
a
bit
if
we
can
get
a
test
in
and
if
it's
not
merged,
I'm
gonna
see
if
we
can't
start
generating
a
for
this
particular
PR,
what
the
increase
would
be.
So
we
can
see
where
to
prioritize
our
PR
margin.
B
You
know
without
having
to
look
at
the
api's
new
code
and
also
having
a
standard
place
where
this
is
documented,
where
the
current
set
of
conformance
like
a
friend
of
a
past
names
page
or
how
we're
documenting
it,
because
we
have
really
three
different
ways
that
we're
looking
at
one
is
the
watch
coverage
metrics,
which
just
should
be
simply
and
remove
soon.
We've
got
our
endpoint
metrics
and
then
we've
got
our
more
complex,
pods
back
covered
metrics
and
eventually
some
behavior
coverage
metrics.
B
But
due
to
this
speed
up
where
we
just
had
for
the
current
metrics
for
API
operations,
we
could
have
detected
when
DPR
from
Clayton's
namespace
changed,
dropped.
Our
current
coverage
metrics
and
at
least
had
a
conversation
and
note
at
that
time.
Oh,
the
namespace
cleanup
now
means
these
other
things
aren't
being
hit,
and
maybe
we
should
start
testing
them
in
a
different
way.
B
B
The
other
thing
that,
after
we'll
work
on
creating
the
stick
in
this
next
two
weeks,
print
for
us,
we've
in
general,
had
help
from
from
John
and
from
Timothy
when
we
ping
them
directly.
But
we're
still
like
he
said
the
the
number
of
people
that
actually
respond
to
our
request
for
helping
to
look
at
goats
is
low
and
we're,
and
and
for
the
things
where
we're
trying
to
promote
the
test
or
slightly
change
or
modify
an
existing
test.
The
amount
of
pushback
is
really
high,
and
so
we
keep
modifying
the
docs.
B
A
Instead
of
taking
so
getting
a
part
of
this
like
when
you
write
new
tests,
they're
owned
by
the
same
so
you
have
to
work
with
the
Satan
question
to
make
sure
that
they
have
being
within
there.
The
problem
with
this
exists-
that's
not
they're,
not
only
right.
So
if
there
are
enough
reviewers,
if
there
are
enough
people,
they
will
make
sure
they're
prioritize
it
accordingly
right
and
if
their
urgent
of
people
in
there
over
books
and
they
might
have
different
priorities,
they
they
won't
treat
this
as
a
high
priority.
A
A
With
regards
to
Limor
age,
one
I
already
commented
I
come
in
to
back
in
September.
Third,
like
right
away
so
long
as
I'm
assigned
I
will
I
will
eventually
get
to
it
within
a
week.
That's
my
standard
turnaround
time
because
I
have
certain
days
of
the
week
where
I
can
devote
towards
just
triaging
my
own
backlog,
but,
like
you
know,
the
comment
still
holds
and
I.
Think
part
of
the
problem
here
is
not
just
not
just
getting
feedback
but
soliciting
understanding
of
what
is
it?
What
is
the
problem
statement
that
you're
trying
to
accomplish
here's?
A
Why
it's
important
and
then
here
is
what
we're
going
to
do
about
it
and
does
this
make
sense
and
I
think
part
of
that
has
been
missing.
So,
if
I
look
here,
this
here
was
submitted
to
do
X,
but
the
problem
statement
is
missing
and
how
you
go
about
it,
and
why
is
also
missing
so
like
it's
very
concrete
to
me
when
doing
this
particular
view,
like
I,
see
what
you're
trying
to
do
I
disagree
with
this
approach.
So
if
we
agreed
upon
the
problem
statement
in
the
beginning,
it
probably
a
lot
less
pushback
right.
A
So
part
of
this
is
having
to
do
a
well-defined
distributive
pipeline.
An
open
source
project
is
one
agree
upon
the
problem.
Statements
make
sure
that
they're
concrete
and
actionable
that
when
you
go
about
executing
against
that
problem
statement,
that
is
clear
that
it's
in
the
there's
somewhat
agreement
on
the
how
right
not
not
to
the
nth
degree,
because
sometimes
people
can
have
created
we're
solving
it.
You
know
some
people
might
be
highly
opinionated
that
I'm
very
widely
on
the
sig
great.
But
you
know
this
individual
problem
state
was
was
ambiguous,
other
ones.
A
They
should
be
a
little
bit
less
ambiguous,
but
you
don't
actually
have
the
coverage
or
you
don't
have
the
person
to
review
it.
So
what
might
be
even
more
beneficial
before
we
start
to
do
this
is
to
have
like
here
is
the
contributor
guide
to
helping
you
make
your
test.
We
have
like
how
to
write
good
tests,
but
like
it's
really
a
people
problem,
it's
a
shared
state
space
and
understanding
and
a
people
problem
so
step.
One
write
an
issue
and
get
make
sure
you
get
buy-in
from
the
people
who
are
gonna,
be
the
reviewers.
A
Make
sure
that
the
problem
statement
for
the
issue
is
well
define,
step
2,
you
know,
make
sure
that
they
have
bandwidth
to
review,
step
3,
submit
the
PRS
and
get
on
the
backlog
and
step
four
is
going
to
be
hound
they're
better
for
worse,
the
the
project
itself
is
under-resourced
because-
and
it's
not
just
under
resources
in
general,
it's
under-resourced
globally.
So
we
we
have
some
other
people
to
have
their
own
agendas.
So
this
might
not
be
a
higher
priority
agenda
item
for
them.
A
The
question
we're
going
to
have
to
face
is:
how
do
we
make
it
a
higher
priority
agenda
item
for
them,
but
going
back
to
the
original
statement,
I
made
it
there's
multiple
parts
here,
I
think
one
of
the
key
fundamental
parts
is
having
a
well-defined
path
for
you
that
eliminates
the
friction
and
that
well-defined
path
is
right.
A
well-defined
issue
before
you
submit
the
PRS
and
make
sure
you
get
buy-in
from
a
singing
question
before
you
submit
the
PRS.
Otherwise
you
could
just
sit
there
for
a
long
time.
B
B
A
C
Can
poke
somebody
here
about
this
if
you
assign
this
to
me,
but
I
think
that
in
general,
like
maybe
as
part
of
what
we're
talking
about
the
email,
you're
talking
about
doing,
Tim
and
I,
think
we
want
to
see
if
we
can
convince
the
SIG's
to
designate
one
or
two
people
as
a
point
of
contact
for
conformance
test
promotion
and
development
from
from
each
sake,
I
think
that
would
be
even
if
they
don't
have
the
time
to
review
everything.
At
least
we
have
it
within
the
sig.
At
that
point,
it
might
be
a
helpful
process.
A
C
A
A
You
can
poke
me
to
try
them
route,
so,
if
I'm
assigned
what
will
happen
to
me,
is
it
eventually
show
up
on
my
my
grenade
aboard,
and
this
is
my
source
of
truth
right
I
do
things
in
priority
order
to
Juba
Nader?
If
you,
if
you
poke
me
on
slack
I,
will
not
necessarily
address
your
issue.
I
will
do
things
in
priority
order
approving
ater,
because
you
know
you're
not
like
if
a
person
folks
me
it's
not
like
I,
don't
have
enough
things
going
on
right.
A
C
A
A
D
D
Is
the
reason
why
has
kind
of
been
sitting
back
and
forth
is
a
I
see
your
comment
about
just
agreeing
with
breaking
out
of
the
function
into
multiple
it
locks
and
I?
Think
that
we
could
have
definitely
clarified
in
the
initial
PR
comment?
The
intention,
if
you
see
in
by
my
first
comment,
addressing
your
comment:
September
4
it
links
to
the
original
PR
for
promoting
this
test,
which
was
closed
because
Smith
XP
Aaron
felt
like
the
test
would
had
too
much
things
going
on
with
it,
and
he
was
what
you
suggest
to
this.
D
Replace
the
bodies
with
it's
essentially
we're
just
Clinton's
online.
So
my
point
would
bring
this
up.
Is
this
is
not
the
only
case
where
we've
had
you
know
somebody
saying
okay?
Well,
he
can't
promote
this
because
X
reason
and
it
this
way
then
try
to
promote
it
again.
We
try
to
make
a
change
and
someone
else
comes
in
says
well
paid.
A
This
particular
the
idea
of
adding
multiple,
it's
the
reason
why
it
doesn't
make
sense.
Okay,
let
me
at
least
illuminate
that
piece.
So
we
can
I
have
clarification
there.
When
you
have
described
limit
rage,
you're,
basically
saying
it.
The
limit
range
should
do
something
by
doing
something
else
right
and
the
problem
with
breaking
up
the
individual.
It's
is
that
your
ear,
you
have
a
conflating
state
space
of
of
problems
that
don't
apply
to
one
another.
It's
doing
something
to
try
and
get
in
to
test
this
in
a
behavioral,
bio
fashion.
A
A
Exactly
so
your
your
actual
test
is,
it
should
create
a
lineage
right.
So
that's
that's
the
reason
why
I'm
saying
it
doesn't
make
any
sense
to
do
it
to
break
apart
and
hits
right.
So
concretely,
this
specific
breaking
apart
doesn't
make
sense,
because
the
way
the
suite
is
structured
and
organized,
it
wouldn't
make
sense
to
the
reader
who's
trying
to
organize
the
set
of
tests
by
the
descriptions
or
the
it's
that
are
there.
So
that's
concretely
why
I
said
no!
This
is
not
the
way
we
should
do
it
now.
B
A
A
Think
that's
a
general
statement
that
could
probably
apply
to
a
bunch
of
this
and
when
he
says
it's
very
busy,
it
is
very
busy
like
the
test
is
too
long
in
my
book,
but
I
think
if
the
the
game
is
broken
up
into
well
factored
functions,
each
function,
doing
a
set
of
concrete
steps.
I
think
it'd
be
easier
to
read
and
it
would
make
a
lot
more
sense.
B
I'm
gonna
add
an
AI
here
for
these
general
descriptions
like
this.
What
we
learned
here,
because
this
is
a
I,
don't
think
this
is
capture
anywhere
and
it
is,
as
you
said,
generally
good
and
particularly
on
how
we
do
our
grouping
of
our
buys
within
the
describes
when
we
get
to
in
it.
What
that
should
look
like
it's
kind
of
described
in
BDD
in
general,
but
more
specifically,
making
sure
it's
there
in
the
guide.
A
A
A
E
Essentially,
it's
a
very
thing
that
I
want
to
discuss
here.
So
I
did
some
CSI
validation
test
weights
before
so
recently.
I
should
have
closed
them,
but
general
problem
is
that
storage
is
looking
for
a
way
to
identify
a
subset
of
the
tests
to
be
close
to
compliments
or
validation
sweetwater
when
we
call
it
and
I
am
I,
told
Michelle
I
was
talking
to
Michelle
file
last
week
and
I
told
her
that
I
would
like
to
help,
and
the
thing
is
I
am
looking
for
a
direction
here.
E
Essentially,
should
I
forget
it
as,
like.
You
know,
API
coverage
for
CSI.
They
have
created
these
test
feeds
currently,
which
are
like
you
know,
kind
of
like
table
desk.
So
that's
not
going
to
work.
If
you
want
to
basically
split
the
test
center
should
I
go
with
coverage
wise
or
since
there
is
a
CSI
spec.
Should
we
go
with
the
BDD
general
discussion
and
direction
that
I
need?
E
C
Is
kind
of
a
general
class
of
problems
that
we
made
little
to
no
progress
in
in
the
last
year
of
how
to
Hana
Hana
leaves
kind
of
pluggable
service
areas
from
a
point
of
view
of
performance.
I
do
have
it
on
my
list
of
things.
I
want
to
get
some
agreement
from
this
group
in
the
next
cycle
of
the
next
quarter
or
so
on
how
we
proceed
with
those
things,
but.
B
A
Example.
Here
is
the
lineage
test,
the
storage
class.
The
storage
tests
are
currently
classified
with
described.
You
know
for
all
of
them
right,
so
you
have
the
whole
regiment
moniker
you
can
hit
for
that
whole
suite.
Now
you
can
modify
the
describe
or
you
can
have
a
set
of
tests
underneath
the
whole
umbrella
of
storage.
That
would
be
like
some
set
of
features
or
some
level
of
verification,
/
validation,
to
use
some
tag.
A
E
C
Probably
move
and
promote
conformance
but
I
think
that
I
guess
maybe
it
may
be
I'm
taking
maybe
I'm
making
it
more
difficult.
I
was
getting
back
to
the
discussion
of
profiles
and
how
we
can
classify
a
set
of
functionality
and
say
you
know
this
said
we
want
a
closer
built
by
some
provider
to
be
able
to
meet
this
set
of
functionality.
C
E
A
I,
take
a
step
back
before
your
bike.
Shed
too
deeply
here
is
that
you
know
something
of
that
nature
should
right
should
have
a
cap
written
or
write
like
I've,
been
writing
a
bunch
of
caps
lately.
This
is
clearly
something
where
having
explicit
set
of
goals
and
non
goals
would
help
to
frame
the
conversation
so
like
what
is
what
exactly
are
you
trying
to
accomplish
with
what
team
and
when?
How
was
the
motivation
where
the
goals
weren't
on
goals,
and
then
we
can
start
to
talk
about
the
logistics
of
that?
A
If
you
frame
the
conversation
first
by
concretely
saying
what
is
the
motivation,
what
is
a
time
frame?
What
are
the
goals
or
the
non
goals?
That'll
help
us
to
search
steer
the
conversation
versus
this
is
pretty
open-ended
right.
We
could.
We
could
talk
about
this
at
length
about
budget
and
multiple
entities
and
virtue
in
this
problem.
C
E
A
I
think
just
defining
what
you're
trying
to
do
and
writing
a
well-defined
proposal
on
it
I
think,
would
be
generally
useful
and
caps
are
listening,
written
for
things
that
cut
across
things
or
to
get
people
on
the
same
page,
so
you're
gonna,
cut
across
storage
and
potentially
conformant.
So
that
makes
sense.
C
B
C
The
general
class
a
problem,
how
we're
going
to
manage
profiles
or
optional
sets
of
not
even
optional
but
pluggable
sets
of
functionality
within
the
conformance
framework,
but
I'm
not
sure
that
fits
three
is
actually
asking
for
here.
So
I
don't
know
if
there's
a
different
cap
for
what
3d
wants
here
or
what
do
you
think
sorry.
E
C
I
mean
I
thought
I
did
Tim.
Suggestion
is
a
good
one
that
if
we
have
a
camp
like
we,
what
we
had
before
for
validation,
Suites,
quit
and
just
closed.
It
was
just
a
PR
and
some
discussion,
and
but
it
was
a
PR
specifically
modifying
a
particular
document
set
of
policies
and
guidance,
and
maybe
we
need
a
little
bit
broader
approach
and
a
little
bit
more
organizing.
B
A
Think
I
don't
think
a
discussion
is
I
somebody
needs
to
take
point
on
a
design
I'm
a
cat-like
tree
and
whether
that
be
you
or
someone
else,
Inc
storage,
that's
helping
you
work
on
these
tests
that
wants
them.
I.
Think
that's
the
appropriate
action
path
here,
because
that
we
have
a
lot
of
conversations
and
stuff
that's
kind
of
problem.
We
need
to
have
more
written
culture
versus
oral
culture.
E
So,
let's
turn
this
turkey
I
mean
from
the
storage
point
of
view.
My
problem
is
basically
identifying
these
tests,
which
is
generally
not
right
from
the
conformance
point
of
view.
I
need
more
better
agreement
on
whether
this
should
this
general
test.
We
should
belong
to
conformance
or
a
validation
suite
of
whatever
we
call
it.
So
those
are
two
different
goals
in
there.
So
I
can.
I
trait
I
can
start
off
with
the
document
I
trade
with
you
and
John,
and
see
where
we
end
up
on
the
storage
box.
Yeah.
C
I
mean
I,
don't
Michelle
and
I
have
talked
about
this
in
the
past.
Like
four
months
right
I
mean
we
haven't
talked
to
a
couple
months
but
like
I,
think
almost
at
least
six
months
ago
we
talked
about
the
idea
of
doing
this
kind
of
thing,
but
you
know
it's
never
resulted
in
concrete
in
a
concrete
document
with
its
describing
what
exactly
we're
going
to
do,
and
therefore
concrete
action
that
comes
out
of
that
document.
So.