►
From YouTube: RATS Architecture Design Team, 2021-02-02
Description
RATS Architecture Design Team, 2021-02-02
A
B
Well
yeah,
so
I
mean
some
of
the
virtual
interims
it
was.
We
had
the
same
kind
of
you
know
experience
where
that
was
the
problem
with
the
next
slide
message
right,
someone
says
next
slide
and
the
chair
does
it
instantly.
B
Taurus
was
really
having
difficulty
with
that
and
it
would
take
a
minute
for
it
to
get
out
to
people
right.
So
that
was
really
weird
like
it
should
just
be
a
single
screenshot,
but
obviously
it
doesn't
do
slide
sharing
with
you
know
it
does
it
as
video
sources.
B
But
it's
silly
because
right
I
mean
most
of
most
of
the
slides.
They
don't
change.
If
I
move
my
mouse
around
here
yeah,
you
want
to
see
that,
but
most
of
the
time
it's
not
important
to
see
that
there
okay.
So
we
seem
to
all
be
trickling
in
here
and
eight
of
us
today.
B
Dave
thank
you
dave
and
hank
for
the
pull
requests.
I'm
meant
to
spend
most
of
sunday
doing
things
and
I
was
unable
to,
but
what
I'm
hoping
actually
is
that
we
will
get
to
a
point
at
the
end
of
friday's
meeting.
That
will
say:
okay,
we
we've
done
as
much
as
we
can
and
we
should
go
ahead.
B
One
reason
I
think
we
should
go
ahead
is
that
I
think
that
you
know
we're
going
to
have
a
bunch
of
of
of
of
changes
that
reviewers,
like
area
directors,
are
going
to
ask
for
us
anyway,
and
it's
almost
dumb
to
make
it
perfect
for
the
simple
reason
that
we'll
have
to
explain
it
we'll
have
to
do
it
all
over
again
in
some
ways.
So
I
I
think
that
we're
wait.
B
I
think
I
would
rather
us
finish
earlier
and
then
come
back
then
try
to
get
it
to
the
point
of
being
perfect.
I
don't
know
if
everyone
else
agrees
with
that
plan.
A
I
think
I
think
what
you
can
do
here
is
you
could
do
that
thing
and
also
not
do
not
the
other
thing.
So
so
what
we
could
do
is
we
could
submit
here
and
say:
yeah,
that's
that's
what
we
will
do
for
the
w,
glc
and
and
still
like,
I
don't
know
not
really
in
secret,
but
in
parallel
iron
out
some
other
nits.
I
mean.
B
You
can
do
that.
What
I'm!
What
I'm
trying
to
say
is
that
that
that
my
experience
with
the
the
isg
in
the
last
two
years,
particularly
our
security
ads,
is
that
they're
extremely
detailed
oriented
and
they
spend
my
opinion
far
too
long
on
working
group
draft
reviews.
B
So
what
I'm
saying
is
that
we
can
fix
a
bunch
of
nits
and
then
have
a
complaint
that
we
that
they,
like
the
original
text
better
without
having
ever
seen
the
original
text
right,
yeah.
A
B
Ought
to
be,
we
ought
to
be,
and
frankly,
the
stuff
that
we're
arguing
about
is
very
nuanced,
and
it
really
doesn't-
and
you
know
probably
matter
on
the
great
deal
of
things
except
that
you
know
we
got
feedback
from
people
who
were
working
with
implementers.
Who
said
that
they
didn't
understand
this.
And
so
we
want
to
clarify
that.
So
I
think
that's
valuable.
B
But
what
I'm
trying
to
say
is
that
we're
going
to
get
a
similar
level
of
detail
from
our
our
area,
director,
review
and-
and
I
I
believe,
it's
roman
as
our
area
director
and
so
we're
going
to
get.
B
B
A
Are
extensive
and
I
still
think
that
that
it's
a
surprising
level
of
detail
from
isp
in
general.
So
in
any
case.
B
What
I'm
trying
to
say
is
that
the
sooner
we
get
it
in
front
of
him
right,
the
sooner
he
discovers
some
issues
that
maybe
we
already
have
and
and
then
we
also
may
get
pushback
on
things
where
we
went
in
a
particular
direction
and
he
doesn't
like
it
and
we
may
have
to
we
it's
better,
for,
I
think
us
to
re-explore
that
text.
While
it's
it's
fresher
in
our
memories
than
later
on,
when
we,
you
know,
we've
settled
it
for
two
months
already.
B
So
that's
why
I
just
really
want
to
get
it
out
of
out
of
this
state
here
already
we've
been
in
this
working
group
when
this
particular
set
of
review
stuff
since
beginning
of
december.
Basically,
so
that's
a
long
time
two
months,
I
think.
C
There's
a
number
of
those
issues
that
we
just
need
to
respond
via
email
to,
and
I
appreciate
william,
did
a
pass
and
mark
dale
six
or
something
of
them,
as
as
he
thinks
that
they
won't
fix,
and
I
kind
of
plus
one
that
and
added
them
to
the
list
of
things.
I
think
we
should
talk
about
but
yeah.
I
I
think,
there's
a
bunch
of
these
that,
rather
than
figuring
out
how
to
fix
the
text,
we
should
figure
out
how
to
respond
via
email.
So
sorry,
who's
in.
C
There's
the
help
wanted
label
that
I
added
and
I
didn't
go
through
all
of
them,
but
so
some
of
them
are
from
me.
Some
of
those
are
from
william
said.
I
think
we
should
won't
fix
this
yeah,
so
you're
on
the
call
here
right
and
so
I
went
through
you
can
see,
there's
a
bunch
of
things
that
are
help
wanted,
and
so
there's
a
couple
of
those
that
are
ones
that
william
had
originally
added
fixed,
then
removed
it
and
said.
C
I
think
we
should
won't
fix
this
and
I
said
well,
I
think
we
should
won't
fix
it
as
well.
So
I
think
he's
got
a
good
argument
so
anyway,
so
I
appreciate
going
through
there
and
I
think,
there's
probably
more.
I
didn't
go
through
all
of
them.
I
just
looked
at
some
of
those,
so
okay.
C
B
Okay,
so
I'm
going
to
interrupt
this
about
10-2
and
I'm
going
to
bring
us
back
to
the
won't
fix
this
and
make
sure
we're
all
agreeing
with
that.
Then.
B
Okay,
let's
should
we
start
with
this
one,
because
I
have
the
tab
open?
Okay,
I
don't
know
maybe.
A
Oh
yeah
dave
had
some
minor
changes
to
examples
being
pulled
in
and
just
a
highlight
idea.
I
I'm
I
was.
A
C
So
my
question
because,
like
I
said
I
can
live
with
the
original,
I
just
like
fewer
words
is
better
if
it
has
the
same
meaning
and
so
right
now
when
he
says
unprotected
eg
not
signed,
so
in
other
words,
for
example,
not
sign
are
there
other
meanings,
because
I
don't
want
to
change
the
meaning.
If
there's
other
meanings
of
unprotected
besides
not
signed,
it
doesn't
say,
I.e,
not
signed,
it
says
eg
as
as
signing
is
just
an
example
is
there.
E
A
B
A
Yeah,
so
the
aircap
scenario
is
a
typical
thing.
We
sometimes
come
back
to
even
here
and
I
just
wanted
to
not
be
inclusive.
That's
all.
I
guess.
A
C
So
I
can
live
with
the
original.
It
just
seems
a
little
verbose
if
there's
a
way
to
shorten
it.
So
that
was
a
suggestion.
If
you
don't,
like
the
suggestion,
feel
free
to
reject
so.
C
So,
michael
or
william,
I
don't
know
I
guess
ned's
in
the
process
of
joining.
He
doesn't
have
audio
yet.
D
B
Oh,
I
have,
I
have
one
of
in
favor
dave
and
one
opposed
ned
and
hank
has
doesn't
care,
and
I
think
I
prefer
I
think
I
actually
prefer
unprotected,
but
I
don't
actually
feel
that
strongly.
C
A
I
I
I
have
as
my
tiny
comment
that
will
come
back,
but
not
maybe
with
this
one
due
to
the
move
of
the
terminology
section.
Some
of
the
terms
introduced
by
that
are
maybe
capitalized
as
a
before
they
are
introduced
in
section
four
yeah.
So.
A
B
C
B
C
A
Sorry,
sorry
to
be
a
nagging
nancy
here,
but
the
version
one
1.0
that
is
hard
coded-
might
not
relate
to
the
target
anymore
if
it
changes
with
latest.
Is
that
correct?
Because
that
is
not
a
stable.
C
A
C
A
C
All
right
so
here
the
original
thing
filed
was:
should
we
define
mutual
mutual
attestation
and
last
time
we
said?
Well,
we
don't
define
mutual
authentication
either,
and
so,
rather
than
that,
I
tried
to
change
mutual
to
in
both
directions,
to
kind
of
define
it
and
layman's
speak
in
context
rather
than
defining
a
term
for
it.
So
hopefully
people
think
that
the
meaning
is
the
same,
because
it
doesn't
mean
that
you're
doing
attestation
at
the
same
time
right
it's
not
like
there's
one
exchange
that
does
attestation
mutually
and
you
at
the
end
of
it.
C
A
They
are
equivalent.
I
think
I
am
surprised
by
having
mutual
associate
with
at
the
same
time.
I
don't
think
that
mutual
implies
that
also.
A
Yeah,
okay,
it's
yeah!
I
see
and
there's.
C
A
At
the
point
of
execution
or
operationalization
you
you,
you
have
mutual
authentication
when
that
happened
is
not
of
interest
to
you.
It
has
to
be
valid
and
fresh.
So,
but
that's
going
to.
C
A
I
see
a
point,
but
the
only
thing
that
that's
actually
of
concern
is
my
question:
is:
would
people
expect
the
term
mutual
here
or
not?
If
that
is
okay
with
both
directions?
I
will
be
fine
with
this,
so
I
because
mute,
the
authentication
mutual
is
like
a
thing.
You
people
you,
they
really
recognize
like
an
intuitive
thing.
C
C
So
you
see
the
words
around
9
15
I
expanded
to
have
the
text
be
the
same
as
in
the
previous
section,
so
you
can
see
previously.
You
said,
as
discussed
in
rp
owner
trust.
Similarly,
and
so
the
lines
923
sorry
921
to
924
are
basically
copied
from
the
previous
section
because
they
apply
equally
here.
So
I
expanded
the
sentence
just
so
that
they
don't
have
to
scroll
back
and
forth.
C
To
get
the
same
point,
so
that's
the
you
know,
in
which
case
typically,
one
side's
identity
or
evidence
must
be
considered
safe
to
share
with
an
untrusted
entity
that
that's
the
phrase
that
would
appear
in
the
section,
the
earlier
section,
the
relying
party
section-
and
I
copied
it
down
to
here.
If
that's
okay
with
people.
C
C
And
then
the
paragraph
that
I
added
in
the
privacy
consideration
section
which
you
can
see
is
ford,
pointed
to
you
from
both
of
those
privacy
considerations,
is
the
one
that
in
it
already
talks
about
pii
and
the
collapsed
part
that
you
don't
see
on
the
screen
right
now.
I
think
it's
in
the
top
collapsed
part
around
1163,
somewhere
up
there,
yeah
pi
1173
on
the
right
you
can
see
is
where
it
starts
talking
about
pii
appearing
in
evidence
and
so
on.
So
that
term
is
already
used
in
the
same
section.
C
So
at
the
bottom
right
after
the
discussion
of
the
the
same
notion
of
attestation
in
each
direction,
you
can
see
the
expansion
in
1194
to
1197
that
talks
about
you
know,
first
and
and
so
on
as
an
example,
and
then
this
is
my
proposed
text
feel
free
to
wordsmith.
That's
the
reference
of
the
ccc
deep
dive
for
the
anonymous
attestation,
and
this
was
my
attempt,
if
you
got
a
better
phrasing,
feel
free.
But
this
is
my
attempt
to
do
it
in
maybe
one
sentence
plus
your
reference,
but.
A
A
A
B
A
Okay,
when
I
do
the
generic
editorial
pass
to
switch
for
example,
then
the
change
above
pii
and
the
actual
term
with
the
actual
term
and
then
pii,
because
you're
introducing
pri
somehow
and
then
explain
it
with
what
it
means.
I
think
it
should
be
the
other
way
wrong,
but
it's
just
editorial
you're
fine
with
that.
We
don't
have
to
cover
this
here.
C
It's
already
expanded
in
the
same
section
before
this
use,
so
I'm
not
sure
understand
your
question.
So
probably
yes,
but
I
understand
what
you're
proposing,
because
it's
defined
it
is
divine
and
spelled
out
before
this,
at
like
three
periods.
A
C
C
C
Because
it's
not
in
your
the
revised
text,
it's
in.
A
This
tiny
knits
so
next
item
we
can
complete
one.
B
All
right,
so
are
we
done
with
this
part,
then.
A
C
C
Okay,
so
hank
under
your
definition,
entity
is
like
an
implementation
or
something
that
fulfills
a
role
in
the
role
is
the
participant
in
the
architecture.
G
F
A
Typically,
I
find
this
is
a
problematic
term
because
of
the
eat
and
and
the
ea
token
basically
uses
reuse
entity
in
a
different
way.
But
yeah.
C
C
Of
entity
is
ubiquitously
accepted,
but
if
we
just.
A
Sorry
that
was
okay,
then
I
have
to
lose
a
few
words
here.
So
entity
is
fine.
That's
right.
Roles
is
fine.
Yeah.
D
A
Problem
so
I
would
like
entity
is:
is
too
broad
it's
a
problem?
The
same
problem
is
with
artifact.
I
understand
why
we
use
artifact,
because.
B
A
Claims
and
conceptual
messages
and
and
some
other
messages
that
are
not
even
conceptual
messages
of
the
architecture.
So
that's
why
artifact
is
there,
but.
A
Roads
aggregate
on
things
that
are
less
than
entities
and
typically
they
would
be
referred
to
as
actors,
but
I
think
we
dismissed
that
at
some
point.
So
I'm
okay
as
a
compromise,
but
still
I'm
a
little
bit
like
really.
Why
entities?
Why
don't
we
define
the
bucket
where
we
put
things
in
and
so
that
that's,
but
that's
the
last
time
I've
erased
this
okay.
C
B
So
this
is
just
fixing
so
that
it's
formatting.
C
Okay,
so
here
I
need
you
to
do
delete
typically
a
device
right,
because
that's
not
the
point
of
that
issue
and
it's
also
very
specific
to
certain
use
cases.
It's
true
for
certain
use
cases
and
false
for
other
use
cases,
and
since
this
is
the
definition
section,
not
the
use
case
section,
it
doesn't
belong
here.
It
might
be
in
the
use
case
section
for
a
particular
use
case.
I
might
say
that
where
we
talk
about
who's
for
black
party-
but
I
believe
it
doesn't
belong
here,
so
I
deleted
that
from
two
places.
In
my
suggestion,.
A
A
C
A
C
Okay,
we
can
go
back
and
get
blame
log
and
figure
it
out
and
not
have
to
repeat
it.
But
I'm
saying
since
that
wasn't
the
point
of
235
that
this
is
meant
to
fix.
I
don't
think
this
per
request
should
be
doing
that
because
there's
some
at
least
in
my
opinion,
there's
a
little
bit
of
a
dispute
about
whether
this
is
actually
typical.
A
Yeah,
I
I
I
skip
a
scope
creep
with
intent
here,
you're
correct
the
the
issue
at
hand.
It
does
not
require
this
fixed.
C
Yeah
just
imagine
a
case
where
you
have
two
vms
on
the
same
machine,
one
attesting
to
the
other
right.
The
relying
party
is
not
a
device,
it's
another
vm
on
the
same
on
the
same
device
right,
so
things
like
that.
I'm
saying
it
varies
my
use
case
just
to
give
you
an
example.
So.
A
Michael
you're
capturing
that
for
removal
or
just
delete
the
as,
except
so,
the.
C
And
then
I
think
I
added
a
period
for
consistency
in
somewhere
near
the
end
right
there,
all
the
other
ones
had
periods,
but
that
one
didn't
that
wasn't
something
you
changed.
You
can
see
that
was
white
text
before
I
just
noticed
it
scrolling
past.
Since
you
were
doing
a
consistency
pass,
I
figured
I'd
help
you.
A
A
A
So
thomas
does
that
in
the
end
satisfy
why
you
actually
raised
this
issue.
C
Much
hank
I
can
see
it
gets
a
little
bit
more
wordy,
but
I
understand
hank's
point
so,
instead
of
the
ones
that
are
humans
right,
we're
used
to
say,
like
a
blind
party
owner,
you
know
an
engineer
that
does
whatever
and
so
often
the
owner
is
referring
to.
You
know
an
admin
or
somebody
acting
as
an
admin
and
hank's
text
changed
it
to
you
know
an
entity
acting
in
that
role,
and
I
think
that's
that's
fine.
C
C
And
yep
and
that
one
did
have
fixes
outside
the
title.
I
don't
think
it
recognizes
fixes
in
the
title,
but
it's
in
the
description.
So
it's
good
it's
good.
It
does
have
it
in
there
and
I
think,
the
la
that
one
hasn't
been
updated
since
last
time.
Right.
C
B
All
right,
so,
let's
walk
through
the
ones.
That
said,
won't
fix
not
right.
C
Or
at
least
help
wanted,
because
I
think
one
or
two
of
them
won't
fix
without
that
label
are
ones
we'd
already
talked
about,
and
so
that
since
yeah
now
this
one
is
not
necessarily
a
won't
fix
this
one
I
think
hank
and
I
had
exchanged,
and
you
can
see
I've
assigned
this
one
to
me
and
since
my
queue
was
basically
drained,
and
so
I
think
hank
and
I
are
in
agreement-
and
I
agree
with
hank's
suggestion,
and
so
let's
just
talk
about
that,
and
if
so
then
I
can
have
a
request
for
friday.
C
So
can
you
see
his
question
is
about
how
it
relates
to
tpms,
and
my
response
is
that
a
tpm
by
itself
is
not
the
testing
environment.
The
tpm,
plus,
whatever
is
feeding
at
the
measurements
together
from
the
environment
and
hank's
point
is
well,
let's
phrase
it
the
positive
sense
instead
of
the
negative
sentence
and
say
what
is
the
testing
environment?
Does
that
sound
like
a
feasible
approach-
and
I
said
yeah
absolutely
I
like
that
idea.
Let's
do
it
just
suggested
there?
C
A
C
Looks
okay,
then,
I
think
hank,
which
happened
while
I
was
sleeping
that
approach
looks
fine
and
if
everybody's
okay,
with
that,
we
can
remove
helpful
for
help
wanted,
leave
it
assigned
to
me
and
I'll
come
back
for
friday.
Wonderful.
B
C
C
C
B
C
I
was
looking
at
taking
this
one,
since
nobody
else
had
looked
at.
It
then
said
I
can't
think
of
any
other
way
to
do
this,
so
I'm.
C
D
D
C
Okay,
so
look
at
the
two
paragraphs:
okay,
so
he's
reading
that
part,
he
says
in
a
tester
consists
of
at
least
one
a
testing
environment,
at
least
one
target
environment
right,
and
I
think
his
gut
reaction
is
well.
You
haven't
kind
of
defined
those
yet
because
you
kind
of
keep
talking
about
that
and
some
of
them
they
might
be
combined
and
so
on.
The
actual
definitions
are
in
the
next
paragraph.
Right
claims
are
collected
from
target
environments,
you
know,
target
and
testing
environments
collect
the
values
and
so
on.
So
that's
really.
C
The
definition
is
there,
but
you
can
see
it's
like
the
next
paragraph
and
I
thought
about.
Is
there
some
way
to
like
swap
the
word
of
the
paragraph
I
said:
well,
no
that
wouldn't
make
sense
to
talk
about
what
they
are
without
saying
how
they
relate
to
an
attester.
So
I
couldn't
figure
out
a
way
to
improve
this,
because
the
definitions
are
there,
but
their
definitions
are
the
paragraph
after
use,
and
so
I
don't
know
how
to
improve
this
text.
B
C
B
C
D
D
B
C
C
D
F
Okay,
I
didn't
fix
it
yeah
do
we
want
to
mark
it
as
fixes.
C
Yeah,
so
these
are
the
two
that
william
looked
at
and
then
I
looked
at
it
and
agree.
So
I
have
my
comment
at
the
bottom
of
here,
which
is.
C
So
if
you
remember
right
in
our
list
of
use
cases,
we
have
a
confidential
ml
model
one
and
then
we
have
a
confidential
data,
one
that
is
immediately
after
that
and
the,
as
I
mentioned
in
my
comment
bottom
here,
the
confidential
data
one
is
immediately
after
that,
and
it
explicitly
says
in
the
beginning.
This
is
a
generalization
of
the
confidential
ml
model,
which
could
be
any
type
of
data.
It's
kind
of
how
that
interest
sentence
begins,
and
so
he
says,
isn't
it
kind
of
a
duplicate.
It
says
well,
yeah.
C
We
point
that
out
ourselves
and
you
can
see
in
my
comment
I
said,
but
I
think
it's
still
useful
to
have
both
of
them
for
two
reasons:
one,
because
the
ml
model
is
a
very
popular
one,
and
so
some
people
might
be
looking
for
that.
So
they'll
get
a
hit
on
that
keyword.
C
That's
one
but
number
two,
because
although
the
line
between
code
and
data
is
fuzzy,
the
confidential
ml
model,
one
is
closer
to
a
confidential
code
distribution
and
the
confidential
data
is
more
like
confidential
data,
which
many
people
think
of
as
different.
Even
though,
like
I
said
often,
the
line
between
data
versus
code
is
really
fuzzy.
B
Well,
people
are
also
very
much
attuned
to
the
concept
that
they
need
to
obfuscate
their
code
to
keep
it
confidential
yeah,
but
they
think
that
they
should
encrypt
their
data
right
right
and-
and
so
they
come
to
the
belief
that
there's
different
solutions-
and
I
think
that
that's
okay-
that
that
they
think
of
them
differently,
but
that
they
really
are.
They
converge
to
the
same
solution,
but
they
start
with
a
different
problem.
Maybe
a
different
risk
assessment,
even.
C
So
both
of
us
think
that
this
is
worth
an
email
response,
but
no
change
to
the
doc,
and
it
sounds
like
you
have
a
similar
but
marginally
different
point,
michael.
If
you
can
capture
it
in
here
just
so,
when
we
compose
the
email,
we
can
make
sure
that
your
point
is
not
forgotten
to
how
about
you
know:
encryption
versus
obfuscation,
but,
like
I
said
I
don't,
I
agree
with
williams.
This
is
not
warrant
to
change
the
doc.
B
C
D
B
B
C
That's
as
far
as
I
got,
I
did
not
go
through
all
of
them,
looking
for
other
things
to
add
help
one
or
two,
but
since
I
saw
everything
else
had
somebody
assigned
to
it,
I
I
did
not
have
this
hank's
brand
new
one,
so
we
could
look
at
things
for
our
new
ones,
see
if
we
have
any
advice
how
to
address
those.
C
Yeah,
okay,
so
here
this
is
what
hank
was
just
alluding
to
my
question:
should
we
change
the
word
so
artifact
just
comes
to
the
section
heading:
it's
not
actually
used
in
the
text.
It's
just
the
name
of
the
section
heading
right.
C
C
C
Yeah,
I'm
just
trying
to
minimize
minimize
new
terms,
minimize
number
of
new
terms
and
you're,
pointing
out
in
terms
of
artifacts,
is,
is
kind
of
spawned
into
existence
and
never
use
here
as
a
second
name,
and
never
again
is
what
you
said:
you're
right,
let's
see
if
we
can
get
rid
of
term
artifacts.
If
there's
some
term
we
already
used,
it
could
be
the
same
thing.
So
data
is
the
closest
thing
I
could
come
up
with.
A
There's
no
good
solution
for
this.
If
you
wouldn't
want
to
invent
terms,
so
that's
alright,
so.
C
C
B
I'm
okay,
with
this
being
artifacts,
we
we,
we
implicitly
use
the
word
artifacts
whenever
we
use
any
of
these
things.
These
are
all
artifacts.
We
just
didn't
need
to
to
pollute
the
document
by
repeating
it
over
and
over
and
over
again.
That's
all,
but
I
actually
really
like
the
word
artifacts,
as
particularly
for
something
that
is
a
signed
object.
A
You
can't
refer
to
it
or
better
that
we
do
not
refer
to
artifacts
at
all.
So
why
use
that
that?
But
we
don't
have
a
catch-all
name
and
we
use
to
use
conveyance
and
conveyed
data
is
more
to
the
essence,
but
I
also
acknowledge
michael's
input
that
yeah,
but
that
basically
is
okay.
Artifact
is
the
catch
all
here.
C
C
So
I'd
say
good
data
point,
but
I
don't
think
it's
usable
here,
yeah,
I'm
fine
with
won't
fixing
this
and
I
don't
feel
strongly
either
way
I
I'd
upset.
I
don't
want
to
spend
much
time
on
this
one,
since
I
think
it's
really
needy.
A
It's
it's,
it's
the
credit,
isn't
it
so
this
is
just
in
it,
but
for
the
sake
of
consistency
at
least.
A
B
A
Go
ahead,
so
that's
still
there
so
so
I
this
is
the
last
time
I've
erased
this
and
never
rest
forever
and
hold
my
peace,
but
entity
is
an
awful
term.
Sorry,
it's
just
everything
and
I
think
for
roles.
Actress
is
like
the
very
intuitive
counterpart
that
we
are
dancing
around
here.
So
why
are
we
avoiding
that?
A
Yeah,
so
we
were
forced
to
relinquish
participants
as
a
computing
science
term,
which
is
okay,
although
other
rc's
most
definitely
use
it,
and
then
we
were
like
okay,
we
bought
it
back.
We
simplified
roles
and
actors,
that's
an
intuitive
pair,
so
roles
act
great
on
actors,
but
somehow
that
was
also
not
okay
and
then
we
were
dancing
around
again
using
entity,
never
defining
it
being
very
broad,
and
so
this
is
the
last
time
we
bring
this
up
back
back
up.
A
So
why
not
use
actor
as
the
bucket
to
aggregate
roles
in
and
say,
and
this
actor
takes
on
these
roles
and
that
actor
takes
on
these
roles
and
then
they
interact
with
conceptual
messages
and
protocols
and
architecture
so
that
that
is
relatively
simple,
I
think.
But
if
it's,
if
dave
just
stayed
highlighted,
maybe
it's
predating
some
of
the
attendees
attendance.
A
C
A
Do
you
need
an
excerpt
so
like
like
about
a
portion
of
the
text
because
going
through
all
the
id
text
and
and
then
making
it
as
an
example,
might
be
a
little
bit
extensive
for
being
rejected?
So.
C
Because
energy
isn't
something
that's
in
that
the
protocol
of
the
architecture
actually
uses,
it's
not
a
term
that
we
have
to
define
to
be
understandable
right.
So
I
think
everything
else
in
the
glossary
section
is
something
that's
inherently
sort
of
special
about
rats.
There's
nothing
special
about
the
word
entity.
So.
A
C
C
The
same
thing
I
mean
in
theory,
your
point
is
valid,
but
it's
like
I
don't
know
I
I
don't
know
until
I
say
it
so.
C
Friday
yeah
this
was
this
one.
I
did
not.
I
saw
this
briefly,
but
I'm
gonna
actually
open
up
the
document
to
look
for
an
example
because
you're
saying
when
we
did
that
reordering
of
putting
things
before
other
things
relatively
recently,
you're
saying
that
may
have
messed
up
four
references
or
something-
and
I
wanted
to
see
a
specific
example
in
the
document
where
that's
true.
B
C
No,
I
thought
when
we
talked
about
it
before
this
is
from
memory,
so
I
could
be
wrong,
but
I
thought
in
section
one
those
terms
were
used
in
context
and
yeah
there
we
go
first
paragraph
one
period:
the
tester
produces
no,
the
previous
paragraph
that
one
that
one
right
there
introduces
in
text
the
meanings
of
a
tester
underlying
party.
So
those
two
terms,
and
only
those
two
terms
become
usable
for
the
glossary.
C
B
C
B
I
don't
have
a
problem
with
that
and
I
and
I
prefer
it
because
I
I
don't
mind
if,
if,
if
readers
don't
quite
understand
the
formal
definitions
of
these
words,
while
they're
reading
the
use
cases,
but
then
when
they
read
the
use
cases,
suddenly
sorry
the
terminology
they're
like
okay,
this
is
tightening
up
those
things
and
I
could
page
back
and
put
it
in
context.
B
But
actually
I've
got
a
much
better
idea
of
what
the
context
is
now
and
I
find
documents
that
are
kind
of
medical
in
the
way
that
they,
you
know,
they
define
all
these
terminologies
up
front
and
I
have
no
idea
which
ones
are
important
terms
or
not
or
how
they're
going
to
be
used.
So
I
can't
even
judge
what
the
glossary
means
right.
I
wind
up
having
to
read.
I
skipped
the
glossary.
B
I
have
to
read
forward
anyway
and
then,
when
I
get
to
a
term
that
I'm
uncertain
about
it,
I
have
to
go
back
to
the
glossary.
So
so
I
think
that
the
glossary
of
terminology
section
is
is
always
kind
of
out
to
the
right
of
the
document.
Right
you,
you
always
wind
up
having
to
go
back
to
it,
whether
it's
up
or
down
it
just
doesn't
matter
you
get
to
something
where
you
realize.
There's
a
term
is
being
used
more
precisely
than
you.
B
C
B
B
B
A
D
A
B
B
C
A
There's
a
term
claims
want
to
get
claims
at
the
fourth
line
and
it's
like
the
eighth
word
or
something
and
two
point
one
and
there's
claims,
and
it's
not
capitalized
because,
but
it
should,
I
guess
so
so
if
you
have
defining
it
in
four,
just
capitalize
it
and
it's
like
oh,
this
is
capitalized.
Maybe
I
remember
that
when
encountering
the
definition
or
maybe
there's
a
definition,
I'm
curious
right
now,
I'm
skipping
forward
to
four.
So
so
that's
the
that's
the
style
question
I
have
here
so.
A
Abstract,
no
abstract
doesn't
count,
you
can
do
it
in
abstract,
but
but
but
beyond
that
in
the
body
of
the
text.
I
think
that's
that's
fine
with
me
because
we
move
the
terminology,
basically
in
the
the
meat
of
it
like
in
the
middle
of
it,
and
that's
fine
as
simon
pointed
out,
but
but
but
I
think
we
should
retrain
the
idea
of
capitalization
of
these
terms.
C
That
term
is
common
in
many
ietf
contexts.
I
think
right
now
that
one
was
in
the
intent
whether
the
intend
is
good
or
bad
was
its
capitalized,
starting
at
the
glossary
and
afterwards
but
uncapitalized
before
that,
and
if
there's
cases
where
that's
not
true,
then
we
can
talk
about
that
or
fix
him
or
whatever.
I
would
say,
claims
collection
is
lower
case
in
my
opinion,
but.
A
C
C
A
Let
the
rfc
editor,
do
it
all
standalones,
not
otherwise.
Semantically
tied
claims
should
be
adhered
to
a
terminology,
and
then
we
do
this
with
all
the
terms
in
the
section
four
and
then
I'm
fine
with
that.
I
think
this
actually
helps
the
reader.
C
C
C
B
Okay,
well
so
we
agree
with
the
the
algorithm
here,
and
so
what
we
probably
should
do
is,
I
think
I've
got
all
the
claims,
but
I'm
going
to
make
a
pull
request
about
claims
and.
A
B
Suggest
that
we
make
pull
requests
for
other
ones
so
that
we
can-
and
I
hope
we
won't
have
to
discuss
this
as
a
group-
we
can
just
review
and
say
we
got
them
right
or
not.
B
A
I
would
have
seen
fix
this
in
the
end.
B
A
C
A
C
B
Okay,
it's
top
of
the
hour.
We
still
have
some
issues
that
don't
have
names
attached
to
them.
This
one
needs
to
be
well.
This
one
is
thomas
and
I
know
it's
pulled
by
there,
but
I
think
everyone.
A
B
All
the
other
issues
have
a
either
it
won't
fix
or
has
a
name
attached
to
it.
So
please
look
through
your
names
if
you're
looking
through
here,
you
can
put
your
name
and
get
just
whatever
is
attached
to
you.
B
So
I'm
going
to
be
trying
to
bug
anybody
on
thursday,
who
I
don't
see,
a
pull
request
for
something
closing,
and
I
would
like
us
to
be
able
to
say
after
friday's
meeting
that
we're
done
with
this
with
this
round
and
that
we
would
like
it
to
go
on
to
the
ad
review,
who
will
no
doubt
cause
us
to
open
or
re
revisit
issues
close.
B
A
And
thank
you
for
raising
that
michael,
I
think
you're
addressing
a
a
host
of
majorities
concerned
here.
B
Great
okay,
all
right
we'll
talk
on
friday.
Remember
we're
an
hour
later.
Then.
We
then
this
time.
Whatever
this
time
is
for
you
and
there
is,
was
an
invite
in
the
mailing
list
and
many
of
you
answered
already.
So
that's
great.
Thank
you
very
much.
B
Yes,
you
know,
or
maybe
yeah
interesting
thing
about
posting
a.
What
do
you
call
it
a
an
invite
to
the
mailing
list?
Is
you
get
all
sorts
of
people
that
aren't
really
that
heavily
involved
and
they
say
no
you're,
like
I've,
never
heard
of
that
person
in
my
life?
But
you
know
it's
great
to
know
they're
on
the
mailing
list,
all
right
kathleen
any
questions
from
you.
E
No
sorry,
I
was
multitasking
I
during
this
entire
call
with
multiple
things,
so
I
listened
caught
some
of
it,
but
when
you're
ready
for
me
to
do
another
read-through,
I
will
do
that
and
we're
having
a
snow.
B
B
B
To
this
call,
my
son
is
back
in
the
school's
open
this
week
again
and
he's
back
in
school
for
the
first
time
in
six
weeks,
and
he
he
doesn't,
it
was
a
it
was
it.
It
was
a.
It
was
well
he's
a
teenager,
so
he
didn't
want
to
go
to
get
up
or
go
to
school
anyway,
but
it
was
extra
difficult
anyway
I'll
talk
to
you
guys
on
friday
at
what's
11
a.m.
For
me,.