►
From YouTube: RATS Architecture Design Team, 2020-02-11
Description
RATS Architecture Design Team, 2020-02-11
B
D
C
C
C
So
what
I
was
sharing
was
the
list
of
issues
again
we're
sharing
what
you're
not
sharing
right
now
but
I
know
I
was
sharing
it
now,
I
think
I'm
sharing
it
yep
like
it
and
let's
just
give
it
another
oversizing
tonight.
Here's
everybody
almost
was
here.
Ladies
here
you're
here
Sarah,
we
haven't
heard
from
you
before
you
get
your
audio
hooked
up.
Is
it
working?
Can
we
hear
you
do
on
a
telephone?
Oh
you
have
time
on
the
telephone.
Okay,
good
just
want
to
you,
say
my
needs
and
done
work.
C
C
D
There
was
a
discussion,
we
should
probably
add
an
issue.
The
question
was
to
what
extent
security
considerations
section
of
the
rats
architecture
should
talk
about
how
long
things
are
valid,
for,
in
other
words,
how
often
should
you
reattached
right,
and
so
you
some
considerations
for
people
who
want
to
use
the
rats
architecture
to
keep
in
mind,
and
we
do
already
have
text
in
the
architecture
document
that
says
you
know
it's
kind
of
a
race
condition
right,
because
the
evidence
can
change
right
after
you
send
it
to
whatever
but
I.
Think
in
the
security
consideration
section.
D
D
C
D
D
C
Where
I
am
okay,
so
he
says
he's
on
his
way.
Ned
is
on
his
way.
He
says
all
right,
you
know
you
guys
were
having
a
meeting
in
Miami.
We
could
have
changed
this
or
something
okay,
so
I
entered
that
issue.
Arm
I
was
trying
to
deal
with
this
one,
which
was
I.
Think
this
is
correct
unless
you
want
me
to
pull
up
the
text.
C
D
D
B
D
D
D
That's
our
present
rock
layer,
devastates
right,
the
ROM,
the
firmware
opti
and
a
trusted
application,
and
let's
say
that
you
have
an
old
version
of
the
firmware,
and
the
verifier
knows
that
firmware
that
there
is
a
known
vulnerability
in
f
is
now
public,
and
so
you
say:
hey
the
rom
is
good.
The
op
t
is
good,
the
ta
is
good,
but
that
trusted
firmware
is
bad
and
nobody
should
trust
this
thing
anymore,
because
the
vulnerability
is
public
and
anybody
could
be
exploiting
in
any
type,
and
so
that
could
come
back
in
the
attestation
result.
C
E
Hi,
this
is
say
then,
inaudible,
yes,
okay,
good,
so
I
think
that
an
attestation
resided
is
always
a
success.
If
you
retrieve
part
of
the
attestation
result
contents.
So,
for
example,
are
you
know
that
the
firmware
on
your
device
is
compromised
because
of
a
new
CV
or
compromisable
yep?
And
now
you
check
that
and
the
expected
result
is
yes,
you
have
compromisable
firmware.
That's
a
success.
I.
D
Was
treating
the
word
success
and
being
it
passes
the
policy,
so
one
example
that
we've
talked
about
before
is
where
the
verifier
just
says.
Yes,
it's
good.
Instead,
it's
a
claim
that
says
yes,
a
test.
The
verifiers
policy
and
airline
party
doesn't
just
simply
trust
the
fact
that
it's
signed
by
a
verifier
right.
That's
not
the
only
way
to
do
it.
That's
one
variation,
so
we'll
discuss,
in
which
case
it,
in
which
case
either
verify
your
signs.
E
B
C
C
D
C
I
agree
with
these:
that
is,
the
status
status,
so
I'm
trying
to
say
that's
like
the
404
response
from
the
verifier,
but
there's
another
case
where
they
issue
a
passport.
Okay,
and
when
you
go
to
present
the
passport
to
the
relying
party
order
agent
at
other
country,
they
say:
I'm,
sorry,
but
you
don't
have
a
visa
or
you
know
we
don't
accept
people
from
your
country,
yeah,
okay,
and
but
it's
a
perfectly
valid
passport.
Okay,
it
was.
It
was
obtained
successfully
from
the
verifier.
It's
just
not
suitable
for
this.
D
E
D
B
D
E
B
D
Depends
on
what
the
verifier
wants
to
put
in
there.
It
could
be
anything
meaning
the
architecture
is
agnostic
as
to
what
that
goes
in
that's
up
to
the
verify
or
whether
it
puts
in
things
derived
from
evidence
whether
it
puts
in
things
that
it
is
created
like
by
the
way.
This
guy
is
a
US
citizen
right,
which
has
nothing
to
the
evidence.
That's
the
value
judgment
that
was
generated
by
the
verifier
or
both
could
be
combinations,
and
so
it's
agnostic
as
to
what
goes
on
the
attestation
result.
All
those
are
possible.
E
D
E
Okay
in
a
passport
model,
I
present
various
form
standardized
evidence
to
a
external
verify,
I'm,
very
sorry,
to
have
mixed
that
up
and
now
the
procedure
is
conducted
and
the
procedure
is
not
successful,
because
the
verifier
was
not
able
to
obtain
all
pieces
necessary
to
appraise
the
evidence.
Then
we
have
a
failure.
Yeah
and
then
we
can
have
the
content
is
successfully
appraised
and
the
result
is
no
way.
You
are
not
going
anywhere
here,
because
you
are
full
of
vulnerable
well
there,
and
that
is
a
policy
decision.
D
Technically,
there's
four
cases,
because
it's
a
case
of
the
very
far
is
not
reachable
I'm.
Putting
that
one
aside.
Okay,
so
you
get
to
the
verifier
and
the
verifier,
the
peddler,
corruption
be
sorry.
You
are
not
eligible
for
me
to
grant
you
anything.
I,
don't
know
who
you
are
I,
don't
trust
you
yeah
I'm
the
wrong
verifier
for
you,
because,
but.
E
D
D
D
A
A
Yeah
so
I
always
differentiate
in
my
mind.
At
the
station
results
from
acetates
attestation
evidence
that
attestation
results
can
we
can
be
just
a
yes/no
decision
from
the
verifier.
Yes,
I
received
a
ballot.
Okay,
now
I
didn't
every
saw
adaptation.
Evidence
is
what
a
relying
party
can
use
to
compare
against
their
policy
right,
so
I'm,
not
sure
that
I'm,
not
sure
result
is
really
what
we're
looking
for
here.
When
we
say
the
resulting
result
is
examined
by
the
relying
party
and
based
upon
appraisal
policy.
The
result
may
not
be
particularly
verbose.
D
So
I
disagree
with
one
of
the
things
that
you
said:
I
think
it
is
a
result.
I
think
both
cases
that
you're
talking
about
are
different
things
that
can
appear
in
an
attestation,
but
you
use
the
term
attestation
evidence
the
evidence
never
goes
to
the
relying
party.
Everything
is
the
ghost
relying
party
is
by
definition,
part
of
the
attestation
result
claim
sets
in
there.
I
would
only
use
the
word
evidence
as
being
the
thing
that
goes
to
the
verifier.
It's
just
a
terminology
point
okay,.
A
D
E
D
It
just
means
that
when
you
do
that,
you
don't
call
it
evidence,
you
call
it,
you
know
it's
still
claim
sets
or
something
maybe
claims
that
it
may
be
the
same
claims,
but
their
claims
in
the
attestation
result
instead
of
claims
in
the
evidence,
this
period
of
terminology
doesn't
mean
it.
You
can't
do
it,
it
could
even
be.
E
D
C
B
D
C
D
A
E
D
D
D
C
C
E
You
say:
yeah,
you
don't
have
to
trust.
The
thing
is
feeding
at
the
values.
Really,
it
is
a
passive
root
of
trust
for
multiple
purposes,
so
I
think.
In
order
to
address
this
issue,
we
have
to
reopen
the
can
of
worms
that
is.
Do
we
want
to
use
the
term
root
of
trust,
yeah,
I
think
because
you're
right,
the
TPM
by
itself,
does
nothing.
It
just
sits
there
and
consumes
it
a
power
alien
and
then,
if
you
use
it,
but
really
it
is
used
by
a
standard
interface
which
is
called
the
trusted.
E
B
I
think
things
hi
is
that
it's
a
garbage
in
garbage
out
kind
of
occurring
opposition
and
rank.
If
you
feed
it
garbage
island
find
your
garbage
break.
Are
we
good?
The
verifier
doesn't
want
to
trust
garbage,
so
there
is
some
assumption
that
the
thing
that's
producing
the
inputs
to
the
TPM
isn't
garbage
correct.
D
B
If
you
go
down
the
TCG
terminology
path,
you
end
up
with
the
route
of
trust,
for
measurement
is
trusted
by
the
other
two
routes,
which
are
the
TPM,
the
route
of
trust
for
storage
and
recording,
and
that
there's
something
called
a
trusted
building
block
which
connects
the
route
of
trust
or
measurement
to
the
root
of
trust,
misdirection,
recording
and
it's
the
combination
of
those
three
roots
and
the
TVB.
That
is
your
quote:
root
of
trust
and
you're.
B
Trusting
everything
and
you're
also
trusting
the
what
they
called
a
trust
chain,
which
is
the
sequence
of
booted
software,
from
the
root
of
trust
for
measurement
to
the
point
where
you
get
to
the
thing
that
is
talking
to
the
TPM
through
the
TSS.
That's
also
trusted,
even
though
the
TCG
doesn't
have
a
name
for
it.
Yeah.
D
B
Sense
memory
does
matter
that
I
think
I
think
that
it's
not
exactly
true
because
from
TCG
perspective,
because
if
you
follow
that
path,
there's
still
the
hey,
I
trust
the
thing
that
I
just
the
code
that's
running
now:
I
trust
it
to
measure
the
next
bit
of
code
that
I'm
about
to
jump
to,
in
which
case
the
attests.
The
target
environment
is
still
not
the
attesting
environment,
although
transit
orally.
It
is
because
you're,
you
know
you're
moving
from
one
transitioning
from
one
state
to
another.
B
D
B
E
D
E
E
E
E
E
B
D
C
E
E
Think
composite
device
is
okay.
If
at
some
point
we
highlighted
every
component,
a
first
of
all
a
composite
device
can
be
a
composite
with
just
one
component
and
then
via
V.
We
covered
the
atomic
device
or
whatever
you
want
to
call
it
and
then
and
then
the
solid
it.
And
then
we
have
the
components
in
the
composite
device
and
all
components
are
entities
that
can
take
on
roles.
Yeah,
I.
D
Do
not
have
a
strong
opinion
this
one
other
than
that
I
would
like
consistency
between
the
sentence
at
the
top
of
the
diagram,
which
says
the
thickness
will
flow
for
a
fill-in-the-blank
right
now.
Is
this
composite
a
tester,
the
thing
that
has
the
the
bottom
line
inside
the
diagram?
Where
has
the
TV
33
after
it
the
title
right
below
the
diagram
which
says
conceptual
software
composite
a
tester
or
whatever
it
is,
and
in
the
sentence
after
that?
This
is
any
composite
of
tester.
D
F
B
A
D
D
C
D
F
For
this
reason,
I
think
the
composite
of
tester
would
be
the
right
thing
for
the
diagram.
Above
one
of
the
reasons
we
had
looking
shoes
as
we
had
an
issue
with
device
boundaries
versus
rural
boundaries
and
I
really
hope.
The
diagram
will
be
talking
about
role,
because
the
tester
B
and
C
to
me
can
exist
from
another
device.
I'll.
C
D
C
D
B
I
think
the
the
use
of
the
term
composite
device
is
describing
an
entity
not
a
role.
It
may
be
that
this
diagram,
taken
as
one
thing
is,
describing
a
role
or
a
use
case.
I'm,
not
certain
but
I.
Don't
think
I
think
I
think
muddies
the
water
to
call
a
tester
at
a
composite,
a
tester
which
essentially
is
defining
a
new
role
versus
okay.
E
B
G
D
E
F
D
E
D
B
E
D
D
B
C
D
B
B
In
the
simple
case
of
an
a
tester
there's
a
target
environment
which
collects
claims
that
are
inputted
into
an
a
testing
environment
that
produces
evidence
to
a
verifier.
That's
we
I
think
there's
another
diagram
to
disk.
That
shows
that
right,
right
and
now
we
have
this
more
complex
diagram
that
shows
a
testers,
B
and
C,
providing
evidence
to
the
attesting
environment
that
produces
composite
evidence
yep
so
either.
So
we
could.
B
If
we
just
wanted
to
focus
on
one
arrow
coming
out,
we
should
be
moving
target
environment
and
collecting
claims,
because
it's
built
with
somewhere
else,
I
guess
or
we
add
the
an
explanation
that
there's
another
arrow
coming
out
with
the
evidence
that
relates
to
the
target
environments,
claims
that
were
collected
was
able
composite
evidence
can
be
devoted
it
or
the
other
way
around
that
evidence
and
consists
of
composite
evidence
or
some
some
nesting
kind
of
structure.
I,
don't
know
I.
D
Don't
think
this
diagram
used
to
change
other
than
the
only
thing
in
this
diagram
now
but
I'm,
not
a
huge
fan
of,
but
I'm.
Okay
with
it
is
the
word
evidence
on
the
lines
from
a
tester
B
over
to
a
testing
environment.
I
think
the
term
evidence
there
is
arguable,
but
I
don't
object
to
it.
The
notion
is,
you
have
different
claim
sets
right,
you
had
the
claim
sets
it's
the
lead
of
testers
claim
set,
that's
in
the
collecting
claims
line
that
comes
in
and
it
appears
inside
the
course
of
evidence.
D
You
also
have
various
claim
sets
that
come
from
a
tester
be
a
tester,
C
and
dot.
Those
are
also
claim
sets,
but
evidence
we
normally
show
is
going
to
a
verifier,
and
we
said
the
attesting
environment
inside
the
composite
device
may
or
may
not
have
a
verifier,
so
if
it
doesn't-
and
those
are
just
claim
sets
that
go
into
the
compositive
and
set
up
to
the
verifier-
and
that's
why
I
said
evidence
is
a
little
bit
overstated,
but
I
don't
object
to
it
in
the
evidence
of
ax,
testers
and
horizontal
lines
and.
E
I
would
disagree
is
that,
because
of
the
a
tester,
B
could
have
been
rude,
of
course,
of
reporting
a
very
subject.
They
founded
signature
with
this
meaning
and
therefore
composes
evidence
by
itself.
We
are
not
just
sending
it
to
a
verifier.
That
is
a
problem
with
the
idea
to
just
sending
it
there
I
assume.
B
D
B
B
G
So,
like
I
have
a
question
that
I'm
confused
on,
because
if
there
is
a
verifier
in
the
intent
in
the
attesting
environment,
then
it
is
potentially
dropping
or
doing
other
things
to
the
composite
evidence,
elements
that
have
been
provided
by
a
testers,
B
and
C
and
contorting
it,
and
it
may
not
be
trustworthy
and
we've
already
talked
about
whether
the
verifier
is
trustworthy
based
on
the
policy
and
I,
don't
see
how
we
can
have
a
policy
that's
applying
to
an
a
testing
environment.
That's
that's
not!
B
F
B
C
Well,
though,
I
don't
forgetting
endure,
we
see
me
going
around
in
circles
and
I
actually
have
like
there's
just
really
subtle
words
that
you
guys
are
using.
It
either
is
just
missing
from
this
diagram,
where
you're
taking
issue
with
pieces
of
the
diagram
that
the
diagrams,
not
even
purporting
to
try.
This
represent.
E
C
Basically
it
is
really
only
about
what
was
your
flat
said.
Is
that
that
a
leader
tester
is
passing
things
on
it
may
be
doing,
it
may
be
acting
as
a
verifier
if
it
has
where
it
may
be
talking
to
a
different
verifier
to
get
attestation
results
from
a
test
for
a
test
or
beam
which
is
passing
on
this
evidence.
But
that's
none
of
those
details
matter
because
they're
all
fought
contained
in
that
in
testing
environment
box
and
right
we're
trying
to
abstract
all
that
so
that
we
just
say
look.
B
C
So
I
had
the
question
a
while
ago.
Was
it
whether
this
is
evidence
from
a
palapa
Temptation
results
from
B,
C
and
D,
etc,
or
whether
or
not
the
evidence
format
was,
would
accommodate
the
fact
that
the
taxation
results
in
fact
no
we're
represented
as
easily
the
same
way,
which
hates
that
I
don't
think
it
mattered
right.
It
did
a
rate
of
result.
Yes,
so
I
just
think.
B
D
D
B
D
B
B
F
Any
description
for
composite
evidence
in
a
couple
weeks
ago:
maybe
it
got
lost,
but
I
can
resend
it
in
I.
Think
it
does
this.
My
big
thing
on
the
other
diagram
is
just
the
only
thing
I'm
worried
about
is
that
we
don't
assume
that
composite
evidence
in
would
come
from
a
composite
device.
So
that's
my
only
thing
that
we
don't
have
a
diagram
that
limits
composite
evidence
just
for
a
composite
device
and
and.
D
D
D
E
E
D
E
D
D
F
C
For
me,
yeah,
okay,
so
in
the
next
five
minutes
or
three
minutes,
we
have
left
I
just
want
to
walk
through
the
other
IDs
Eric.
If
you
can
send
your
guitars
and
Hank
your
github
ID,
so
we
can
define
you
that
issue
Frank
open
a
whole
bunch
of
these
I
guess
is
Wednesday
after
a
call
tried
to
get
the
text
here.
This
one's
closed,
I
think
those
issues.