►
From YouTube: RATS Architecture Design Team, 2020-05-19
Description
RATS Architecture Design Team, 2020-05-19
A
sarah?
So
we
spent
a
lot
of
time
last
week
on
86,
which
was
a
bunch
of
updates
from
ned
that
net
had
initiated
an
update
to
freshness.
A
I
don't
want
to
start
with
that
one,
because
I
think
that
maybe
there
was
86.
I
think
that
maybe
we
can
deal
with
some
of
the
other
pull
requests.
First,
I
thought
that
is
there.
Anyone
here.
Can
anyone
hear
me?
Am
I
talking
into
the
world
hello.
A
Okay,
they're
all
very
quiet
and
muted.
I
wondered
if
82
could
be
merged
at
this
point,
but
hank
singh
feels
strongly
against
this
in
word
and
then.
A
So
I'm
looking
for
low-hanging
fruit
that
we
can
deal
with
quickly
to
get
them
off
of
there.
What
about
this
issue?
Number
77.
A
All
right,
so
I'd
like
to
start
with
this
one
dave.
You
had
a
bunch
of
complaints
about
about
grammar,
which
I
thought
I
damn
it.
I
thought
I
fixed
this
one.
Oh
yeah.
D
A
It
is
my
fix,
is
here
about
grammar?
I
don't
know
if
I
can
get
that
all
on
the
same
screen
here.
A
C
C
A
Cancel
still
get
that
so
basically,
it
introduces
text
only
changes
here
at
the
beginning.
B
B
Yeah,
that's
an
interesting
point.
Yeah.
I
don't
think
we
have
to
define
that.
I
tried
not
to
define
it,
but
maybe
the
the
quotes
might
be
confusing,
so
we
can
just
remove
the
quotes,
but
there
is
a
there
is
a
not
internet
protocol
using
aspect
here.
So
I
I
try
to
yeah.
C
I'm
suggestion
we
just
delete
that
one
sentence:
that's
in
the
middle
there,
unless
there's
some
place,
that
we
actually
need
to
use
the
term
local
conveyance
for
something,
because
I
think
the
other
sentences
around
it,
I
think,
are
okay.
B
Yeah
we
can
remove
this
conveyancer,
sometimes
yeah,
okay.
This
is
basically
yeah
yeah.
C
B
Yeah,
if
there
is
no
benefit
to
this,
conveyancer,
sometimes
referred
to
which,
if
you
start
a
sentence,
this
way,
probably
it's
just
a
afterthought.
In
any
case,
they
can
just
we're.
B
That's
fine,
it
does
not
take
anything
out
of
the
text,
basically
and
sorry
for
not
reviewing
your
reviews,
dave.
That
was
an
oversight
again,
but.
B
B
E
C
Was
there
more
below
that?
I
don't
remember
no
leave
that
on
the
screen
for
a
minute
before
I
say
no
objection,
one
more
time
now
that
you
give
me
an
answer,
I
can
read
the
rest
of
it
in
that
context,.
B
B
So
it's
basically
based
on
the
assertion
before
so,
if
you,
if
you
have
an
a
tester
and
a
verifier
on
the
same
system,
but
they
are
fully
isolated,
they
behave
like
separate
entities.
Their
faults
we
are
talking
about.
These
functions
are
merged
code
wise.
They
are
not
isolated,
and
so
now,
if
you
have
a,
if
you
think
ahead
now
and
you're
now
going
with
the
oh,
my
network
quality
is
low,
sorry
and
think
about
the
attestation
results
that
are
created
by
that
verifier.
B
That
is
also
in
a
tester.
Yes,
then
it
might
have
an
impact
on
how
the
evidence
is
composed.
So
this
is
more
on
the
composed
evidence
side.
I
assume
and.
B
B
Yeah,
this
is
a
french
case,
so
maybe
it's
more
confusing
than
illuminating.
Would
it
take
anything
away
from
you
if
we
just
skip
my
after
40
also
about
the
citation
results,
because
it's
just
a,
I
think
it's
an
implicit.
So
you
want
to
just
remove
the
sentence.
Yeah
I
mean
you.
I
would
have
to
add
more
text
to
explain
this
text
and
then
it's
a
red
state.
You
know
so
maybe
just
removing
it
doesn't
hurt.
C
A
C
C
If
I'm
understanding
your
point,
then
you're
saying
it's
acting
as
a
verifier
or
things
on
the
system
bus,
you
know
so
yeah.
C
If
I
understand
right,
you're,
saying
things
in
the
system
bus,
you
might
have
say
one
thing
on
the
system:
bus
that
is
a
a
tester
and
another
thing:
that's
on
a
system
bus,
it's
a
like
party
and
you're,
acting
as
the
verifier
for
the
communication
between
those
two
devices
are
those
two
things
on
the
system:
bus
and
you're,
saying
completely
independent
from
that
you're,
acting
as
a
tester
for
things
for
things
you're
talking
to
across
the
network,.
E
C
Just
making
sure
my
understanding
is
correct,
you're
saying
there's
two
things
going
on
they're
completely
unrelated
in
the
scenario.
One
thing
is:
it's
a
device
on
a
network
and
it's
in
a
tester
across
the
when
it's
talking
to
other
devices
across
the
network,
separate
from
that
there's
a
system
bus
inside
and
it's
acting
as
a
verifier
between
devices
on
the
system,
bus,
exactly
support.
So
it's
for
shattering.
A
C
A
C
I
have
no
objection,
it's
basically
ways
to
improve
it,
but
I'd
say:
there's
nothing
broken
about
it
right
now.
So
yeah,
I'm
okay
with
it.
C
I
don't
know
if
anybody
else
has
comments,
but
thanks
for
addressing
mine.
A
Okay,
I
I
it
has
to
get
rebased
before
we
can
merge
it,
so
I'm
gonna
do
that.
So,
let's
move
on.
D
F
With
let's,
let's
do
freshness,
I
think
that
was
a
simpler
one.
F
There
were
some
so
two
things
going
on
here:
one
was
just
some
white
space
was
being
deleted
and
then
the
other
one
was
replacing
a
wall
of
text
with
sort
of
three
bullet
items
summarized.
A
A
F
This
was
the
we
have
context.
Was
there
was
a
bunch
of
discussion
and
gary
provided
a
wall
of
text,
lots
of
discussion
and
then
finally,
my
comment,
I
commented
saying
well,
it
seems
like
there's
just
three
basic
ideas
that
are
being
discussed
here.
I
captured
that
in
notes
and
then
said:
hey
those
notes
are
pretty
good.
A
A
C
If
you,
there
was
something
that
did
recently,
because
I'm
going
to
have
some,
hopefully
minor-ish
wording,
suggestions
on
the
the
text.
That
ted
actually
did
that
net
actually
posted.
So.
C
Yeah,
that
is
it
scroll
up
yeah.
This
is
this
is
the
right
thing:
okay,
yeah,
because
this
is
the
the
here's,
the
three
bullets
that
came
out
of
it
right.
Okay,
so
the
tank
commented
on
some
before
me,
so
we're
getting
close.
I
think,
on
the
closer
on
this
one,
but
there's
still
a
couple
hurting
mods
to
do
so.
F
C
Seeing
the
bullets
one
bullet
703
is
your
point
number
one.
C
Let
me
just
say
what
mine
is
I,
since
I
didn't
do
the
suggestion
on
mine,
I
think
environment
and
environment,
both
of
those
should
have
a
parenthesis
s
after
the.
C
B
That's
fine,
oh
listening
to
me
so
well,
okay.
I
think
this
is
very
straightforward,
so
the
text
says
a
conveyance
protocol
that
is
authenticated
and
integrity
protected.
B
So
if
I
would
take
this
literally,
I
would
say
I
don't
care
if
this
protocol
is
authenticated
or
integrity
predicted,
I
want
evidence
to
be
protected,
basically,
integrity
and
maybe
even
obfuscated
and
remote
peers
are
hopefully
authenticated.
So
it's
not
the
protocol.
I'm
talking
about
all
these
attributes
are
somehow
associated
with
the
protocol,
but
I
want
authenticated
peers.
I
want
protected
evidence
and
it
is.
The
protocol
has
to
do
all
this,
but
the
text
doesn't
say
this.
F
But
I
think
that
was
the
point
of
the
discussion
was
that
the
the
protocol
had
the
protocol
had
these
properties
and
the
evidence
and
so
forth
had
no
protections.
B
Yeah-
and
that
is
what
I
wrote
so
what
I
will
write
is
a
convent
value
must
enable
to
protect
evidence
and
to
afford
okay
authenticate
the
peers
involved
in
the
communication,
or
something
like
that.
I
don't
know
so,
not
just
the
phrases
just
off.
That's
my
only
point.
I
didn't
provide
a
alternative
because
I
was
not
sure.
Maybe
we
are
really
talking
about
an
authenticated
protocol,
but
who
is
the
authenticator
of
the
protocol?
So
I
think
this
is
just
phrasing,
but
I'm
not
100
sure.
D
I
have
a
question
here:
you're
talking
about
authenticating
peers
and
protecting
evidence.
I
mean,
isn't
and
and
both
peers.
B
D
F
Yeah
the
point,
I
think,
though
I
don't
think
the
christian
wording
is
incorrect.
I
don't
think
that
was
hank's
point
he's
just
saying
you
could
word
it
differently.
There's
another
there's
another
correct
way
to
word
it
where
the
focus
is
on
evidence
instead
of
the
protocol,
but
either
way.
The
point
is
that
the
the
the
authentication
and
integrity
protection
that's
applied
to
the
to
be
semantically
the
same
and
apply
to
the
evidence
in
in
the
context
of
the
roles,
in
other
words,
you're
authenticating.
F
The
the
authentication
of
the
tester
in
the
protocol
is
the
same
as
authenticating
the
tester.
As
of
the
end
point,
it
could
be
different,
but
the
point
of
what
the
what
the
text
is
trying
to
say
is
that,
if
that,
if
they're
the
same,
then
then
everything
works.
The
geary's
point
is.
I
D
Quite
following
all
this,
and
but
it
seems
to
me
that
the
core
goal
here
of
attestation
is
to
protect
the
evidence
which
means
kind
of
authenticate
where
it
came
from,
I
mean
and
yeah.
Okay,
the
the.
F
D
I
understand
that,
but
so
the
requirement
is
to
protect
the
evidence,
and
that
could
be
done
by
signing
the
evidence
as
sort
of
an
independent
data
unit.
That's
going
to
be
trans
transported
over
a
protocol
that
has
no
particular
security
characteristics
or
it
could
be
done
because
the
protocol
authenticates
itself,
I
mean
or
authenticates
the
attester
in
a
way.
That
is
good
enough
for
the
verifier
in
an
equivalent
way
for
the
verifier.
D
That
seems
like
the
two
choices
here:
yes
and
that
and
that
that
authentication
of
the
the
the
a
tester
in
the
protocol
has
to
be
sort
of
equivalent
to
what
is
being
done
for
attestation.
So
if
the
key
set
key
set
up
for
that
is
not
sort
of
in
the
same
way
that
the
attestation
key
would
be
set
up,
then
you
that
protocol
is
not
good
enough.
A
Well,
I
I
think
that's
up
to
the
the
appraisal
policy
right,
so
I
think,
in
the
case
of
some
of
the
case
that
jerry's
exam
imagine
the
it's
a
symmetric,
sim
card
type
key
that
is
used
to
create
the
conveyance
channel
and
the
verifier
has
fairly
high
confidence
in
in
that
in
its
immutability.
A
And
so
the
fact
that
this
is
a
symmetric
key
channel
is,
which
would
you
might
think
of
as
weaker
is
not
in
this
case.
It's
not
in
this
case
weaker.
It's.
It's
maybe
stronger
from
that
guy's
point
of
view.
D
Yeah
and
the
question
wasn't
symmetric
versus
asymmetric
and
I'm
trying
to
think
of
some
examples
here
but
which
I
don't
have
at
the
moment,
but
seems
like
that
that,
like
it
might
be
a
tls
channel
and
the
the
the
testers
tls
key
is
a
private
key
is
not
the
same
kind
of
thing.
You'd
expect
for
the
setup,
for
it's
not
the
same
kind
of
thing.
You'd
expect
you
know,
maybe
it's
that
that
key
is
put
in
there
by
a,
but
it's
an
ad
vendor,
rather
than
it's.
A
Not
it's
it's
not,
but
lawrence.
It's
appraisal
policy
is
what's
going
to
determine
that
right,
it's
not
arbitrarily!
I
just
swap
one
for
the
other
and
say
it's
going
to
be
equivalent.
It's
no
in
this
protocol,
we're
expecting
a
secure
channel
and
we're
investing
implicit
trust
in
that
channel
because
of
the
characteristics.
So
it's
not.
It's
not
arbitrary,
which
I
think
that
you're
trying
to
struggle
with.
D
I'd
like
to
say
I
mean
it
seems
like
we
should
list
some
characteristics
of
that
that
that
key,
I
I
think.
D
I
can
see
why
the
appraisal
policy
is
a
factor
here.
I
agree
with
that.
I
guess
I
need
to
think
think
through
some
examples
here
of
how
this
key
might
get
set
up,
because
the
intent
of
the
the
I
mean
the
way
that
the
key
material
for
that
communication
protocol
may
have
been
set
up
may
have
had
nothing
to
do
with
attestation
itself
and
it
may
not
serve
absolutely.
A
D
Right
but
there's
there's
some
characteristics
of
maybe
some
examples
or
some
characteristics
of
communication
protocols
that
some
in
some
cases
that
they.
A
D
D
D
Saying
that
I'm
just
trying
to
get
get
some
general
guidance
characteristics
of
of
a
key
key
material
that
that
definitely
works
key
material
that
doesn't
work
set
up
so
that
that
work
and
some
stuff
that
don't
work
just
get
some
sort
of
calibration
or
some
sense
of
what
that
what's
going
on
there.
So
these.
F
B
Your
co-author,
what
was
in
his
intent
here,
I
think
it
is
all
about
the
conveyance
of
eat.
I
think
that
is
a
good
example
for
you
internally,
and
maybe
you
can
you
can
converge
on
that
topic
and
find
out
and
what
was
the
intent
here,
because
I
assume
it's
about
the
secure
channel
set
up
and,
of
course,
keys
have
to
manage
in
a
in
a
meaningful
way,
and
but
that
is
probably
again
as
we
just
established
appraisal
policies
decision
if
that
is
meaningful
to
it.
B
A
C
I
on
that
point
two
yeah,
so
I
there
point
two.
I
just
added
that
after
we
did
after
you
did
your
parenthesis
s.
We
need
to
make
a
similar
point.
Two
has
the
same
issue
as
point
one
right
when
it
talks
about
via
testing
environment,
and
so
I
I
don't.
C
Way
to
rephrase
this
one
is:
is
it
a
testing
environment,
parentheses
s?
I
don't
know.
If
that's
is
that
correct
now
it
says
the
evidence
in
the
conveyance
channel
is
supplied
by
testing
environments
right
now.
Evidence
is
by
definition,
supplied
by
a
testing
environments
right,
and
so
what
is
this
point?
I
was
trying
to
recollect
what
was
point
number
two
actually
intending
to
be
about,
because,
right
now
it
sounds
like
ball.
That's
always
true
by
definition,
isn't
it
so?
How
is
it
a
constraint?
C
A
I
guess
it
might
not
be
true
in
situations
where
the
the
conveyance
channel
is
opened
by
an
entity
and
then
is
then
takes
data
from
other
places,
because
this
conveyance
channel
maybe
has
other
purposes
as
well.
A
C
Right,
I
think
we
need
to
have
make
sure
that
the
text
says
that,
because
I,
when
I
read
it
this
morning,
it
did
the
text
didn't
say
that
to
me.
But
I
think
your
point
is
exactly
right
and
we
just
say
it
at
meaning.
I
think
that's
what
nid
was
trying
to
capture
when
I
read
it.
The
first
time
and
I
said
hey,
I
agree
with
the
articulation,
but
I'm
ready.
A
C
A
Channel
is
I'm
not
sure
if
I
want
to
say
he's
used
exclusively
for
conveying
evidence
or
no.
A
If
the
conveyance
channel
has
multiple
purposes,
then
all
evidence
comes
supplied
by
the
testing
environment.
Evidence
is
always
applied
by
a
testing
environment.
C
A
Since
it's
not
signed
right,
yes,
then
the
point
is
that
that,
if,
if
okay,
I'm
imagining
something
like
you
know,
I
was
just
on
the
http.
This
call,
you
know
imagining
something
like
the
results
of
of
different
web
endpoints,
sending
data
in
a
single
channel
like
quick
right
and
so
it's
quite
possible
for
things
to
come
from
different
places
and
for
some
of
them
to
tend
to
be
evidence
when
they're
not.
D
C
I
agree
with
the
intent
I'm
trying
to
find
wording
that
does
not
exclude
the
notion
of
weird
attestation
or
perhaps
composite
device
where
there's
multiple
testing
environments
but
they're
still
like
in
the
lairdon
one,
there's
still
one
at
the
bottom
right.
That's
the
one
that
you
in
the
layered
attestation
case
you're
talking
about
the
one
at
the
bottom
right,
the
one
that
signs
the
other
ones
right
in
the
composite
device.
C
A
So
if
we
had
a
layered
well,
I'm
not
sure
I'm
gonna
come.
A
Environment
that
the
that
the
outer
connection
upwards
to
the
verifier
could
be
and
a
channel,
but
that
any
evidence
that
was
from
other
in
testing
environments
that
was
conveyed
would
have
to
have
explicit
signatures.
C
Yeah
right
in
the
layered
attestation,
you
have
explicit
signatures
for
everything,
except
for
the
bottom
one
as
the
bottom
one
is
the
only
one
that
could
be
possibly
unsigned
this
would
be
applying
to,
but
we
haven't
defined
weird
attestation
here,
so
I'm
trying
to
find
a
a
definition
that
actually
makes
sense
if
you
were
to
go
back
to
it
after
another
section
without
having
to
break
into
that
discussion.
Now,.
C
C
Okay,
no,
I
think
you're
right
ned.
I
I
didn't
realize
that
you're,
probably
right,
and
so
maybe
that,
because
I'm
trying
to
figure
out
what's
the
technical
point
here,
the
technical
point
might
be
as
long
as
it's
as
long
as
the
channel
is
tied
to
any
one
of
the
testing
environments.
That
might
be.
Okay.
B
A
C
Maybe
that's
too
complicated,
now
yeah,
I
think
the
with
the
key
stuff
was
fun
to
leave
for
point
three,
but
the
a
testing
environment
is
a
problematic
phrase,
because
there's
only
one
where
later
on,
we
say:
there's
not
just
one
well.
A
That's
what
I
was
trying.
That's
that's
what
the
width
was
about
right.
It's
it's!
That's
what
I
see
what
you
mean
right,
that's
the
point.
The
point
is
that
there
are
other
testing
environments,
but
they
don't
control
this
channel
key
and
therefore
they're.
They
can't
send
unprotected
evidence
they
have
to
send
protected
evidence.
I.
C
Get
your
point:
I
had
misread
the
phrase
as
the
preposition
applying
to
a
different
point
in
the
sentence.
C
C
So
your
point,
which
I
misread
the
grammar
on
was
a
testing
environment
with
the
key
that
protects
the
conveyance
channel
right.
That
would
be
the
equivalent
to
what
you
typed
before.
C
Okay,
so
the
way
that
I
had
misread
this
before
was
that
with
applies
to
like
convey
or
supply.
So
I
read
it
as
all
unprotected
evidence.
That's
conveyed
is
supplied
exclusively
by
the
investing
environment,
along
with
the
key
that
protects
the
conveyance
channel,
which
is,
of
course
not
the
intent
right
right.
C
C
And
so
you're
trying
to
do
it
by
way
of
constraint
on
widget
testing
environment.
So
that's
why
that
has
instead
of
with
would
be
correct.
So.
C
Except
for
with
can
be
misread,
I
would
so
first,
if
you
change
with
to
that,
has
it
has,
is
unambiguous
and
it
has
michael's
meaning
and
now
the
only
possible
way
to
wordsmith
it
better
is
because
that
appears
twice
in
the
sentence.
It
would
be
nice
to
get
rid
of
one
of
those,
but
actually
three
times
in
the
sentence
protecting
this
work
yeah.
Well,
yes,
but
that
wasn't
that
I
was
referring
to
now,
I'm
basically
it's
good
enough
right
now,
but
but
let
me
just
emphasize
which
words
I'd
say.
C
C
This
to
say
using
no,
I
think
that
goes
back
to
having
the
same
problem
as
with
did
it's
applied
using,
so
I
think
it's
other
than
the
fact
that
that
appears
three
times.
It
is
actually
correct
and
unambiguous
right
now,
and
so
I
would
commit
your
suggestion
first,
okay,
and
if
somebody
can
find
a
way
to
get
rid
of
the
fact
that
it
has
three
that's
in
the
sentence
without
going
ambiguous.
That
would
be
fine,
but
I
can
live
with
it.
The
way
it
is.
C
All
right
so
now
we're
back
to
here.
I
had
one
of
the
questions
at
the
very
bottom,
which
is
a
non-technical
question.
F
A
B
White
spaces.
What
so?
I
will
just
if
everybody
understands
what
the
sentence
is
meaning,
so
that
the
protocol
is
not
authenticated
by
provides
authentication
to
the
freemore
peers
and
is
integrally
protecting
the
content
and
other
content
as
a
it's
like
a
weird
sentence
again,
but
if
everybody
thinks
that's
okay,
we
can
do
it
in
a
polished
path.
Later,
that's
just
editorial.
I
assume
we
didn't.
A
B
C
B
C
B
C
C
B
C
C
A
F
A
And
then
so
is
this
section?
Is
this
section
now
butting
against
what
we
just
committed
from
hank
yeah?
We
talked
about
roles
yeah.
F
G
F
So
so
I
would
say
you
know,
commit
the
try
to
try.
I
don't
you
know,
commit
something
and
then
try
and
bring
bring
the
bring
the
the
threads
closer
together,
but
we
spent
a
lot
of
time
on
on
the
definitions
above
which
were
basically
just
intended
as
sort
of
clean
up
in
order
to
make
this
other
paragraph
in
order
for
the
other
things
to.
A
Not
be
you
know,
what
you'd
like
to
do
is
is
commit
this
so
that
we
can
get
all
the
updates
that
were
good,
recognize
that
the
wall
of
text
at
the
bottom
is
overlaps,
with
what
hank,
what
we
just
committed
with
from
hank
and
then
and
then
essentially
we.
Why
does
it
stay
closed?.
A
Closed
does
it
say,
closed
no
okay,
I
didn't
close
it
that
I
remember
I
can
still
commit
it.
A
C
I
don't
know
at
the
bottom,
it
says
to
reopen
pull
request
and
it's
grayed
out.
I
wonder
if
I
can
reopen
it.
C
A
Oh,
you
know
what
I've?
Okay?
Okay,
no!
I
did
merge.
Okay!
Well,
you
undelete
the
oil
composition
branch
on
the
hang
on
here.
It's.
A
A
Because
I
believe
I
I
merged
this
manually,
I
believe
I
just
merged
this
manually
and
that's
what
made
me
re-delete
the
branch
because
you'd
like
to
delete
the
branch
afterwards.
C
Oh,
this
one
has
a
brainchild.
This
is
a
yeah.
This
is
not
merging
in
the
master
I-86
right.
This
is
from
issue
55,
dural
composition,
but
the
branch
it's
merging
into
has
been
deleted,
so
I
think
yeah.
So
I
think
the
next
thing
is
to
rebase
it
off
with
master
yeah.
You
have
to
rebase
this
one
off
of
master
right
now.
It's
based
off
of
wrong
opposition,
but
you
have
to
rebase
it
off.
A
Yeah
I
merged
I
merged
I
merged
77
manually,
which
was
also
on
the
roll
composition
branch,
and
now
we
have
this
issue,
55
branch
which
those
right,
if
you
look
at
the
variation.
A
C
C
A
C
Yeah
there's
a
number
of
things
in
here
that
I
didn't
like
I
saw,
I
think,
responded
to
or
commented
on.
One
of
them,
and
it
looks
like
thomas,
had
a
bunch
of
comments
on
this
one
separately
from
unrelated
to
mine.
J
C
Thomas,
did
you
want
to
talk
about
your
one
of
your
issues
on
this
one.
J
Yeah
well
yeah,
maybe
better
to
to
address
it
as
a
meta
comment.
So
if
we
really,
my
point
is
that
we
really
really
need
two
sub
terms
to
specify
what
freshness
means.
J
Then,
then
we
probably
need
to
make
them
to
make
sense
in
in
a
non-ambiguous
way
right,
otherwise
we're
screwed,
apparently
hank,
and
I
can't
reach
an
agreement
on
on
these
two
sub-terms,
which
are
recentness
and
uniqueness
related
to
freshness
and
and
and
I
think
we
managed
to
drive
the
conversation
to
to
a
dead
end
right
now.
So
so,
unless
someone
else
jumps
in
and
resolves
the
conundrum,
I
suggest
we
forget
about
this
pr
and
and
just
live
with
one
true
freshness.
You
know
independent.
There
is.
B
At
the
station
right
on
adjustments,
there
are
different
types
of
freshness.
Then
we
will
talk
about
evidence
only
here,
which
is
not
enough.
I
think
that
was
pointed
out.
We
have
multiple
solutions
out
there
and
they
have
different
types
of
freshness.
So
we
cannot
talk
about
flashes
in
the
architecture
at
all.
If
you
want
to
have
the
freshness
defined
by
the
solution,
so
either
we
prune
freshness
from
the
architecture
entirely
or
and
leave
it
to
solutions
to
talk
about
it
or
we
have
a
relatively
stable
basis
here.
C
On
one
point
I
agree
with
thomas,
which
is,
I
think,
some
of
the
content
in
this
pr,
I
think,
is
probably
more
than
what
I
would
want
in
the
architecture
document
you
see.
One
of
my
points
is
also
on
the
screen
or
whatever
around
uniqueness
and
things.
I
didn't
see
why
that
needed
to
be
in
this
document.
C
C
I'm
not
necessarily
opposed
to
adding
more,
but
I
want
there
to
be
a
good
justification
and
I
haven't
seen
the
good
justification
yet
and
so
I'm
trying
to
be
open,
but
I
didn't
see
that
that
this
was
necessary
and
if
we
don't
merge
this
one,
I'm
okay
with
that,
but
I'm
not
sure
which
problem
we're
trying
to
solve
with
this
one.
That's
why
I
said
I'm
open
discussion
so.
B
Three
different
kinds
of
basic
interaction
models
here,
which
is
a
thing
based
quote
thomas
recently,
is
based
on
uniqueness,
which
is
the
challenge
response
thing
that
uses
a
non-value
that
is
unique.
B
In
terms
of
I'm
very
sorry
for
that
correct,
so
we
have
this
freshness
based
on
uniqueness
as
the
let's
call
it.
That
is
the
the
unique
value
we
call
it
the
nonsense,
cryptography
cryptography,
but
you
can
basically
call
it
a
unique
thing
that
you
handshake
here
and
then
we
have
the
synchronized
time-based
stuff.
That
does
not
need
it
and
therefore
does
not
provide
you
uniqueness,
very
necessary
thing
for
one
solution,
but
it's
super
unnecessary
for
the
other
solution.
C
B
Basically,
just
after
thought
and
might
and
and
illustrative,
we
have
to
have
an
understanding
for
for
more
than
that
that
you
can
base
stuff
on,
and
I
think
it's
very
important
to
understand.
The
freshness
has
two
characteristics
here
and
typically
what
you
sub
everybody's
talking
about
nuns
and
is
thinking
uniqueness,
but
it
is
actually
for
attestation.
The
recentness
is
the
interesting
part
uniqueness
you
can
always
eliminate
by
by
state.
B
If
you
get
the
same
thing
again,
you
should
start
to
get
suspicious,
so
state
can
solve
your
uniqueness
problem,
but
sometimes
you
don't
have
state
so
unique
is
very
important
to
the
quality
of
evidence,
and
that
is
what
I'm
trying
to
establish
here.
It's
just
a
fundamental
point
of
reference.
How
evidence
characteristics
are
retained,
even
if
you
have
small
solutions
or
big
solutions,
and
I
think
that
is
what
an
architecture
is
for
in
general.
It
helps
you
to
build
a
solution
you
with
with
prone
to
error.
B
Then
you
fail
with
creating
solutions
from
an
architecture,
and
that
is
the
only
reason
why
I
pulled
it
in
here,
because
if
you
leave
that
to
another
document
yeah
well,
it
could
be
the
interaction
model
in
theory.
So
we
could
then
again
remove
the
concept
entirely,
because
it's
not
up
to
the
architecture
anymore.
That's
my.
C
Point
I
mean
we're
about
out
of
time,
like
I
said
right
now.
My
opinion
is
the
text
in
the
document.
Right
now
tries
to
strike
the
right
balance,
meaning
there's
a
freshness
section
in
the
main
body,
there's
an
elaboration
in
the
appendix
which
talks
about
both
nonces
and
time
stamps.
There's
two
different
ways
to
do
things:
that's
the
appendix,
because
those
are
starting
to
bleed
into
mechanism
more
like
the
interaction
model
stuff.
C
I'm
also
noting
that
we're
about
out
of
time,
because
it's
eight
o'clock
right
now
and
I
don't
know
where
we're
at
with
the
one
that
michael
was
trying
to
rebase,
but
I
don't
know
if
this
is
a
good
place
to
pause
and
pick
up
again
next
week.
I
think
so.
A
So
I
I
did
manage
to
rebase
it,
but
I
had
to
open
a
new
issue,
it's
89
and
it
has
both
hank
and
ned's
rewrite
of
the
roles,
because
I
that
was
a
merge
conflict,
so
I
put
them
both
in
for
now
and
we'll
pick
one
and
then
there's
some
other
text
that
is
needlessly
different
all
right.
So,
let's.
A
A
A
A
I'll,
let
you
do
that
and
I
there's
there's
hank
text
and
ned
text,
and
so
you
may
actually
just
prefer
that
what
we
just
we
did
at
the
beginning
of
the
hour
anyway,
so
you
may
just
vote
against
the
revised
text.
Gotcha,
because
you're
saying
there's
two
different
variations.