►
From YouTube: RATS Architecture Design Team, 2022-05-31
Description
RATS Architecture Design Team, 2022-05-31
A
D
B
Anyway,
good
morning,
let
me
see
here
good
morning:
everybody.
B
B
A
few
issues
here
we
said
last
time:
we
wouldn't
fix
these
two,
this
one
on
freshness
details,
I'm
just
gonna
put
that
up.
While
we,
I
hope,
we're
gonna
get
hank
is
gonna,
join
us
because
he's
a
freshness,
expert.
B
B
Falling
trees,
yeah
down
lines
plus,
like
I
think,
some
people
there
was
like
three
or
four
three,
like
their
transformer
blew
up
and
a
tree
took
out
the
distribution
network
and
a
tree
took
out
their
local
network
and
so
yeah
we
had
you
look
on
you
look
on
the
internet.
There's
like
dozens
of
photos
of
you
know
these
200
foot
tall,
you
know,
pylons
that
have
just
crumpled.
B
It's
like
kind
of
a
stephen
was
a
tommy
knockers
or
something.
There
was
a
stephen
king
film
with.
B
B
Okay,
so
I
was
looking
through
the
issues
yesterday
and
trying
to
figure
out.
If
there's
anything,
I
could
suggest
text
for,
and
I
actually
came
up
empty
myself.
B
So
we
already
said
we,
we
wouldn't-
and
I
answered
these
two
by
email
last
week
with
16
that
I
posted
in
the
afternoon,
and
so
I
don't
know
if
there's
really
anything
that
I
think
what
we
need
to
do
is
we
need
to
just
go
through
these
comments
and
agree
as
a
group
that
either
we're
not
going
to
solve
them
or
we
need
to
come
up
with
some
proposal.
What
we're
going
to
do.
B
I
I
think
that
so
I
I
think
that
this
is
this
part.
This
part
is
the
part
that
I'm
I
don't
have
an
answer
to
this.
I
do
yeah,
okay,.
F
The
the
freshness
section
has
three
examples
of
how
you
can
solve
freshness
right
right.
The
other
roles
or
architecture
diagrams
are
things
that
are
inherent
in
attestation
that
you
either
that
you
always
have,
although
they
may
be
combined
with
their
roles
in
any
particular
implementation,
but
they're,
always
there,
epic
id
distributor
is
something
that's
not
in
that
category.
This
is
an
example.
F
Yeah
right,
the
goal
of
the
appendix
was
to
tease
out.
Are
there
actually
any
architectural
requirements
that
need
to
be
incorporated
into
the
main
architecture?
And
this
is
showing
that
the
appendix
right
now
is
saying:
hey
here's
three
ways
to
get
freshness
that
don't
change
the
earlier
stuff
that
you
could
use
any
of
these
or
even
though
a
fourth
one
to
build
a
solution
out
of.
F
Typical
ones
that
you
might
want
to
reuse
one
of
these
and
if
you
do,
then
here's
some
considerations,
you
want
to
use
when
building
your
actual
solution.
I
think
that's
how
I
understand
the
purpose
of
appendix
right
now.
Well,.
B
I
would
say
I
want
to
go
into
the
third
dimension
and
I
want
to
put
it
above
all
of
the
other
items
such
that
it
rains
down
the
epoch
ids
to
everybody.
And
so
that's
why
it
doesn't
fit
in
very
well
and,
as
you
say,
it's
optional
right
and
the.
F
B
D
F
D
F
D
E
B
F
You
could
any
way
you
draw.
The
diagram
would
be
wrong
in
some
implementation,
of
course,
of
course,
so,
for
example,
for
example,
what
I
mean
is
you
could
have
two
epic
id
distributors,
one
that
distributes
ids
to
a
test,
run
verifier
and
a
different
one
that
distributes
ids
to
verifying
relying
party
and
they're.
You
know
ships
in
the
night
right.
That
would
be
valuable.
F
D
E
F
C
Conte
comment
really
is
hey:
are
these
like
examples
or
whatever
you
know,
so
I
guess
an
objective
question
would
be.
C
A
Could
point
to
daa
as
an
example
of
a
you
know,
a
role
that
wasn't
in
the
original
architecture
document.
It's
still
valid
to
talk
about.
A
C
Yeah,
so
if
you
wanted
to
hit
them
in
their
face
a
little
bit
harder
right
after
where
you
have
highlighted,
you
could
say
that
these
examples
don't
constrain.
You
know
the
architectural
choices
they're
really
just
to
illustrate
right,
because
it
sounds
like.
The
comment
really
comes
from
a
belief
that
maybe
that
the
example
is
more
than
what
it
is.
F
I
don't
know
I
sort
of
feel
like
the
second
sentence
of
the
paragraph.
There
is
the
one
that
was
added
to
the
first
sentence
after
the
same
discussion
happened
in
the
second
second.
C
C
C
B
Why
the
reason
why
I
you
know
why
I
why
I
wanted
well.
B
I
agree
that
this
time
considerations
belonged
here
was
two
reasons,
but
one
of
them
is
that
it
became
obvious
to
me
that
if
you're
implementing
and
you
haven't
thought
about
freshness
that
you
will
come
up
with
it
with
some
stuff
that
suddenly
you,
if,
when
you
want
to
add
freshness
to
it,
you
suddenly
discover
you
have
to
add
a
round
trip
which
you
hadn't
thought
of
before,
and
the
round
trip
may
in
fact
be
in
a
you
know
in
a
place
or
just
a
way
that
you
know
violates
privacy
or
something
else,
and
so
you
have
to
think
about
freshness
at
the
very
beginning
of
the
process,
and
here
we've
thought
about
freshness
in
some
detail
so
that
you
won't
go
and
do
something
less
well
thought
out
and
that's
why
we
put
it
in
the
in
this
architecture
as
an
appendix
as
a
you
know.
B
Here
are
three
thought
out:
three
well
thought
out
examples
of
how
to
do
it
and
if
you
have
another
one
of
course
do
that.
But
you
better,
you
know,
know
what
you're
doing
right
and
that's
why
you
know
we
did
this
and
we
all,
but
we
also
you
know
we
didn't
explain
how
to
use
tls
right,
because
other
people
did
that
we,
you
know,
I
mean
just
in
general,
how
to
use
tls
right.
B
We
didn't
explain
how
to
use
certificates
yeah,
we
didn't
explain
a
whole
bunch
of
other
things
because
there's
other
good
documents,
but
there's
no
other
good
abstract,
and
this
is
still
very
abstract.
This
section
it
doesn't,
you
know,
tell
you
anything
about
formats
or
the
epoch,
ids
or
really
how
they
travel,
or
you
know
what
you
can
do,
or
you
know
how
the
timestamps
do,
or
the
nonces
or
whatever
so
anyway.
So
the
question
is:
are
there
any
you
know?
B
Are
there
any
other
responses
here
there,
and
here
the
daa
introduces
a
role.
Do
you
should
we
reference
that
in
informatively.
A
C
Freshness
available,
like
hugely
important,
I
mean
there's
other
topics
that
we
don't
talk
about
like
ordering
and
stuff
like
that.
That
are
important
as
well,
and
you
know
we.
I
don't
know
that
we
where
we
go,
and
you
know
I
think,
last
week
I
was
talking
about
not
going
down
the
path
of
adding
a
whole
lot
more
to
the
document.
We're
trying
to
close
it
out.
F
I
would
probably
do
is
that
we
take
that
first
full
paragraph
and
use
that
in
an
email
response
to
the
question
it
says
in
the
past
we'd
edit,
the
second
sentence,
because
this
question
was
asked
before:
how
do
you
want
us
to
say
it?
It
seems
like
it's
good,
as
the
other
thing
that
I
would
do
is
on
the
comment
up
above.
If
you
once
you're
ready
scroll
up
to
where
you
can
see
the
0.3
there,
an
attendee,
a
tester
is
shown
as
consuming
attestation.
Results
shows
for
land
party
producing
evidence.
F
That's
exactly
the
same
as
the
other
issue
that
we
said
won't
fix
where
it's
the
intermediary
that
caches
or
relay
stuff.
Okay,.
D
F
That's
one
where
it
should
be
clear
in
context
if
you've
understood
the
previous
section.
If
you
don't
understand
the
previous
section,
and
you
asked
the
comment
that
we
said
we
weren't
going
to
fix,
then
this
section,
if
you
look
at
it,
it
doesn't
repeat
the
sentence,
and
so
we
could
actually
repeat
a
sentence
in
there.
I
don't
know
if
we
need
to,
but
if,
if
other
people
thought
it
would
be
useful,
then
I
would
be
fine
adding
it.
F
So
if
you
got
to
like
example,
five
and
you
can
see
where
evidence
goes
through
the
airline
party
and
it's
exactly
the
same
as
the
as
the
background
check
picture,
we
talked
about
last
time,
but
the
text
below
it
doesn't
repeat
that
sentence
we
had
up
above
and
so,
if
anything,
we
could
repeat
the
sentence
down
here
so
down.
B
F
B
E
So
you're
writing
into
an
issue
right
now.
Yeah.
F
F
F
Well,
you're,
working
on
that.
I'm
looking
for
some
place
in
16.1
to
put
the
equivalent
for
the
passport
model
sentence,
which
I
would
probably
do
in
the
same
purpose,
but
to
see
okay.
B
F
B
F
Five,
you
could,
we
will
want
to
do
something
assuming
we
accept
this
and
think
this
is
a
good
idea
to
add,
and
we
want
to
do
the
same
thing
in
section
6.1,
with
passport
sentences.
E
Just
the
stations
are
the
same
way
you
mean.
F
There's
a
we
already
have
a
sentence
back
there
about
passport,
because
this
is
one
with
the
I
test.
Yeah.
The
attestation
result
is
cashed
by.
E
Impact
you
eric
about
the
question
of
evidence
and
he
was
like
sure
why
not
but
treat
the
debbie
dragons.
You
know
freshness
will
be
hard,
but
you
could
catch
evidence
for
a
definite
amount
of
time
dealing
with
a
interesting
challenge,
how
to
establish
freshness,
emotions
but
possible.
It
is
yes.
F
B
I
would
say,
after
this
this
between
the
second
and
third
paragraphs.
F
F
B
B
B
Remind
myself,
okay,
all
right,
so
editorial
comments,
so
I
I
puzzled
through
this
one
and
yesterday,
actually
I
I
went
to
find
rfc
4949,
section
2.6,
and
this
is
the
this.
Is
this
the
text
from
4949
and
I'm
confused,
because
I
think
we're
doing
exactly
what
4949
says
we're
supposed
to
do
so.
B
F
So
here
is
questions.
What
does
compare
mean-
and
you
quote
it
down
below
some
entries
containing
a
mark
of
the
forum.
Compare
where
x
is
a
list
of
terms
that
are
either
antonyms
or
different.
Some
other
manner
worth
noting.
F
And
you're
arguing
that
we're
using
it
in
exactly
the
same
sense
looks
like
roman
is
saying:
well,
can
you
say
more
about
how
it
differs
and
I
think
you're
saying
well,
49.49
doesn't.
B
B
C
Yeah,
I
think
I
think
he's
taking
an
exception
to
the
the
way
that
it
reads
and
that
it's
that
it
doesn't
parse
in
english,
and
so
I
think
the
solution
to
his
comment
is
to
make
sure
that
it
parses
properly
in
english
and
don't
worry
so
much
about
what
it
says
in
the
reference,
because
reference
is
there
the
compares
there.
Well
I
mean
let's,
let's,
let's
just
critique
the
sentence
itself.
C
A
A
B
B
F
B
C
B
B
C
C
C
I
I
think
the
solution
here
is
to
say
something
right
before
to
say
that
what
are
we
doing
and
then
then
say
it
is
a
compare
as
in
the
rough
in
the
reference
here
right.
So
so
so
this
again,
the
objection
is
to
that
person
who
has
not
read
that
and
is
reading
this
document
and
has
stopped
this
track.
Saying
I
understand
english,
but
I
don't
understand
what
the
sentence
in
this
document
means.
C
F
Would
argue
that
I
don't
think
that
it's
grammatically
correct
as
it
was
before
right,
it's
phrased
as
an
imperative
and
it's
correct
grammar
there's
nothing
incorrect
about
it
in
in
the
old
stuff.
I
I'm
fine
with
matching
the
precedent
or
rewarding
it
or
whatever.
I
just
don't
think,
there's
anything
wrong
with
the
way
it
was
before.
E
G
E
No,
no,
I
think
it's
a
nonsensical
argument,
but
what
I
think
is
what
our
woman
wants
us
to
do
is
to
reason-
and
I
think
the
reasoning
has
to
be
in
written
words.
What
is
of
interest
there?
The
the
the
pointer
is
too
abstract
and-
and
I
think
that's
his
problem-
there's
no
syntactical
content
problem.
Here
we
have
to
pull
over
the
part
or
the
the
meaning,
at
least
from
rbc
4949,
that's
interesting,
and
we
should
have
to
look
at.
I
think.
F
F
The
syntax
is
the
same
as
this
in
4949.
Wouldn't
you
have
exactly
the
same
comment
when
reading
4949
whenever
it
uses
the
word,
compare
so
like
on
the
screen
where
it
says:
reliability,
yeah,
compare,
availability
and
survivability.
What
do
you
mean?
Why
does
it
compare?
What's
the
difference,
I
mean.
Why?
Wouldn't
you
have
exactly
the
same
command
here?
That's
not
the
precedent
I
mean
it
seems
like
roman
doesn't
like
49.49.
B
F
In
many
ways,
so
if
we
do
it
exactly
the
same
way
as
4949,
I'm
saying
it's
defensible,
you
can.
C
It's
else
dangerous
kind
of
mode
going
forward
in
any
document.
To
just
basically
say:
hey
all
rfcs
that
came
before
just
because
it
passed
muster
from
the
current
reviewers
is
the
way
to
do
it,
and
so
I
I
think
the
comment
really
is
about
the
idea
that
you
can't
read
it
and
make
any
understanding
of
it,
and
so
I
I'm
not
sure
exactly
what
the
fix
is,
but
I
think
it's
to
make
the
compare
sentence
longer
not
just
put
a
colon.
The
colon
doesn't
fix
the
problem.
It's
not.
F
Yeah,
if
you
think
you
can
do
it
in
a
way
that,
because
remember
you
know
you
and
I
were
arguing
minimal
changes
or
whatever.
If
you
can
suggest
text
that
wouldn't
have
to
reset
working
group
consensus,
then
please
do
so.
You
can
remove
it.
We
can
change
punctuation
or
we
can
extend
additional
words
with
meaning
that
would
make
reset
consensus.
So
what
you
suggest.
B
So,
let's
just
let's
just
compare
relying
party
from
4949
and
what
we
write
right
so
here
it
says
it's
a
synonym
for
certificate
user,
which
I
would
agree.
It's
compare
because
we're
our
term
relying
party
is
really
nothing
to
do
with
the
with
the
it's,
not
a
certificate
user.
It's
not
a
synonym
for
certificate
user,
something
much
more
complicated.
B
B
E
C
E
C
I,
I
think,
that's
that's
the
right
idea,
because
that's
what
that's?
What
the
point
of
the
compare
statement
is:
that's
why
it's
there,
and
so
the
the
the
syntax
of
the
compare
is
not
really
the
issue,
which
is
what
is
it?
What
is
the
information
that's
trying
to
convey
and
we
can
make
it
easier
on
the
reader
by
just
saying
what
what
the
hank
just
said.
B
A
A
E
Yeah
he
never
liked
this,
but
from
the
very
first
review
he
somehow
robbed
this
statement
and
we
were
not
able
to
capture
what
if
we
kept
saying
a
resolution,
I
think
this
is
going
into
the
right
direction.
F
C
Yep,
so
so
this
fix
definitely
tells
the
reader
what
to
expect
and
gives
them
a
means
of
solving
it.
So
problem
solved
in
that
sense
problem
not
solved
in
the
sense
that
it
forces
them
to
go
to
the
other
definition
and
to
reason
about
what
the
difference
is,
rather
than
less
telling
them
what
that
difference
is,
and
so
it's
more
work
to
do.
The
latter
and
the
group
should
decide
whether
we
do
that
or
not.
G
C
D
C
But
but
if
we
really
think
it's
a
problem,
so
I
I
think
you
just
throw
it
to
the
group.
What
do
we
think
should?
Is
it
enough
to
go
with
this
fix
and
leave
it
to
the
reader
to
follow
the
reference
and
say
yeah?
It
is
different.
I
understand
what
they're
saying
and
it's
different
what
they
said,
or
is
it
important
to
have
the
no
misunderstandings
and
to
have
us
decide
what
that
difference
is.
F
So,
even
if
we
just
take
the
example
that
michael
had
up
on
the
screen
about
the
definition
of
relying
party,
I
think
any
comments
about
the
4949
version
is
fraught
with
rat
tools,
because
you
get
into
discussions
of
what's
a
certificate
and
what's
a
user,
and
none
of
those
are
central
to
the
point
of
of
what
we're
trying
to
talk
about
the
rat's
document.
F
B
C
B
Is
please
please
suspend
your
4949
definitions.
B
A
D
B
Unlearn,
you
must
unlearn,
you
must.
Yes
exactly
so.
A
A
So
I
think
what
we
just
said
is
true
for
both
verifier
and
relying
party
in
the
case,
which
is
that
we're
we're
not
we're
not
changing
those
definitions,
we're
building
on
top
of
them.
In
the
case
of
relying
party
the
the
component,
you
know.
The
thing
that
comes
up
most
often
in
my
experience
is
that
the
relaying
party
is
more
than
just
what
we
describe
as
the
role
that
as
it
pertains
to
processing
of
attestation
information.
D
D
E
C
Well,
if
the
reference
says
reliant
party
is
just
a
certificate
user,
it's
so
much
more
than
that,
and
in
fact
it
doesn't
even
have
to
be
a
certificate
user.
All
right,
yeah.
You
can
certainly
figure
use
cases
where
there
are
not
certificates
involved,
and
you
know
relying
parties
deal
directly
with
evidence,
that's
put
to
them
and
make
their
own
decisions.
C
B
Ten
years
you
know
we'll
have
only
five
different
kinds
and
we'll
be
able
to
describe
them.
As
you
know,
banana
apple.
You
know
some
other
abstraction
thing
that
there'll
be
five
of
them
and
if
I
say
that
it's
a
it's,
a
a
grapefruit
attestation
system
that
everyone
will
know
what
it
is
at
that
point,
I
I.
C
B
C
Point
you
still
will
have
variation
in
the
actions
you
take
at
a
relying
party
and
how
you
interpret,
of
course,
you
think
of
attestation
as
a
statement
of
health
there's
different
reasons
to
have
different
health,
and
I
think
we'll
have
a
tougher
time
enumerating
what
those
are
in
those
side.
But
as
far
as
the
mechanisms,
I
think
the
industry
will
settle
out,
which
ones
work
in
the
context
of
the
kinds
of
things
it's
doing,
and
you
know
and
the
story.
F
There
is
a
wording
tweak,
I
just
posted,
I
hope
to
this
issue.
Yes,
I
think.
D
B
I
have
some
some
dirt
on
my
screen
that
turned
reference
into
reference
with
an
h
rather
than
an
n,
but
it
was
dirt
on
my
screen.
B
I'm
gonna
close
that
for
now
repeat
that,
maybe
I
can
close
this
and
this
and
issues
okay.
So
the
last
one
is
topological
patterns.
What
did
he
write?
Okay,
so
I
think
we've
dealt
with
this.
We
added
some
sentences
to
it
already,
I'm
just
not
convinced.
We've
dealt
with
all
of
the
comments
there.
F
B
We
we
we,
we
found
that
we
already
had
sentences
on
them
yeah,
and
so
my
email
pointed
out
that
we
already
said
that,
and
you
just
didn't,
read
the
text:
okay,.
B
And
we
copy
those
some
of
those
sentences
to
the
appendix
now.
Yes,
so
I
think
this
is
truly
dead,
but
I
feel
it's
nebulous
enough.
B
F
F
C
E
That
be
resolved,
okay,
so
the
problem
here:
okay,
that's
gone,
okay,
yeah
and
the
resolution
is
that
we
assume
this
is
still
worth
and
we
probably
have
to
go
through
another
round.
E
He
has
to
highlight
in
a
more
illustrative
way
what
what
is
what
an
actual
problem
is
with
it.
So
so
the
question
the
main
question
was:
what
does
an
implementer
take
away
from
this
and
I
would
say
the
whole
guiding
principle
of
this.
E
A
A
E
Yeah
and
changes
the
fundamental
definitions
of
the
roles
which
say,
and
if
dave
dave
made,
that
very
clear
at
the
very
beginning
of
this
restructuring,
yeah
that
every
row
is
specifically
consumed,
some
producers
specific
conceptual
messages
and
that
never
changes
whatever
the
examples
are
whatever
the
compositions
are.
In
the
end,
the
attestation
result
will
arrive
at
the
airline
party
and
will
be
consumed
there
not
earlier
later.
C
So
here's
a
question:
you
know
the
complication
really
comes
when
you
start
talking
about
combining
roles
with
each
other
and
putting
them
away
in
ways
that
don't
look
just
like
the
diagrams
we
we've
done.
Have
we
done
a
good
job,
saying
that
that
kind
of
thing
is
architecturally
possible,
but
don't
get
hung.
A
C
So
then
I
I
I,
if
that's
the
case
and
that
that's
kind
of
what
our
responses
need
to
be
is
that
you
know
there
is
a
complication
in
architecturally
and
when
you
go
to
implement
this,
what
you
really
need
to
do
is
think
about
where
these
roles
are
and
what
the
issues
are
of
the
roles,
and
you
know.
Furthermore,
what
we
might
want
to
say
is
that
combining
them
in
certain
ways
brings
about
other
issues
that
we
may
not
have
written
about,
and
you
really
need
to
think
about
what
happens
when
they're,
not
distinct
densities,.
F
F
F
C
Use
cases,
but
when
you
do
this,
you
need
to
think
about
all
the
issues
right
and
that
that's
that's
kind
of
like
a
warning,
an
architectural
warning
that
you
know
all
of
them
are
not
created
equal
as
far
as
things
like.
Can
you
get
a
trusted
result
item.
A
A
C
E
C
E
Yeah
maybe
write
a
document
about
it,
but
I
know
that
can't
go
into
the
text
I'm
just
kidding,
but
but
I
assume
that
other
solutions
that
might
be
more
complicated
would-
and
there
are
others
already
out
there-
in
the
working
corporate
thing
or
even
operating
and
so
yeah.
We
have
another
role
and
we
have
another
interaction
model.
C
That
statement
says
a
lot
about,
I
mean
the
freshness
we
were
just
talking
about
and
the
ability
to
cash
stuff.
You
know,
where
is
that
cash
for
different
there's
different
places
where
caching
makes
sense,
and
that
has
a
different
impact
on
trust,
and
I
mean
it's
actually
a
much.
C
C
So
but.
E
C
E
E
B
A
E
C
B
B
E
E
Did
that
sorry,
I
was
sorry
yeah
yeah.
Do
we
want
to
have
more
external
review,
or
do
we
want
to
push
this
without
other
area
director?
You.
E
Oh,
that's
all
it's
just
so
that
can
relay
this
question.
E
E
E
Okay,
so
so,
okay,
yeah,
that's
fine!
I
I'm
just
I'm
yeah,
okay,
okay!
Let's
do
it
this
way,
it's
way
late,
anyways!
Let's.
B
Will
post
the
17
as
in
an
hour
or
less
and
answer
write
another
reply
to
roman
and
say
I
think
we're
done:
okay,
michael
okay!
Thank
you.
Bye,
bye,.