►
From YouTube: RATS Architecture Design Team, 2021-04-09
Description
RATS Architecture Design Team, 2021-04-09
A
A
D
A
B
A
That
kind
of
suspect,
but
apparently
the
some
of
the
orders
of
magnitude
stuff,
are
correct.
It's
just
what
is
it
stored
in?
I
don't
know,
but
it
does
make.
You
think
of.
I
don't
know,
makes
me
think
of
doctor
who
yeah
right,
because
there
was
that
guy
adrick
who
could
do
math
in
his
head
and
cause.
You
know
things
to
appear
anyway,
dave
you
are
here
andrew
you
are
here
hank.
I
have.
A
Yes,
hank
says
he
wants
the
link,
it's
the
same
one
as
always
hank.
Let
me
give
you
the
link.
A
A
F
A
As
we
discussed
yesterday,
google
calendar
was
my
kool-aid
that
made
me
join
google,
and
now
it
doesn't
work
very
well.
So
I
guess
that
means
I
should
leave
okay
and
I
don't
think
we'll
probably
won't
get
way
on
friday
night
at
11
p.m
at
night,
but
maybe
they
work
saturdays,
a
lot
there.
It's
not
a.
A
The
workers
paradise
somehow
doesn't
have
a
lot
of
vacation
days
anyway.
Okay,
so
we
have
seven
of
us
here.
Peter
doesn't
have
his
mic
on
yet
maybe
connect
there.
He
is
peter
okay.
So
thank
you
to
everybody
who
took
up
the
assigned
issues
and
we
have
nine
pull
requests
to
go
through
today.
A
I
don't
know
what
I
think
this
is
the
lowest
hanging
fruit,
but
no
one
else
has
reviewed
it
yet
so
anyone
want
to
nominate
one
that
they.
A
A
Okay,
I
oh,
I
had
reviewed
it,
oh
yeah,
so
it
won't.
Let
me
approve
unless
with
changes,
unless
I
tell
you
the
change
and
I
already
wrote
the
change
so.
D
Yeah
on
your
comment
there,
I
will
accept
your
change
as
long
as
you
change
n
to
a
a
tamper.
B
D
D
A
D
G
D
A
D
A
separate
issue
that
this
one
does
not
address
that
penguin
and
I've
been
commenting
on,
which
is
the
first
one
on
the
whole
list
and
it's
easy
to
without
flipping.
If
you
scroll
up
just
to
sorry
scroll
down
just
a
little
bit,
so
you
can
see
more
of
the
picture.
D
Okay,
the
issue
there
that
I
will
address
in
a
different
pro
request.
It
was
observed
now
that
it's
actually
clear
that
b
is
boot.
Loader
right,
you
can
see
a
line
coming
out
of
the
attestory
environment
and
going
up
to
the
verifier.
That's
labeled
layered
evidence
for
boot,
loader
and
kernel
right.
So
it's
observed
that
boot
loaders
usually
can't
communicate
directly
to
the
verifier
right,
and
so
that
line.
D
Needs
to
come
upwards
through,
say,
kernel
and
then
come
up.
Instead
of
coming
directly
out
the
side,
it'll
probably
be
changed
to
be
going
upward
vertically
up
through
the
kernel
and
things
because
it
would
have
to
be
passed
on
up
to
the
kernel
and
then
either
the
kernel
or
an
application
would
have
to
be
the
one
that
actually
sends
it
or
whatever
it's
just
signed
by
the
testing
environment.
So
anyway.
So
that's
in
a
different
issue.
I
won't
do
it
in
this
one,
but
I'll
do
it
when
addressing
the
separate
issues.
Okay,.
C
Is
that
going
to
cause
confusion
where
we've
tried
to
avoid
circularity
and
trust.
A
I
I
and
I
I
would
concur
that
that
this
doesn't
have
to
be
physicality.
It's
it's
it's
the
passage
of
information.
I
would
say
that
better
is
to
put
some
kind
of
a
note
right
around
here
that
says
boot
loaders,
clearly,
can't
probably
don't
have
network
stack,
so
it
needs
to
be
conveyed
through
another
layer,
physically
conveyed
through
another
layer,
and
that
just
acknowledges
that.
But
but
there
could
be
other
layer
a
right
or
b
in
other
scenarios
where
there
actually
is
a
path
right.
D
D
Yeah
293
and
then
scroll
to
the
bottom
of
there,
where
you
can
see
the
comments
we're
talking
about
that
yeah
there
and
then
just
add
a
comment
there.
Just
saying
that.
D
G
G
D
We're
talking
about
just
adding
text
to
address
that,
but
not
changing
figure.
Three,
that's
enacted
that
may
mislead
and
we
would.
We
would
do
that,
not
by
changing
the
figure,
but
by
putting
text.
That
explains
your
your
point
and
different
ways
that
it
could
be
addressed.
G
Okay
and
that's
done,
but
I
have
a
question
that
if
we
use
this
layer
concept,
so
the
evidence
should
be
protected
and
it
should
be
protected.
The
configuration
integrity
and
the
confidential
validity
is
that
right.
D
G
A
Okay,
let's
go
to
this
issue,
I'm
just
gonna.
You
know
pick
one
top
one
from
each
person
and
we'll
come
around
yeah.
D
D
D
C
A
To
tip
arch
and
tcg
arch
as
references
for
it
here
and
trusted
execution
environment,
that's
so
non-controversial,
I
think,
just
go
any
objections.
C
D
There,
probably
is
one
from
global
platform,
but
I
have
to
do
some
searching.
Did
you
look
around
the
global
platform
website.
A
I
I
I
don't
think
it
was
a
question
of
so
is
the
document
public.
I
prefer
citing
things
that
are
public,
because
I.
A
A
A
A
D
It's
gone
through
last
call:
it's
got
its
shepherd
assigned.
I
think
that's.
D
A
So
so
I
just
I
I
I
thought
the
point
as
I
was
trying
to
say
is
I
I
thought
that
this
document
was
ahead
of
us
in
the
queue
and
that
we
were
likely
to
wind
up
referencing
the
rfc,
because,
but
you
know
so,
I'm
literally
there's
literally.
I
have
a
document,
an
auth
for
it.
I
have
two
documents
in
auth
48,
but
one
of
them
doesn't
have
it
isn't
quite
in
authority,
one
references,
the
other
informatively
and
I'm
just
like.
D
That
tea
park
has
actually
gone
through.
Two
working
group
last
calls
lecture
thing,
so
it's
actually
slightly
ahead.
A
Still,
okay
anyway,
I
I
this
will
be.
They
automatically
become
references
to
the
rfcs,
assuming
it's
published
unless
the,
unless
the
that
that
that
one
document
is
still
in
also
in
process,
in
which
case
they
don't
create
new.
You
know
dependence
dependencies,
but
it
is.
It
is
unfortunate
because
it
doesn't
wind
up
as
a
citation
of
the
rfc
in
that
case,
which
is
unfortunate
for
that
rfc
anyway.
It
we're
merging
this.
I
think,
there's
no
conversation.
A
Yep
thanks
ned,
no,
no
conflicts
confirm
merge.
E
C
Unless
we,
you
know.
D
F
F
Ask
jeremy
I
mean
he
should
know.
F
A
Okay,
all
right,
let's
go
on
to
one
by
hank.
F
Is
this
one
addressing
that
this
was
about
conveyance,
mostly
and
not
about
the
implications
of
trust
model?
Actually,
and
so
I
you
said
this
is
about
security
uccs,
so
you
should
do
that.
F
D
A
Lowercase,
so
the
other
joke
is
that
the
government
of
canada
spent
over
a
billion
dollars
on
something
called
secure
channel
in
the
decade
ago
and
took
a
a
10
million
report
to
even
figure
out
what
they
had
bought.
A
A
Yeah,
so
that
line
looks
fine
okay,
so
I
rewrote
this
984.
I
found
it
really
awkward.
I
removed
some
words
and
I
hope
it
preserves
this
the
sense
with,
while
being
a
little
bit,
I
say,
less
convoluted.
I
don't
feel
strongly
about
it.
F
Yeah,
it
was
a
long
one.
I
have
to
admit
that,
but
also
I
only
had
like
I
don't
know,
half
an
hour.
F
So
what
you
I
like,
I
can
live
with
a
possessive
detonation
of
terms
being
removed,
although
I
think
don't
think
it's
about
the
trusted.
Verifier
you're
really
trusting
the
verifiers
capabilities
here
and
then
I
am
fine
with
renders
and
unprotected
again
you
don't
know
which
evidence
you're
not
talking
about
the
sentence
before
also
I
was
talking
about
just
simply
evidence
so
qualifying
it
felt
less
ambiguous
to
me
and
at
the
at
the
end.
I
don't
know
even
if
the
sense
it
since
it
makes
sense
with
removing
the
tail.
E
A
So
the
tail
I
removed-
because
I
thought
it
always
was
the
case
that
it's
always
believable
at
the
time
of
attestation.
In
other
words,
that's
not
new.
The
fact
that
the
the
that
the
evidence
has
to
be
believable
at
the
time.
Well,
that's
always
the
case
and
there's
nothing
to
do
with
whether
the
the
the
uccs
part
right.
It's
always
the
case
it
yeah.
I.
D
I
I
I've
now
read:
I
like
michael's
wording
better
other
than
it
needs
a
grammar
fix.
It
needs
the
word
that
in
between
trust
and
the
verifier,
but
otherwise
I
I
find
the
meaning
the
same
and.
F
D
F
Important
that
the
ucc
that
the
evidence
is
it
is
does
not
matter
that
it
is
tampered
with
after.
F
D
F
F
A
Yes
make
sense
of
it,
so
I
just
assumed
that
was
sort
of
contained
in
the
concept
of
freshness
which
we've
dealt
with
elsewhere,
yeah,
okay,
and
that's
why
I
felt
it
was
just
a
little
bit.
The
extra
that
extra
run-on
part
makes,
I
think-
and
maybe
the
non-english
speakers
in
the
room
will
tell
me-
makes
me
think
that
there
is
something
else
here
that
I'm
missing
that
isn't,
that
is,
that
is
subtle
and
somehow
I've
missed
it,
and
so
I'm,
but
I
don't
think
that's
the
case.
A
We're
just
talking
about
freshness,
and
I
think
the
major
part
about
this
is
that
an
independent,
because
the
evidence
was
not
signed.
The
audit
log
of
the
verifier
is
not
as
good
quality
if,
if,
when
it
logs
the
evidence
right
because
it
could
have
have
messed
with
it
before
it
logged
it.
C
D
Model
section
right,
and
so
your
points
there
are
not
about
the
core
trust
model
per
se
other
than
the
the
fact
that
it's
abstracted
says:
you're
trusting
the
verifier
is
going
to
do
the
right
thing
right
doing.
The
right
thing
is
elaborated
in
other
sections.
C
A
Right,
so
all
I'm
trying
to
say
is
that
I
don't
think
there's
a
nuance
intended
by
this
other
than
the
things
we've
already
said,
but
that
extra
little
bit
sort
of
tells
me.
Oh
there's
some
nuance
that
I've
missed
and
I
don't
know
what
it
would
be
and
that's
why
I
feel
we
should
remove
it,
because
I
don't
think,
there's
a
nuance
here
that
we're
missed.
That
were
that
I
don't
think,
there's
a
nuance
for
untrusted
or
unsigned
evidence
that
is
different
in
this
case.
Yeah.
D
Line
982
uses
the
term
via
a
trusted
conveyance
channel,
whereas
the
earlier
line
says
via
a
secure
channel.
I
would
like
us
to
use
the
same
term
between
those
two
sentences,
whichever
one
we
pick.
F
I
did
not
try
to
mess
with
the
trusted
conveyance
channel
because
I
think
it
is
used
in
other
texts
places
also.
G
C
E
F
Yeah
and
then
the
uccs
document,
it
is
yes.
D
F
F
F
That
is
potentially
correct,
but
I
would
okay.
That
is
true.
Without
that
knowledge,
you
have
to
read
the
other
document
with
that.
The
three
terms
words
here
or
four
you
can,
you
can
maybe
make
people
intuitively
get
the
context
better.
That's
why
I
edit
it,
but
I
don't
insist
on
adding
it.
You
can
still
make
people
go
to
the
other
document.
Of
course,.
D
F
D
F
F
D
D
Then
just
a
question:
this
is
not
a
request
for
change
in
982.
It
says
a
trusted
conveyance
channel.
Is
there
a
difference
between
a
trusted,
conveyance
channel
and
a
secure
conveyance
channel.
D
F
D
A
Okay,
okay,
so
let's
go
to
this
one
from
me.
A
A
A
D
That's
just
explaining
why
I
think
the
pr
should
be
merged
and
this
issue
should
not
be
closed.
Okay,
because
there's
a
related,
the
alternative
is,
I
could
follow
a
new
issue
which
yeah
figure
that
simple
comment
at
the
bottom
of
that
one.
So
this
one
does
literally
what
hank
was
complaining
about,
but
there's
a
related
closely
issue.
That
means
there's
probably
a
additional
change.
I
should
make
okay.
A
D
So
tr
the
term
trusted
appeared
in
two
places,
one
up
in
the
architecture
section
and
then
one
down
to
the
rules
section.
So
in
the
architecture
section
you
can
see
the
451
red
and
all
I
did
there
was-
I
removed
it
and
replaced
it
by
a
forward
reference
to
the
verifier
section.
Okay,
because
the
two
ways
to
establish
it
was
already
discussed
in
the
verifier
section
so
rather
than
explaining
it
twice.
I
just
did
a
forward
reference
to
the
place
that
had
the
longer
elaboration
about
that.
D
So
that
was
the
change
at
451
is
just
replacing
the
trust,
anchor
store
sentence
with
a
forward
reference
and
all
the
rest
was
in
the
verifier
section
and
the
verifier
section
had
a
bunch
of
discussion
of
trust
anchors
more
than
just
what's
on
the
screen,
there's
more
up
down
below
that.
You
don't
need,
and
so
I've
had
to
explain
it.
The
first
use
okay
by
so
the
first
use
was
in.
D
I
don't
remember
if
it's
in
950
or
it's
probably
right
below,
where
it
expands,
and
so
I
put
the
definitions
up
right
before
the
use,
and
so
the
paragraph
that
starts
at
9
41
is
the
paragraph
that
is
literally
copied
out
of
the
tp
architecture
document
and
the
changes
before
and
after
the
paragraph
are
to
make
it
read
in
context.
D
D
A
Right
so
it's
also
in
4949,
and
I
don't
know
if
that's
a
better
definition,
that's
trust.
Yeah,
okay,
just
anchored
not
well,
it
might
have
stored
it
further
down.
A
E
B
A
D
And
it
became
useful,
so
I'll
explain
it
once
I
did
this,
then
I
realized.
So
this
is
touching
the
places
that
use
the
term
trust
anchor.
Okay,
yes,
the
reason
I
don't
want
to
close
it
is
what
I
noticed
is
that
the
relying
party
section
appears
right
before
the
verifier
section
has
very
similar
wording,
but
does
not
use
the
term
trust
anchor,
and
so
what
I
want
to
do
is
I
want
to
move
the
definition
up
in
the
relying
party
after
I
make
the
language
match.
D
So
ryan
party
needs
to
have
similar
wording
because
it
has
a
trust
anchor
store
too,
and
then
I
noticed
that
the
security
considerations
section,
which
has
trust
anchor
store
information,
the
cheap
architecture
document.
We
don't
have
anything
any
place
that
I
could
put
that.
I
could
copy
that
right.
There's
nothing
in
the
in
our
security
consideration
section
that
has
verifier
considerations.
D
We
only
talk
about
the
a
tester
considerations,
and
so
I
would
want
a
place
in
the
security
considerations
to
add
another
little
subsection
about
considerations
on
the
verifier,
slash
relying
party
and
that's
where
I
would
copy
the
trust,
anchor
store
considerations
out
of
the
team
architecture
document.
So
that's
what
the
two
things
that
do
in
a
separate
pull
request.
D
D
Also
referred
to
in
and
then
at
a
security
considerations,
paragraph
that
talks
about
securing
the
trust,
anchor
story,
it's
like
one
paragraph
or
something
that
I
just
need
to
copy
in
the
right
text,
but
all
those
were
not
fixing
the
original
issue.
It's
now
going
to
once
you
merge
this
one,
it
will
be
clear
what
a
trust
anchor
store
is
yeah.
So
I
think
it
fixes
hank's
point,
but
I
figure
we
can
just
use
the
same
issue
so.
A
Okay
works
for
me
all
right,
so
that's
done.
D
D
A
Some
duplication
right
so
in
this
one
we
assigned
it
to
you.
So
it's
my
bad,
not
yours!
So
in
this
pull
request,
dave
moves,
appraisal
policy
and
reference
values
down
to
here
and
then.
D
Notably
you
see,
one
difference
is
the
order,
so,
if
you
scroll
down
you
can
see,
I
did
it
with
as.
D
D
D
D
D
We
can
look
at
ned's
because
he
made
another
change.
It's
not
part
of
the
reordering,
which
I
at
least
wanted
to
discuss
so
okay.
Well,
let's
do
that.
A
D
A
I
did
was
the
move,
and
so
ned
did
more
than
what
I
did
you
you
just
moved
the
text
and
never
changed
is
what
you're.
D
Saying
correct,
that's
correct
other
than
the
only
thing
that
I
changed
in
mine
was
the
one
sentence.
That's
green
389
right
and
I
think
ned
has
a
different
version
of
that.
So
just
notice
the
the
my
phrasing
389,
I
think
ned
has
a
longer
piece
of
text.
There.
D
D
Think
this
is
a
change
that
I
said.
Ned
did
that
I
didn't
see
that
this
one
was
related
to
the
moving
of
text
here
and
to
me
this
changed
a
meaning,
and
so
this
was
the
part
that
I
wasn't
that
happy
about.
Even
if
I
liked
the
rest
so.
D
You
can
see
previously.
There
was
three
possibilities
in
ned
screen,
there's
four
possibilities
of
which
I
claim
the
first
two
are
the
same
thing,
so
the
adding-
and
I
guess
he
changed
might
be
to
is
in
two
of
the
cases,
and
so
I
just
didn't
understand
the
reasons
for
the
change
for
this
one.
So
I
think
the
the
red
text
was
fine.
I
don't
have
a
problem
with
changing
if
I
understand
what
the
problem
is,
but
I
just
looked
like
it
was
to
me.
I.
A
D
A
D
C
G
C
C
D
That
was
a
very
different
meaning
386
is
is.
I
was
going
to
ask
you
about
this
one
too,
because
these
are
two
different
things
and
it's
possible.
The
386
is
not
worded
right.
This
means
programmed
into
the
relying
party
by
you
know,
hard
coded
by
the
relying
party
developer.
So
385
is
it's
configured
by
the
relying
party
owner
386
as
it
comes
from
the
relying
party
developer.
Those
are
two
different.
C
Entities,
but
that
would
be
wrong
if
it's
programmed
by
their
by
the
developer.
If
the
developer
was
not
working
in
under
the
under
the
delegation
of
the
owner.
D
Sure,
no,
it's
not
wrong,
because
here
there's
use
cases
where
people
ship
or
lying
parties
that
might
be
headless.
You
know
iot
devices
or
something
like
that,
and
they
are
hard
coded
to
have.
The
verifier,
for
example,
be
the
manufacturer
of
the
device
which
is
also
the
the
programmer,
and
so
then
they
sell
it
to
somebody
who's,
then
they're
buying
party
owner
their
lane
party
owner
does
not
get
to
override
the
verifier
right.
The
relying
party
says
myverifier
is
the
manufacturer
and
sorry
you
can't
change
that
yeah.
So.
C
D
It's
a
nuance
right
that
you
could
also
take
the
view
that
that's
not
the
case
where
there's
still
a
relying
party
owner,
that's
separate
from
the
manufacturer.
They
just
don't
get
to
decide
who
the
verifier
is.
They
get
decide
other
things,
but
they
don't
decide
who
the
verifier
is,
and
so
they
can't
affect
the
appraisal
policy,
but
they
can
affect
lots
of
other
things.
C
Right
so,
rather
than
you
know,
we
could
write
a
much
longer
document
about
about
multiple
ownership
use
cases,
but
it's
probably
inappropriate
for
this.
At
this
time,
owner.
D
Can
change
anything?
That's
configurable,
just
saying
mayor
the
on
the
relying
party,
your
attestation
results
may
or
may
not
be
configurable.
Unlike
the
verifier
you're
allowing
parties,
the
appraisal
policy
for
attestation
results
might
simply
be.
Did
the
verifier
give
it
a
thumbs
up
or
thumbs
down
right
and
no
more
configuration
and,
if
you're,
in
a
small
iot
device
that
might
be
fairly
common
to
have
very
limited
sets
of
functionality.
C
So
the
right
so
this,
but
we're
in
the
context
of
describing
what
and
what
the
appraisal
policies
are
right,
and
so
it's
confusing
to
say
the
owner
of
the
policy
can
do
all
of
these
things,
except
for
when
the
manufacturer
does
something
different
and
then
it's
still
considered.
You
know,
though,
it's
still
considered
appraisal.
D
C
D
D
D
D
A
It's
just
the
fix
is
not
clear,
but
the
movement
of
the
text.
The
other
text
is
good,
so
the
question
is,
what
can
I
do?
Can
I
I
guess
I
can
put
another
commit
on
top
of
this,
that.
A
D
What
do
you
want
to
do?
Well,
let's
scroll
down,
let's
put
that
on
the
stack
for
just
a
second,
because
there's
two
possibilities
there.
If
you
look
at
this
part,
the
question
is:
do
people
like
what
ned
did
here
or
are
there
any
issues
that
anybody
else
has
down
here
right,
because
what
we're
trying
to
figure
out
is
net
head
of
different
ordering
of
sections
down
here
is
ned's
ordering
better
than
my
ordering,
which
is
sort
of
arbitrary
just
to
match
the
order.
D
The
things
were
mentioned
in
the
architecture
is
a
data
flow,
the
architecture
section
where
ned's
is
in
a
different
order
that
may
be
more
intuitive
or
something
based
on
how
you
read
it.
Do
people
like
this
one
better
are
the
original.
If,
for
some
reason,
people
like
mine
better,
then
what
one
could
do
is
one
could
use
mine
to
do
the
reordering
and
then
use
nez
to
do
the
discussion
that
we
were
having.
A
D
Or
the
other
at
this
point
yeah
so
down
to
391,
is
that
same
good
discussion
we
were
just
having
because
that's
related
to
the
might
be
obtained
or
is
configured
and
might
be
versus,
is
and
stuff
on,
not
the
register.
The
green
part
is
part
of
that
same
discussion.
D
C
Conceptual
messages
yeah.
C
So
when
you're,
when
you
the
result,
then
was
a
reference
values
section
that
had
one
sentence
in
it.
So
I
elaborated
on
that
to
add
more
clarity.
C
Yeah
and
then
appraisal
policies,
I
added
a
paragraph
at
the
end
to
add
more
context
to
it.
D
So
I
I
I
like
the
paragraph
under
conceptual
messages,
myself.
C
D
Okay,
that's
what
I
was
going
to
ask
is
because
I'd
put
the
appraisal
policies
in
between
reference
values
and
attestation
results,
because
it
uses
evidence
in
a
reference
value
since
used
to
generate
attestation
results
and
so
in
the
data
flow
diagram.
That's
why
it
appears
in
between
those
two
in
that
section,
because
it's
the
you
know
the
order
in
which
they're
used
right
yeah.
I
don't
feel
strongly
about
the
ordering.
C
D
Well,
I
should
say
we
correct
what
I
just
said,
because
there's
two
appraisal
policies
right:
one
of
them
is
used
in
between
there
and
the
other.
One
is
used
afterwards,
where
you
have
it
right
now.
So
if
you
think
about
the
appraisal,
policy
for
evidence
is
used
in
between
reference
values
and
attestation
results,
but
the
appraisal
policy
of
attestation
results
is
obviously
used
after
the
attestation
results.
So
right.
D
D
D
So
yeah,
I
I
don't
have
a
preference
on
this
one
because
of
the
covers
the
appraisal
policy
for
attestation
results.
Then
your
location
looks
fine
to
me.
Ned.
D
So
I
guess
I
like
ned's
changes
of
the
order
better
than
mine.
Now
that
I've
heard
this
and
I
like
ned's
change
around
the
conceptual
message,
paragraph
intro,
I
think
I,
like
my
one
sentence
forward
reference
from
the
section
that
stuff
moved
out
of
and
I
think
we're
still
discussing
the
other
text.
That
net
did
so.
I
don't
know
how
to
factor
it,
because
these
parts
of
deads.
A
I
would
like
to
merge
now
so
yeah,
so
I'm
I'm
just
trying
to
get
make
a
piece
a
commit
that
we
can
just
merge
that
part
and
so
no.
D
D
D
A
D
D
D
D
Okay
and
so
I'm
going
to
go
back
to
mine,
I'm
going
to
copy
the
one
sentence
out
of
it
and
then
abandon
it.
A
So
in
the
the
the
non-controversial,
so
I
think
I'm
supposed
to
keep
this
hunk.
So
I'm
supposed
to
keep
this
hunk
in
in
the
the
tech
in
the.
D
A
A
Is
that
that
right,
what
shall
I
call
that
pull
this
other
one
is
issue
325.
A
A
D
D
A
D
Okay,
I
have
closed
my
request
and
the
two
places
that
I
had
bits
of
text.
I
copied
them
into
comments
on.
A
D
And
then
I
just
wanted
to
be
able
to
see
side
by
side,
ned's
text
there
versus
my
text
there
I
said
I
was
fine
with
ned's
text,
but
I
don't
know
if
there's
anything
in
my
wording
that
people
like
there,
if
so
you
feel
free
to
you,
know,
merge
between
those
two
thing.
In
my
second
comment,
there.
A
I
don't
know,
I
don't
know
if
that
was
intentional,
that
was
in
the
in
the
in
there,
so
I
can
put
it
back.
A
I
think
yeah
probably
put
it
back,
okay,
I'm
putting
it
back
and
that's
what
I'm
doing.
This
is
my
editor
and
then
you
had.
I
missed
you.
D
A
D
The
reference
is
wrong
as
you're
using
things
you're
already
tagged,
yep,
okay,
so
reading
heads
read
my
and
see
if
you
want
to
just
ignore
mine
or
if
there's
any
of
my
text,
if
you
want
to
merge
into
nets.
A
And
then
your
other
comments,
ned's
other
comments
are
now.
I
hope
I
do
this
right.
A
D
That
is
the
only
thing
that
ned
did
differently
than
me,
and
I
wanted
any
comments
on
whether
there's
any
of
my
text.
You
want
to
do
here.
I
am
fine
with
merging
this
one,
but
if
people
like
any
of
my
texts
better,
you
might
want
to
put
this
into
the
same
pr
before
merging
it,
but
I'll.
Let
you
guys
decide
that.
C
D
C
A
In
the
meeting
all
right
yeah,
so
let's
come
back
to
this,
but
let's
I
would
love
to
yeah.
So
if
if,
if
give
us
some
reaction
to
here
and
then
I'd
like
to
close
I'd
like
to
merge
this
one
by
email
or
okay
sounds
good
all
right!
Okay,
thanks
michael
thank
you
bye-bye!
It's
great.