►
From YouTube: IETF-SCITT-20230522-1500
Description
SCITT meeting session at IETF
2023/05/22 1500
https://datatracker.ietf.org/meeting//proceedings/
A
A
No,
no,
it
got
changed
yeah,
but
the
short
story
is
we
tried
to
find
a
a
suitable
slot
for
everyone,
not
everyone
paid
attention
and
then,
when
I
announced
the
the
selected
or
preferred
date,
people
woke
up
and
said:
no,
no,
no,
that's
not
what
we
want.
D
Yeah,
it's
less
convenient.
This
thought
is
less
convenient
for
me,
but
I
was
more
concerned
that
I
had
screwed
up
my
calendar.
Okay,
because
there
were
some
problems
with
the
time
zones
with
the
upcoming
calendar
in
early
April
and
if
you're
using
Google
Calendar,
it
has
not
caught
up.
It
still
has
not
fixed
all
of
its
places,
really.
A
D
Yeah
and
that's
a
Google
Calendar
bug
that
unfortunately
yeah.
A
That
happens,
yeah,
okay,
yeah.
We
have
a
smart
group
today,
but
doesn't
matter
so
welcome.
Welcome,
anyway,
to
this
next
virtual
intro
meeting
for
the
skid
working
group.
As
usual,
official
intro
meeting
recordings
are
being
made,
the
all
the
IDF
rules,
Supply
IPR
rules,
Etc
we
last
week
we
talked
about
the
registration
policies
and
this
week,
as
I
distributed
in
the
meeting,
invite
in
the
updated
meeting
invite.
A
We
want
to
also
take
a
look
at
the
outcome
issues
and
maybe
make
some
progress
on
those,
or
we
can
also
look
at
the
PRS,
which
are
the
only
currently
three
PRS
but
they're
a
long
list
of
open
issues.
Unfortunately,
I
see
John.
Do
you
want
to
add
something
to
this.
E
Oh
yes,
thanks
Hannah,
so
one
bit
of
agenda
hacking
just
arose
from
email
list.
This
week
is
a
quick
discussion
about
anybody
here
who
knows
the
t-trends
but
they're
going
for
chartering,
and
it
looks
to
me,
like
the
key
transparency
folks,
really
just
want
to
be
a
use
case
on
top
of
skit
and
if
they're
not,
then
it's
something
like.
We
really
need
to
make
sure
that
the
Primitives
are
aligned,
because
it's
really
really
darn
similar,
so
yeah.
E
A
quick,
quick
chat
about
that,
because
their
Charter
is
in
the
discussion
phase.
Right
now,.
A
Yeah
thanks.
That's
that's
a
very
good
point.
B
I'm
having
a
little
trouble
hearing
John,
maybe
John-
can
can
talk
about
more
loudly
next
time.
Thanks.
A
Steve,
do
you
have
any
issues
that
you
would
like
to
I
know
you've
been
on
vacation
and
so
on?
So
probably
you
haven't,
spend
a
tremendous
amount
with
skit,
but
maybe
there's
something
because
you
created
a
number
of
the
open
issues
as
well
or
at
least
try
to
mix
well.
Actually,
you
did
create
many
open
issues.
C
Mean
I
just
copied
I
did
the
copy
paste
authoring
of
them
so
they're
not
but
I'm.
Looking
I'm
clicking
through
the
pr
requests
and
the
issues
now
I
I
guess
it's
hard
for
me
to
have
an
opinion
because
I
don't
know.
What's
transpired
over
the
last
couple
of
weeks,
I
had
literally
just
got
in
late
last
night.
So
what
I
see
the
Merkle
tree
proves
issue
45.
C
And
issues
I
mean
we
could
certainly
triage
the
items
in
there,
including
the
ones
if
they're
so
old.
You
know
that
the
stuff
that
we
copied
over
from
the
work
that
Hank
had
done
that
if
there
is
no
activity
on
them,
we
could
close
them.
C
A
Yeah
for
sure,
there's
yeah
there's
one
new,
very
new
open
issue
which
Cedric
created.
So
that's.
A
Set
Cedric,
since
we
are
looking
at
these
open
issues.
Obviously
yours
is
the
most
recent
one.
Do
you
want
to
say
something
about
this
specific
of
measures
URLs
for
Trans,
transparent
statements.
A
Want
to
share
this
your
screen
or
to
show
the
the
issue,
because
you
added
a
lot
of
text.
F
Yeah,
so
do
I
do
that.
So
let
me
try.
A
Yeah,
it's
the
not
the
sort
of
like
the
join
icon,
but
then
there's
the
pre-loaded
slides,
which
you
don't
have,
but
then
there's
the
share
screen
icon
on
a
bar
icon
bar
below
your
name.
That's
that's
where
you.
G
F
F
H
Okay,
are
we
just
looking
at
the
the
issue
screen,
so
it's
there's
a
URL
for
it.
Obviously,
yeah.
A
F
Okay,
that's
this
one!
Yes,.
F
Okay,
I
get
rid
of
this,
yes,
so
so
yeah.
So
it's
something
we
discussed
that
at
many
places
and
the
meetings
but-
and
it
should
be
relatively
simple,
but
in
fact
the
details
are
not
that
simple.
When
you
look
at
them,
so
the
the
purpose
is
to
find
to
identify
what
is
the
standard
way
to
refer
to
a
producer
register
statement
and
that's
something.
That's.
F
For
many
reasons,
but
in
particular
it's
very
important
for
registration
policies
as
soon
as
you
have
a
multi-statement
policies
and
it's
also
very
important
for
Federation.
As
soon
as
you
have
multiple
transparency.
F
So
please
have
a
look
at
the
notes,
so
I
stated
some
use
case.
Some
girls
are
some
proper
syntax
which
consists
of
a
locator
together
with
the
commit
smarter.
F
F
So
so,
when
you
used
to
say
that
this
discoverability
is,
in
which
case
just
the
opaque
and
from
the
hash,
it's
a
Quantum
based,
so
you
can
have
an
index
somewhere
else
that
actually
retrieves
the
transparent
settlement
based
on
the
hash,
no
matter
where
register
the
the
current
design
is
more
explicit
about
the
identity
of
the
transparency
service,
in
the
sense
that
this
is
the
official
locator
where
that
was
registered,
and
when,
where
you
may
be
able
to
query
it
if
needed,
the
difficulty
of
the
controversy
is
that
we
don't
want
to
require
that
the
transparency
service
be
able
to
Forever
on
the
rich
queries.
F
So
so,
in
some
cases
that
will
be
required
in
some
cases
that
would
be
too
expensive
or
limited
to
six
months
or
or
two
years
or
something
like
that,
and
so,
if
you're
concerned
with
that
design
is
that
people
might
interpret
the
locator
as
a
promise
that,
yes,
you
will
keep
the
transparency
statement
available
and
that
URL
and
again
it's
ideal
when
it
happens,
it's
not
something
not
too
strongly
around
there.
G
Thank
you,
honest
yeah
this.
This
is
an
issue.
We
also
experience
with
our
own
sag
Center,
where
you
have
a
single
shot,
six
hash
value,
but
there
are
different
types
of
statements
or
declarations
that
may
be
posted
into
the
registry,
and
so
we
ended
up
creating
essentially
What's
called
the
Declaration
type
and
that
that
may
be
what
you're
referring
to
here.
Cedric
is
the
need
to
have
a
different
type
identified,
different
types
of
statements
for
the
same
exact
object
identified
by
a
shot.
256
hash
value
am
I
understanding
your
concerns
correctly.
F
So,
okay
I'm
not
completely
sure
so
there
is.
H
F
Need
to
be
able
to
query
the
transparency
service
or
the
statements
registered
transparency
service,
and
for
that
you
may
you
may
have
a
more
advanced
query
language
that
says
I
want
to
know
what
is
the
latest
query
from
that
issue
for
that
purpose,
for
example?
So
so
here
the
goal
is
not
to
provide
the
Syntax
for
querying
the
the
transparency
service
in
general,
that's
more
to
identify
a
specifically
one,
previously
registered
Clan,
and
just
that
that
climate,
so.
F
Is
intensible
good
enough?
We
it's
not
meant
to
be
a
little
fast,
caring,
the
the
battery
transferences.
It's
meant
to
be
a
mechanism
to
authenticate
and
designate
when
specific,
transparent
register
statement
and
that
yes,
so
innocential
correct.
But
the
purpose
of
this
is
more
limited.
It's
really
identifying
a
register.
Transparency.
G
G
F
A
F
Okay,
so
so
what
argument
is
being
compact,
but
it's
not
just
it.
So
if
I
replicate
the
transference
statement
everywhere,
then
in
some
Advanced
attacks,
maybe
that's
a
transparent
settlement
that
was
not
actually
a
registered
of,
maybe
the
transparency
described
and
is
inconsistent,
and
so,
even
if
you
forget
about
compactness
in
terms
of
security,
being
able
to
refer
to
a
commitment
to
a
shared
transform
statement
is
arguably
more
transparent
and
more
secure.
So
there
is
that
transference
sharing
that
is
important.
F
No
I
I
yeah,
let's,
let's
say
that
I
I
think
yeah
I
think
you
you
can
in
many
cases
put
the
full
transfer
on
set
mode
I,
think
because
it
may
be
recursive
and
that
transference
Advantage
May
itself.
Is
you
really
don't
want
to
unfold
a
full
day
of.
F
And
you
have
a
statement
and
you
want
to
say
this
is
the
new
statement
for
my
firmware,
and
this
is
replacing
this
particular
statement.
F
Then,
in
that
case,
you,
including
it
would
I
think
defeat
the
purpose,
because
you
want
to
to
refer
to
something
that
is
registered
at
the
position
in
The
Ledger
or
not
not
to
it
as
a
value.
And
if
you
were
doing
it
by
value,
then
you
need
to
rebuild
the
full
story
of
all
the
firmware.
You
have
released.
A
Okay,
thank
you
Hank.
I
Yeah
hi
everybody
so
for
being
late,
I'm
in
a
weird
place
in
Orlando
right
now
that
has
curious,
Wi-Fi
and
curious
doors,
and
so,
if
you
scroll
down
here
and
I,
think
Cedric
already
said
something
different.
If
you
scroll
down
yesterday
on
this
page
on
this
issue
and
F5,
this
page
you'll
see
a
comment
that
I
prepared
an
hour
ago,
but
only
said
press
send
when
I
was
arriving
in
my
room.
So
if
you
scroll
down
yeah,
oh
it's
already,
there
yeah,
wonderful,
it's
auto!
Updating,
wonderful,
so
I
thought.
I
When
I
read
this,
what
you
wrote
here
that
we
have
normative
text
about
that
in
the
architecture
I'm
not
going
into
the
depths
if
we
should
use
moment
of
language,
but
it
says
in
the
architecture
that
payload
content
may
be
encrypted
or
opaque,
which
in
essence
means
that
a
transparent
survey
is
basically
never
required
to
understand
it.
But
it
must
understand
this
metatar,
which
goes
into
the
cozy
header,
for
example.
I
So,
having
said
that,
if
you
are
not
putting
the
actual
payload
say
cat
video
into
the
store
by
the
reference
to
it,
where
the
locator
to
me,
that
sounds
that
must,
if
you
use
a
transfer,
a
structure
that
should
maybe
transparent.
I
So
it's
non-opaque
using
the
language
here,
it's
non-opaque,
then
it
must
be
verified,
I've
already
known
here,
because
I
was
really
fast,
but
I
assume
that
you
need
a
standardized
statement
structure
that
is
deliberately
well
defined,
not
very
known,
but
also
well
known,
I,
guess
that
points
to
the
correct
location
and
I
think
we
can't
say
yeah
well.
I
Some
of
these
referrals
are
proprietary
and
you
have
to
know
how
they
work
okay,
but
we
also
need
a
standard
way
to
say
what
Dixon
said:
that's
the
content
type
or
is
you
call
it
something
different,
but
basically
you're
saying
you
don't
have
to
understand
the
type
you
were
looking
at.
I
At
least
you
know
what
the
semantics
are
to
some
extent
and
then
where
to
find
it,
how
to
resolve
it,
how
to
retrieve
it,
maybe
even
so,
with
the
scheme
and
authority
and
whatever,
and
so
thank
you
all,
and
so
my
assumption
again
to
a
long
story
short
to
me.
It
sounds
like
this
is
a.
This
is
a
few
of
the
statements
that
we
can
Define
actually
as
a
civil
structure.
Isn't
that
correct.
F
I
think
I
I
think
there
is
a
sometimes
I
I'm,
not
sure
in
particular
I
think
you
really.
Your
comment
seems
to
be
assuming
that
we
are
talking
about
the
payload
of
the
statement
and
that's
not
at
all.
We're
then
trying
to
do
there.
So
so
you
can
have
references
to
many
things,
including
to
the
the
payload.
If
you
want
to
do
the
detached
envelope
and
kind
of
things,
but
that's
not
what's
What's
Happening
Here.
F
What
you
are
referring
to
here,
oh,
is
to
a
full-fledged
constant
on
statement,
including
the
the
statement
the
in
inside
the
Sur
envelope.
That
makes
it
a
sign
set
not
and
including
the
receipt
that
makes
it
transparent.
So
it's
it's.
The
full
signed,
registered
and
transparent
statement,
not.
F
I
Thank
you.
That
is
me
just
not
reading
this
intense
enough,
but
trying
to
comment
on
it
early.
So,
okay,
so
just.
F
To
to
to
elaborate
on
that,
so
so
those
references,
of
course,
the
transparency
service,
to
agree
on
the
contents.
So
it
may
make
sense
instead
of
a
bit
or
to
have,
instead
of
an
object
period,
to
have
references
to
other
statements,
yeah
the
menu
scales,
where
the
transformation
service
will
be
that
place.
If
we
include
references.
F
In
particular,
to
refer
to
a
police
blessing,
for
example,
or
to
some
transparent
statement
to
another
part,
so
we
are
building
this
kind
of
things,
and
so
the
the
use
case
I
think,
will
be
for
those
being
included
in
headers.
But
but
again
as
it
is
written,
the
reference
could
go
in
a
policy
in
the
payload
in
the
headers
anywhere
you
just
want
to.
We
don't
want
to
say,
okay,
what
they
consistent.
Okay,.
I
Sorry
for
the
Detroit,
then
to
to
just
reflect
us
back
real
quick
I
think
the
did
web
portion
is
actually
nice
because
we
have
the
identification
of
the
transparent
statement
and
potentially
can
resolve
that
into
getting
it
which,
let's
like
like.
The
second
point,
is
the
Discovery
so
yeah.
I
Why
not
so
especially
and
you're
correct
if,
if
that
would
include
the
receipt
you
can
now
refuse
this
reference
here
in
another
tree
that
might
have
different
algorithm,
layouts
yeah
and
still
would
be
able
to
use,
make
use
of
the
statements
stored
somewhere
and
I
think
that's
an
excellent
choice.
Yeah
and
this
can
go
anywhere
even
to
the
header.
Yeah.
Okay
got
it.
Yes,.
F
I
agree,
but
to
be
fair,
several
people,
including
Roy
and
Summer
for
engineers
yeah,
complained
that,
assuming
that
the
transparency
service
will
be
able
to
reserve
this
kind
of
references
forever
is
very
dangerous
and
is
misleading.
I
Mm-Hmm
yeah,
so
if
you
do
not
so
if
you're
making
statements
about
an
artifact
that
or
sorry
about
a
I
want
to
say,
artific
right
now,
we're
saying
a
product
object,
whatever
the
thing
that
you're
talking
about
in
the
supply
chain,
the
software
effectively
in
our
case,
if
that's
to
start
somewhere,
it's
because
it
already
exists
for
20
years
and
you're.
I
Now
in
hindsight,
documenting
things
with
Sketch
and
then
from
that
point
on
start
to
use
it
as
a
live
current
feed
so
to
speak,
then
you
you
pretty
much
would
be
obligated
to
fetch
all
that
from
the
Internet
or
from
other
Authority
or
for
the
source,
and
that
does
not
make
all
this
sense
also
size
and
scalability
and
usefulness
right.
So
so
I
think
we
need
to
deal
with
verification
processes
that
say,
oh
and
by
the
way
on
my
regular
I,
don't
know
safe
order.
Check.
I
I
am
unable
to
resolve
this
anymore,
and
maybe
there
could
be
I
know:
maintenance
processes
that
that
indicate
that
these
statements
are
now
I,
don't
know,
stale
to
some
extent,
I,
don't
know
how
that
would
work,
but
is
that
a
round
forward
that
that
there
would
be
I,
don't
know
intervals
of
safe
verification
that
the
verifier
checks
all
this
again?
This
is,
of
course,
hard
work
and
it
takes
a
lot
of
effort,
but
would
that
be
a
way
to
something
prune?
These
statements
that
are
not
available
anymore,
but
we're
true
at
some
point.
F
F
Statements
for
for
three
months
and
they
have
an
issue
that
needs
to
be
available
for
five
years.
Then
the
issue
I
really
like
to
point
the
verifier
to
a
star
that
is
going
to
keep
the
statement
for
five
years
as
opposed
to
the
original
construction
service,
and
we
don't
get
a
way
to
do
that.
I
think
that
can
be
done
by
putting
locators
and
as
interesting
information
elsewhere
in
the
envelope.
But
but
if
we
do
that,
then
we
need
to
agree
that
the
distance
outside
of
this
particular.
I
So,
just
sorry
for
just
another
great
question,
but
this
would
fall
in
the
registration
policy
bucket
then
so,
because
we
have
to
Define
who,
who
is
carrying
the
duty,
who's
the
burden
for
the
duty
of
checking
that
and
now
here
now
the
issuer
is
the
verifier
just
node
to
Rises
that
those
records
it
puts
it
into
the
append
Only
log
right.
So
so
the
the
responsibility
to
take
care
they're
there
for
five
years
is
on
the
issuer
side.
You
say.
F
Yes,
okay
of
course,
I'm
making
that,
but
yes,
that
would
be
an
example.
I
I
believe
that
in
fact
it's
a
very
good
point.
The
fact
that
some
references
may
be
interpreted
by
some
registration
policies,
I
think
that
that's
going
to
be
read
about
on
whichever
the
reference
included.
F
For
example,
if
there
is
a
radar
that
says
persist
for
five
years
and
then
you,
the
translation
service,
accepts
it,
then
of
course
it
will
need
to
maintain
it
in
the
in
the
future,
but
that
would
be
the
semantic
of
a
specific
header
in
the
registration
information.
I.
Don't
think
that
would
be
attached
to
the
semantics
of
the
reference
itself.
F
A
Okay,
Charlie
you're
next.
H
Hey
everybody,
so
yes
Cedric
this
is.
This
is
interesting.
I
guess
my
my
question.
I
guess
primary
question
is:
is
this
the
is
this
a
proposal
for
the
methodology
by
which
to
build
the
data
structure,
or
is
it
a
proposal
for
an
option
that
could
be
used
within
that
data
structure?
H
In
other
words,
this
the
idea
that
we
would
have
references
to
transparent
statements
as
the
default
and
supported
version,
or
is
this
one
method.
F
So
so
it's
it's
meant
to
be
the
recommended
interoperable
method
for
reference
statement
and
I'm,
not
spitting
your
couple
of
places
where
we
will
need
in
this
architecture
to
agree
on
that
format,
and
one
of
them
is
Federation.
F
When
you
add
multiple
transparency,
services
that
refer
to
another
another
will
be
some
Advanced
registration
policy,
where
you
need
the
transparency
service
to
interpret
with
the
reference
means
and
to
check
to
check
its
resolution.
So
I
think.
H
F
H
So
I
I
guess
I.
Disagree
with
that
because,
while
we're
talking
about
internet-
and
you
know
it's
replacing
certificates
and
things
like
that,
that
works
perfectly
fine.
H
However,
in
you
know,
if
you
look
at,
for
example,
I'm
trying
to
think
of
something
else,
that
would
be
useful
here-
art,
maybe
right
so
somebody
authenticates
a
work
of
art,
and
that
goes
into
a
supply
chain
of
some
kind,
and
that
happened
in
1850,
and
you
know
that
where's,
the
where's,
the
where's,
the
URL
the
transparent
statement
there
there
has
to
be
an
option
to
maintain
everything
and
also
you
know:
Ray's
got
his
hand
up
as
well.
H
H
Think
that
is
a
fundamental
problem
when
you're
not
dealing
with
software
or
Internet
and
you're
dealing
with
you
know,
steel
or
or
or
artwork,
or
you
know,
peaches,
so
that
my
statement
is
that
I
guess
my
my
position
on
this
is
that
we
need
to
have
an
alternative
in
which
the
contents
of
the
authentication
are
present
in
the
statement
and
the
compactness
I
do
not
see
as
a
good
argument
for
that,
because
you
know
compact
does
not
necessarily
equal
better.
H
In
my
opinion,
there's
not
it's
not
that
much
information
in
the
payload
to
add
the
full
transparent
statement,
as
attested
by
the
the
authentication
Service.
F
Okay
thanks
so
so,
if
I
know
somebody
that
spout
of
the
discovery
ability,
so
you
think
discoverability
in
the
air
gap
system
or
in
many
systems,
is
going
to
be
to
be
broken.
What
I
was
suggesting
as
a
as
a
useful
patterns
for
that
is,
is
in
that
case
that
you
will
distribute,
distribute
the
statement
that
refers
to
statement
that
refers
to
state.
F
Only
together
are
we
statement
B
so
that
so
you
pass
statement
and
said,
don't
be
and
a
refers
to
be
and
then
look
at
everyone
can
check
that
the
reference
that
b
Ashes
to
what
is
regarded
as
the
hash
of
statement,
B
hashes
to
the
reference,
but.
F
Guess
both
are
included,
but
both
are
included
as
of
flatly
instead
of
one
being
invested
in
the
other.
F
F
H
Do
we
need
compact?
That's
that's
I!
Guess
my
fundamental
heartburn
over
this
compact
isn't
necessarily
all
that
great.
If
you
want
you
know,
if
you
want
a
tool,
you
know
you
know
where
the
tool
is
but
you're
not
in
that
location.
You're
not
going
to
be
able
to
get
the
tool
and
I
think
that
you
know
these
are
not
gigantic
amounts
of
data
in
the
attestation
space,
so
I
think
they
should
be
carried.
You
know,
or
at
least
have
an
option
to
carry
them
in
full.
F
Data
I
agree,
but
I
think
that
option
is
there.
So
in
a
sense
you
can
see
that
as
a
as
a
detachable.
Only
the
transaction
statement,
you
can
always
pass
the
the
list,
the
the
set
of
France
referring
to
another
okay.
H
Well,
maybe
maybe
I'm,
maybe
I'm
misinterpreting
an
earlier
statement
which
is
which
you
said,
I
think
this.
You
said
this
is
the
proposal
for
the
data
and
how
the
data
will
be
built.
H
So
if
that's
not
the
case,
then
I
feel
okay,
but
I
do
think
there
has
to
be
a
standalone
option,
which
has
the
full
attestation
built
in
with
all
the
you
know,
with
all
all
the
right,
all
the
right
data
both
to
go
off
and
re-authenticate,
if
necessary,
or
to
judge
whether
at
some
point
in
the
past
it
was
authentic
enough.
F
Yeah
I
think,
and
for
that
you
have
the
translation
and
I'm
assuming
you
lots
of
users
will
keep
the
response
balance
saved
on
for
about
team,
and
that's
that's
an
equal
thing
to
do
now.
I
guess
I
was
thinking
also
on
what
happens
on
the
on
the
transparency
service
side.
So
what
you
put
in
the
and
I
think
in
the
log,
putting
statements
that
include
full
copies
of
other
states
is
is
wrong
in
the
extreme.
F
If
each
statement
has
to
include
this
all
three
of
the
abundancy
of
other
statements,
then
you
get
very
large
statements
and
also
to
verify
the
law.
You
need
to
share
the
consistency
of
every
Nest.
Is
that
not
against
those
that
I
mean
regarded
in
the
past?
So
that's
more
verbose
and
less
secure.
G
H
Okay,
well,
yeah
I
mean
I.
Think
that
there's
I
don't
want
to
dominate
the
floor
here,
but
there
are
differences
between
something.
If
I
was
you
know,
if
I
was
a
reputable
organization
in
1850,
but
today
I'm
not,
then
the
references
are
gonna.
Do
you
any
good,
so
I
think
there's
so.
H
I'll
leave
it
at
that
I'll.
Let
take
the
floor
and
Express
his
issues.
F
A
Yeah
I
guess:
we've
beaten
that
one
that
Ray
is
next
in
the
queue.
B
Thanks
I
think
that
okay,
so
one
thing
is
I'm
resolving
a
little
bit
more.
What
the
use
case
is
for
the
selection,
data
situation
and
I
the
discoverability
and
the
reference
are
going
to
be
pretty
important,
I
think
because
this
data
is
going
to
be
developed
incrementally
and
so
there's
going
to
be
an
initial
say,
a
hunk
of
data,
it's
not
very
big,
which
is
the
list
of
all
the
public
keys
of
the
devices
and
that
needs
to
be.
B
You
know.
The
transparency
service
is
perfect
for
making
that
public
and
locking
it
down,
but
that
has
to
be
locked
down
earlier
than
you
know.
We
get
the
data
from
those
devices,
then
later
we
get
incremental
data
from
some
devices
and
not
others,
and
and
then
there's
this
Central
devices
that
are
that
are
and
we
we
would
like
to
get
that
data
incrementally
and
and
then
so
we
at
the
end,
we
need
to
have
a
way
to
either
as
part
of
the
skit.
B
B
That
will
allow
us
to
get
to
all
of
the
data
and
and
say
okay
for
this
particular
in
this
case
election
at
a
certain
District.
What
is
all
the
data
you
know
give
me
all
the
data
for
it
for
a
software
use
case.
It's
more
like
give
me
all
of
the
software.
You
know
stuff,
plus
all
the
tests
and
the
results
and
from
the
testing,
agencies
and
stuff
and
and
I
need
it
all,
and
so
there
needs
to
be
a
way
to
number
one
have
something
that
is
outside
of
the
Obscure
area.
B
If
you
want
to
use
it
and
then
a
way
to
say
to
for
for
for
the
submitter,
maybe
to
say,
I'm
I
want
to
group
all
this
together
like
maybe
maybe
we
can
have
a
fairly
simple
skit.
You
know
machine
that
doesn't
know
very
much,
but
the
users
of
it
can
know.
Okay
I've
submitted
all
this
stuff.
It
has
these
references
and
I'm
combining
that
all
together.
B
Finally,
into
this
one,
final,
submission
or
or
at
least
incremental
submission,
and-
and
here
it
is
and-
and
so
I'm
saying
my
new
claim-
is
that
these
other
submissions
are
all
part
of
this
larger
thing,
and
if
we
do
it
that
way,
then
we
can
kind
of
not
have
to
worry
about
the
discoverability
and
other
things
like
that.
We
kind
of
put
that
on
the
shoulders
of
the
users
where
they
they
would
know
what
they've
submitted
and
they
would
say.
B
So
if
it's
a
software
use
case,
you
would
say:
okay,
here's,
the
binaries,
here's
the
you
know,
ports
to
XYZ
and
and
then
here's
some
of
the
testing
data
and
then
and
I'm
grouping
that
all
together
into
this
into
this
sort
of
medicine,
Mission
and
then-
and
it
refers
to
the
other
submissions,
so
that
people
outside
of
the
scope
of
that
particular
use
case
may
not
be
able
to
find
out
like
what
what
it
is
that
they're
grouping
together.
B
There's
there
may
be
not
too
much
exposed
in
the
non-obscure
part
of
it,
and
you
would
still
be
able
to
find
that
if
you,
if
you
had
the
the
sort
of
meta
submission
concept
that
that's
sort
of
my
idea,
but
this
it's
kind
of
important
for
for
us
to
be
able
to
get
to
to
number
one.
Have
this?
B
Have
this
get
registry
be
incrementally
something
we
can
incrementally
submit
to
have
those
all
locked
in
and
then
have
some
way
of
having
a
final
submission
that
also
refers
to
the
incremental
ones,
because
we
want
to
be
able
to
say,
for
example,
that
that,
like
back
in
time
at
time
x,
we
did
submit
this,
and
this
is
exactly
what
was
submitted.
We
can't
rely
upon
a
new
submission
to
say:
okay,
now
I'm
submitting
a
new
and
by
the
way,
I
changed
everything
this.
So
it
complies
with
my
new
hacking
scenario.
B
So
so
it
would
have
to
be
able
to
refer
to
older
entries,
and
so
that's
why
I
think
the
sort
of
meta
entry
can
solve
the
discoverability
and
and
not
not
provide
too
much
of
the
obscurity.
I
hope.
Maybe
I've
said
enough
about
that
idea:
okay,
next.
F
So
I
I
think
I
understand
your
the
use
case.
I
I
think
we
are
so.
F
Is
really
centered
on
the
transparency
service,
and
so
in
particular
what
we
got
a
statement.
It
really
is
the
unit
of
registration,
so
you
describe
a
providing
technology
that
are
about
what's
happening
and
so
I
think
from
the
transformative
each
time
you
register
something
and
to
to
make
it
transparent,
yeah.
It
is
a
set,
a
high
statement
and
I
think
that's
okay
in
particular,
it
means
the
transparency
that
is
not
know
about
this
meta
level.
Grouping
of
of
State
nodes
and
I
believe
that's.
F
That
fits
your
your
need,
assuming
that
the
user
is
okay
to
deal
with
collections
of
statement
of
related
statements
for
given
purpose
as
opposed
to
a
single
statement.
I
I
think
if
you
try
to
first
everything
or
the
scenario
or
the
conclusion
of
of
a
protocol
statement.
F
Is
in
case
then,
then
we
have
a
problem,
because
the
unit
of
registration
and
the
unit
of
what
needs
to
be
stated
for
the
users
are
quite
difficult
but
again
I'm
hoping
that
as
long
as
the
user
is
okay
to
use
to
do
the
meta.
B
B
It
was
said
in
the
past
and
that's
what
that's
what
they
said
and,
and
so
they
can't
say:
okay
well,
this
you
know
this
is
the
new
data
that
that
we
with
that,
without
also
referring
to
the
older
data
and
say
this
is
what
used
to
be
you
know,
maybe
you
could
do
it
that
way,
but
it's
I'll
tell
you
what
I
I
think
the
the
the
so
the
discoverability
maybe
like
if,
if
you're
someone
outside
and
you
and
you
want
to
say,
find
all
of
the
products
that
had
this
type
of
airbag,
then
you're
gonna
need
to
somehow
I.
B
Don't
know
how
you
would
do
that
if
you
have
obscure
statements,
unless
you
were
able
to
somehow
see
into
them
a
bit
more
and
find
that
airbag
or
or
that
you
know
library
of
in
the
bombs.
You
know
you
would
need
to
be
able
to
to
find
that,
and
so
the
some
of
the
functionality
of
looking
in
in
in
unless
you
you
somehow
could
I
just
that.
That's
the
part
that
I
think
is
a
problematic
with
with
having
a
you
know,
completely
obscure
submission
statements.
Then
how
do?
B
How
do
people
actually
find
that
library
that,
like
who's
using
this
Library,
we
needed
to
find
out
what
the
you
know
that
that
would
be
a
good
challenge
to
to
to
it?
B
I'll
tell
you
what,
though,
I'm
I'm,
getting
closer
to
understanding
what
you're
doing
here
and
what
my
needs
are
in
this
and
I
will
I'll
make
a
comment
on
this
in
written
form.
Okay,
thanks,
good.
F
Thanks
yeah,
in
fact,
I
just
want
to
as
a
commenter
I
think
using
references
when
you
want
to
replace
statement
that
you
made.
This
is
a
very
good
use
case.
So
if
you
have
two
statements
signed
by
the
same
issue
for
the
same
purpose,
then
you
really
don't
know
which
one
follows
from
the
other.
Now,
if
you
have,
if
one
of
them
explicitly
refers
to
the
other
one
and
you
have
God,
then
you
clearly
you
can
compare
both.
F
You
can
understand
whether
that
was,
and
you
really
know
very
clearly
and
you
can
promise
everyone
that
you
have
the
latest.
F
That
will
have
a
reference
where
we
can
probably
Define
the
quite
a
specialized
useful
registration
policy
for
checking
that-
and
it's
also
a
case-
we
are
putting
the
the
full
statement
that
you
replace
by
value,
doesn't
scale.
B
Yeah,
okay,
so
let
me
give
you
an
example:
within
the
software
use
case,
let's
say
we
have
a
piece
of
software
and
we
have
let's
say
it's
a
something:
that's
critical
like
a
you
know,
self-driving
car
software
that
they
have
to.
They
have
to
submit
and
say
this
is
our
self-driving
car
software
and
it
was
tested
according
to
these
criteria
and
here's
the
test
results.
And
then
you
know
later
they
had
an
update
and
they
updated
and
fixed
stuff.
But
some
of
the
cars
were
still
running
on
the
old
stuff.
B
There
was
an
accident.
People
died
so
they're
in
court
and
they
want
to
find
out
what
version
of
the
software
and
what
is
it
tested.
So
not
only
they
need
to
go
back
and
find
out,
not
the
new
stuff
but
the
old
stuff,
and
what
the
test
results
were
there.
So
we
need
to
have,
as
you
can
see
there,
we
need
to
have
the
full,
like
even
the
older
results
that
they
can't
re.
You
know
we
don't
want
them
to
be
able
to
replace
that
and
and
never
and
or
change
it.
B
B
They
would
have
to
say
no
I
can't,
and
this
is
what
we
want
them
to
have
to
say,
I
can't
because
it
was
only
tested
to
this
according
to
the
skit
Ledger,
because
it
locks
in
where
my
test
results
were
and
that's
comparable
to
what
I
need,
so
that
so
there
would
be
a
distributed
set
of
things
in
The
Ledger
not
distributed
I'm.
B
Sorry,
that's
a
bad
name,
because
it
implies
something
else,
but
say
a
number
of
entries
in
The
Ledger
that
refer
to
the
same
product
at
different
stages
and
different
releases,
but
they
all
need
to
be
kept
there
because
of
the
different.
You
know,
levels
of
functionality
and
whether
they've
been
tested
and
so
forth,
and
so
they
can't
replace
each
other.
But
now
we
need
to
find
them,
and
so
now
now
we
need
to
be
able
to
say
okay
now
for
this
product,
so
somewhere
there
needs
to
be
a
marker
of
I.
B
Think
that
is
not
obscure
for,
and
this
is
what
I
expressed
in
that
Sky
View.
We
need
to
know
who
the
manufacturer
is
and
what
the
product
is,
at
least
so
that
we
can
track
those
things
down
and-
and
maybe
there's
a
thing
where
you
can
say:
look
I,
don't
want
anyone
to
know
anything
going
in
and
so
I
don't
want
to
use
that
metadata,
but
I
think
most
of
the
time
we'll
say:
okay,
look
you're
going
to
have
to
identify
who
you
are
you're
submitting
this
and
what
your
product
is.
B
A
Yeah
two
issues:
one
is
and
that's
a
sort
of
question
to
Cedric,
to
or
request
to
Cedric
to
sort
of
fine
tune.
The
the
motivation
I
think
the
compactness
is,
is
it's
an
optimization,
but
it
also
obviously
has
some.
Some
disadvantages
makes
things
more
complicated,
Etc,
but
I
I
hear,
but
there
are
other
use
cases
from
the
discussion
so
far.
A
It's
just
the
difficulties
to
draw
the
line
between
referencing
and
a
query
language.
So
there's
there's
a
slippery
slope
here.
That's
the
first
point.
The
second
issue
is
more
technical.
One
I
believe
that
did
web
URI
scheme.
Here
is
not
what
you
should
be
using
for
this
purpose,
because
I
would
I
would
just
use
a
urine
and
Define
how
the
discovery
works,
because
they
did
that
at
least
that's.
A
My
understanding
of
the
specification
specifically
refers
to
identifying
a
did
rather
than
a
statement,
so
you
actually
looking
at
a
at
using
or
defining
a
different
URI
scheme
and
that's
not
a
big
drama.
It's
just
a
different,
a
different
name
essentially.
F
F
F
Of
where
it
was.
A
Quite
aware
of
it,
yeah
yeah
I
could
I
can
in
the
comments
I
can
post
a
message
on
how
I
think
it
should
be
done.
It's
a
small
technical
issue
rather
than
sort
of
the
earlier
discussions,
were
more
on
a
use
case
and
broader
architectural
scope
rather
than
this
one.
So
it's
it's
a
small
thing.
F
On
this
issue,
so
yes,
comments
and
yeah
Andre
is
welcome.
F
G
Thank
you
Hamas,
so
one
thing
that
concerns
me
is,
you
know:
the
software
supply
chain
is
very,
can
get
very
complex,
very
quickly
and
I
think
we're
being
a
little
bit
too
abstract
in
our
discussions.
G
I
think
we
can
flush
out
The
Core
Concepts
of
a
trans
Currency
Service,
using
a
very
simple
model
based
on
something
we
can
all
relate
to
and
I'll
give
you
a
very
quick
example:
Johannes
is
going
to
be
the
transparency
service,
that
registers
people's
weights,
all
people,
information
and
so
I
want
to
do
a
look
up.
I
want
to
say,
I
want
to
know
the
weight
of
Dick,
Brooks
and
I
need
to
get
it
from
harnesses
registry
transparency
service.
G
There's
two
things
that
come
right
off
the
bat
we
need
to
know
we
need
to
identify.
We
need
to
be
able
to
identify
the
transparency
service
entity
that
we
can
query
and
we
need
to
know
the
subject
that
we
want
to
submit
the
query
on
so
there
you
have
two
concepts
that
both
apply
to
software
as
well.
With
regard
to
this,
whole
concept
of
weight
weight
is
a
great
example
of
going
back
to
what
Hank
was
saying
earlier
about
staleness
and
freshness.
G
You
know
if
you
record
my
weight
from
today,
it's
very
different
from
my
weight
that
we
recorded
you
know
in
1979
when
I
got
married,
so
the
concept
of
freshness
and
Stillness
would
play
out
with
this
simple
use
case
as
well.
So
what
I'm
suggesting
is?
Maybe
we
can
just
start
with
something
simple
that
we
can
all
relate
to,
as
we
try
to
flush
out
what
this
transparency
service
thing
is
and
what
it
provides
and
I
think
the
just
using
that
weight
example
is
a
good
one
that
we
could
easily.
You
know
flush
out.
G
Some
key
Concepts
from
the
things
like
you
know,
for
example,
in
a
wake,
is
a
unit
of
measure
right,
so
that's
data
that
would
have
to
be
carried
in
the
receipt
there's
a
time
stamp
of
when
that
rate
date
rate
was
recorded.
You
know
so
there's
all
kinds
of
other
information
that
we're
going
to
the
statement
in
that
one
case,
so
you
know,
even
though
the
weight
one
is,
is
not
germane
to
what
we're
trying
to
accomplish
here.
G
F
So
I,
okay,
so,
first
of
all
I
I
agree,
that's
something
we
need
to
do.
I
I
think
that's
out
of
scope
in
this
particular
issue.
So
as
as
mentioned
a
few
times,
this
is
not
meant
to
be
a
very
long
wedge
and
what
you
are
I,
think
informally
asking
for
me:
what
are
the
most
natural
requirements?
What
made
up
what
may
have
about
statements
so
this
this
is
just
a
way
to
identify
and
refer
to
one
part,
one
specific
transfer
on
set
mode.
So
back
to
your
horizontal.
F
It
provides
me
a
way
of
referring
indefinitely
to
each
of
your
360
weights
register
that
a
particular
transference
service
not
well,
not
less,
and
if
you
want
to
know
what
is
the
to
get
the
history
or
to
get
the
letters,
that's
not
the
purpose
of
a
concern
segment.
That
will
be
something
else
so.
E
It's
experimenting.
G
F
Building
tool
to
refer
to
specific
transferences
statements-
and
there
are
use
cases
for
that.
Many.
G
Yeah,
it's
it's
really
a
proposal
to
try
and
simplify
this
to
something
that
we
can
all
say.
Okay,
here
are
the
key
Concepts
that
we
know
we
need
to
capture
for
any
transparency
service.
That's
going
to
produce
anything,
but
you
know
if
that's
not
going
to
work
for
people
and
obviously
that
then
that
will
fly.
So
if
it's
not
going
to
work,
then
I
I
withdraw
my
suggestion.
Thanks.
A
Crank
stick
next,
one
is
Hank.
I
I
Charlie
said
some
of
this
historicity
is
not
captured,
digitally
correct,
even
the
software.
Sometimes
this
is
in
a
very
weird
format.
That
is
that
we
wouldn't
call
it
digital
in
exercise.
I,
don't
know,
that's
not
post-processible,
effectively
and
or
weirdly
formatted,
and
then
and
then
we
have
this
weight
example
from
dick,
which
is
the
same
same
because
dick
is
recording.
I
This
I
know
in
his
own
EXO
sheet
again
so
so
now
we
have
skit
and-
and
now,
if
you
want
to
add
a
a
ancient
software
to
the
skid
Ledger
and
its
historicity,
there's
an
authority
doing
that
that
is
the
issua.
So
Charlie
highlighted
that
this
information
is
not
really
available
on
the
internet,
but
well
someone
issued
the
statement
so
why
that
statement
is
issued
by
the
issuer.
Now
this
issue
is
responsible
of
vouching
for
the
authenticity
and
the
rest
of
the
historicity
of
that
statement.
I
Made
transparency
by
the
service
I
think
so,
whatever
you
do,
even
if
it's
hundreds
of
years
old
and
about
a
painting,
let's
call
it
like
ancient
software
that
ran
on
floppy
disk
with
the
air
gap
naturally
built
in
because
there
was
no
network
I,
think
that
is
something
you
know,
burden
the
issue
of
it.
I
So
the
issue
is
doing
all
the
Providence
research
providing
his
historicity
and
then
then
issuing
that
a
statement
as
the
initial
statement
about
the
product
or
the
object,
and
then
we
go
from
there
right
so
I
think
it's
fine
as
long
as
we
burden
the
issue
with
that
responsibility
of
providing
that
data
in
statement,
format,
of
course,
and
and
then
go
from
there.
So
I'm
I'm
a
little
bit
surprised
that
this
this
perceived
as
a
problem,
because
now
the
issue
would
be
responsible
to
resolve.
A
H
I'm
sorry
go
ahead:
yeah.
B
H
Should
have
put
my
head
down
but
since
I
since
I'm
up
I'll
just
answer
Hank's
question.
H
Concern
I
had
was
not
so
much
with
the
issuer,
but
the
availability
of
the
internet
to
go
off
and
look
it
up.
That's
that's
my
concern.
I,
like
the
idea
of
a
standalone
set
of.
I
H
I
Coming
back
to
that
really
quick
again,
the
the
issua
takes
on
the
responsibility,
if
you
don't
trust
the
issue
and
the
transparency
service.
Of
course,
the
offline
very,
very
fiber
receipt
won't
help
you,
you
have
to
put
trust,
put
trust
in
this
treasure
Service
as
a.
I
Their
work
correctly
because
they
they
pass
the
registration
policy
of
the
transparency
service
I
mean
then
at
this
point
you
can.
You
can
believe
that
someone
was
at
this
point
of
of
of
making
it
transparent
at
least
providing
it
on
the
internet,
maybe
for
two
days,
I,
don't
know,
but
but
that
should
work
right.
I
H
D
I
mean
there's
it's
very.
H
Intriguing
use
case
I
had
to
say,
but
yeah.
H
To
it,
but
I
just
think
that
there's
going
to
be
once
we
get
into
other
Supply
chains,
you
know
in
the
middle
of
nowhere
if
you're
going
to
trust
a
load
of
Steel
that
becomes
harder
to
you
know,
authenticate
the
endorser
right
things
like
that,
so
yeah.
So.
G
A
H
Not
I'm
not
objecting
to
having
that
available,
but
I
also
think
that
it
would
be
useful
in
many
cases
to
have
the
full
data
set
available
as
well.
F
Okay,
so
yeah
I,
I
I,
think
that's
super
important.
So
let
me
let
me
check
if
we
we
are
here
on
this
one.
So
so
to
me,
this
proposal
is
really
not
as
Humane
that
you
need
to
do
a
lookup
in
order
to
resolve
a
reference.
So
to
me,
the
most
comment
here
is
will
be
that
as
a
verifier.
A
F
Receive
two
statements:
two
transparent
statements,
one
referring
to
the
other
and
you
recompute
the
hash
of
the
sermon
statement
to
check
that
it
matches
the
one
in
the
first
state
mode,
so
that
you
know
that
the
visual
of
the
first
set
month
is
really
meant
to
refer
to
the
second
segment
that
you
have
it
on,
and
that
does
nothing
going
to
the
internet.
That's
something
you
can
do
on
your
own
or
as
soon
as
you
have
a
copy,
either
received
or
obtained
from
any
other
source
of
the
second
transform
claim.
So
the.
F
That
discular
ability
will
be
required
on
first
unnecessary
to
share
your
reference.
The
easiest
case
is
you
have
a
transform
set
mode,
you
have
it
and
then
you
check
that
it
matches
a
reference
in
another
transference.
A
But
that's
only
half
of
the
story
Cedric,
because
because
in
in
your
write-up
you
kind
of
assume
that,
for
example,
in
this,
what
you
show
on
the
screen,
DS
dot
software.org-
is
that
there
is
some
that
this
is
at
some
place
right
at
a
transparent
statement
and
in
on
some
transparency
service
somewhere
on
the
internet.
At
some
point
in
time.
Right.
E
F
Need
to
to
check
if
that
transparency
service
is
still
alive.
A
Right,
but
if
and
maybe
maybe
what
we
need
to
do
is
really
as
Charlie
and
Ray
mentioned.
We
need
to
work
through
a
specific
use
case
here,
but.
F
I
I
a
couple
of
use
cases.
The
discussed
today
would
be
really
good
and
I
can
try
to
do
that
as
a
comment
as
well.
Yeah.
A
I
I
I
give
I've
noticed
like
we
got
Tracked
Down
with
the
discussions,
which
is
which
is
good
to
talk
through
this,
to
get
a
better
understanding,
it's
a
important
architectural
concept,
but
Ray.
You
want
to
say
the
last
few
words
before
we
have
to
finalize
the
meeting
for
today.
B
Well,
yeah
I
know
that
we're
out
of
time,
but
I
do
think
the
transparency
service
we'll
have
we'll
need
a
functionality
that
will
say
this
reference
I'm
give.
These
are
all
of
the
submissions
for
this
reference
that
I
have
and
there
are
no
more
and
there
are
no
less,
because
otherwise,
a
person
who
claims
to
be
honest,
emitter
can
say
okay.
This
is
this
is
all
the
information
that
we
have
about
this
and
we
didn't
do
anything
in
the
past,
even
though
they
did
so
anyway.
B
Yeah
yeah
I
will
I'll
write
it
up
because
I
I
am
getting
closer
on
this
use
case
and
I'll
put
the
link
to
for
where
I
am
because
there's
still
a
few
loose
ends,
but
it's
getting
locked
down
some
more
so
I'm
starting
to
understand
how
to
use
these
pieces.
Thank
you.
A
Okay,
cool
yeah
thanks
a
lot
for
the
discussion
we
I
took
some
notes
and
obviously
there
are
some
some
action
items,
apparently
not
the
end
of
the
story,
but
I
I
think
we
all
now
much
understand
much
better.
A
What
you
said
Rick
tries
to
accomplish,
with
with
your
proposal
and
thanks
for
the
details
right
up
and
for
the
explanations
today,
and
we
also
I
think
we
looked
into
some
additional
use
cases
or
more
advanced
use
cases
which
we
also
need
to
look
into
which
are
related
to
this,
but
obviously
not
the
same,
so
even
more
work.
So
we
went
out
to
solve
One
open
issue
and
we
created
an
extra
one.
B
G
A
Yeah,
thank
you
all
and
have
a
good
Monday.