►
From YouTube: IETF110-DRIP-20210309-1600
Description
DRIP meeting session at IETF110
2021/03/09 1600
https://datatracker.ietf.org/meeting/110/proceedings/
A
C
Good
good
and
that's
wanna
make
sure
that
works
for
later.
Okay,
I'll
mute
myself
back
for
now,.
B
B
A
A
Please
take
the
time
to
read
the
notewells
and
if
you
have
any
feel
free
to
to
raise
your
question,
I'm
gonna
go
through
the
slides
and
medic
is
gonna
handle
the
line,
the
queue
I
hardly,
I
can
hardly
see
the
chats,
but
I
will
leave
med
to
do
that.
So.
A
So
logistics
we
have,
we
need
a
mini
taker,
so
it
says
two,
the
minute
you
can
help
him
by
clicking
on
on
the
link
here,
I'm
gonna
put
the
link
into
the
chat,
so
everyone
can
see
that
and
you
can
help
him
would
be
pretty
appreciated.
A
You
don't
have
to
sign
the
blue
sheet.
It's
gonna
be
automatically
automatically
done.
You
know,
I
guess
how
to
use
the
different
tools
we
have.
You
should
not
be.
A
You
shouldn't
maybe
not
need
to
share
your
screen,
but
if
you're
talking
you
can
put
the
video
on
and
and
the
the
mic
on
it's
it's
it's
better.
A
You
can
put
yourself
in
the
queue
by
raising
your
hand,
if
we
need
any
hum,
we
have
a
tool
here.
I
don't
know
if
you
see
my
mouse,
but
that's
if
we
are
asking
a
question
to
the
working
group.
This
is
basically
what
we're
going
to
use.
A
So
the
agenda,
I
think,
is
there
any
comment
regarding
the
agenda.
Otherwise
we
should
go
through
the
different
drafts
and
see
we
have
basically
four
presentations.
So
it's
ongoing
draft
implementations
updates
and
new
coming
drafts.
A
I
currently
we
have
two
working
group,
two
working
group
document
that
are
in
working
group
last
goal.
We
have
a
two
document
that
are
working
documents
that
are
making
progress.
The
working
group
lascal
received
many
comments.
One
question
I
would
have
is
I
propose:
I
mean
how
we're
going
to
use
github.
I
don't
want
to
open
the
debate
on
whether
how
good
is
github
and
so
on.
So
I
propose
that,
instead
of
just
having
a
huge
pr
as
I
did,
we
should
rather
have
a
list
of
open
issues.
A
We
could
put
them
into
the
github
tracker,
but
the
discussion
should
happen
on
the
mailing
list
so
which
means
that
every
issues
that
gonna
be
raised
in
the
tracker
will
be
open
and
discussed
on
the
mailing
list.
So
what
I
propose
to
is
that
I'm
gonna
add
issue
regarding
my
comments
and
I
will
let
the
co-author
to
open
the
discussion
on
the
mailing
list.
A
So
how
do
we
I'm
wondering
if
people
would
like
to
introduce
more
interim
meetings
or
if
that's
a
reason,
if
the
interim
meetings
are
helping
to
make
progress
on
the
draft?
That's
something
I
would
be
happy
happy
to
know.
If
those
interim
meetings
are
needed,
we
can
set
some
from
last
atf.
I
don't
remember,
we
had
any
interim
meetings
and
we
should
try.
I
mean
we
should
try
to
see.
A
If
that
makes
a
difference,
and
last
thing
I
had
a
discussion
with
the
rfc
editors
and
one
other
thing:
they
they
told
me
that
would
help
them
is
to
use
xml
the
three.
So
if
we
have,
I
mean
when
you're
submitting
your
draft,
you
could
basically
try
to
submit
the
v3
version
instead
of
xmlv2.
A
A
Basically,
you
could
use
the
command
line
that
is
provided
by
casten,
or
there
are
also
online
tools,
but
just
keep
in
mind
that
it
may
be
something
to
keep
any
comment.
Any
question.
Any
thanks
before
we
move
to
the
next.
E
Hey
daniel
yeah,
just
just
this,
is
regarding
the
recent
comment
you
posted
on
the
github,
so
the
the
general
comment
is
someone
we
use
github
very
frequently,
and
I
do
feel
like
putting
comment
inside
of
the
documents
it's
pretty
hard
to
follow.
E
A
Okay,
so
the
comments
I'm
gonna,
the
comments
I
made,
I'm
gonna
put
them
into
the
the
git
tracker.
E
Comment,
you
have
already
you
just
delivered
there.
I
will
organized
right
for
for
now.
You
don't
have
to
spend
the
time
to
do
that
again,
because
I
okay
I've
gone
through
a
couple
times
already
initiated
the
discussion
in
the
middle
east.
So
that's
five,
but
for
the
following
up
comments,
I
would
rather
you
put
a
pull
request
then,
which
basically
means
this
is
one
of
the
open
issues.
Then
either
small
one
is
fine
right,
so
we
can
address
each
of
them.
Then
you
know
how
many
views.
A
A
E
A
Okay,
nice
feedback.
I
will
proceed
as
as
you
suggest,
okay
and.
C
C
Yeah,
I
pretty
much
have
moved
us
to
version
three
of
xml
to
rfc.
I
think
we're
we're
good
on
that
on
current
drafts.
Just
just
just
to
let
you
know
about
that.
I
made
that
transition
pretty
much
once
hendrix
said
that
version
three
was
ready
to
go.
A
Okay,
good,
so
just
to
make
sure
I'm
understanding.
Schwei
is
gonna,
handle
the
comments
I
provided
in
in
into
the
tracker
and
you're
gonna
open
the
issues
and
bring
the
discussion
to
the
mailing
list.
So
I
I
don't
need
to
do
anything
on
that
side.
Is
that
correct.
E
Daniel
that's
a
firm
team.
Yes
for
now,
for
the
current
account
that
you
have,
I
would
definitely
do
that.
A
F
Actually
imagine
that
we
will
start
with
the,
I
would
say
with
the
architecture,
but
if
it's
hy
the
that
would
be
next
on
the
floor.
Okay,.
E
E
No
okay,
thank
you.
So
everybody
know
the
architecture
now
using
welcome
to
blast
call.
So
here's
just
provide
a
updated
summary
things
on
ietf109.
Yes,
we
don't
have
an
intermediate
between
and
also
bring
up
some
of
the
important
comments
from
last
revision.
That's
just
key
for
us
to
move
forward.
I
do
have
a
slide
in
the
end
to
just
discuss
how
to
should
we
move
forward
next
slide.
Please.
E
So
here's
a
quick
summary:
we
have
done
six
revisions
fence,
109,
the
thanks
to
all
the
reviewers,
daniel
and
eric
and
matt.
Just
all
provide
a
good
awesome
comments
and
the
welcome
blast
school
is
on
revision
11..
So
far
we
have
received
comments
from
daniel
in
the
github
and
stefan
from
million
list
and
eric.
E
E
Next
slide:
okay,
so
here
is
not
the
entire
list
that
we
have
done
so
far,
so
obvious
summary
center
for
youtube
revisions
in
the
middle
list.
So
if
you
are
interested
to
get
a
chance
in
energy
each
revision,
you
can
definitely
look
through
the
mini
list,
but
some
of
the
major
objects
needed
to
be
emphasized
here,
the
first
one.
We
address
all
the
editor
notes,
which
is
greater
progress
and
bob
did
a
thousand
review
and
proposed
text
for
the
fa.
E
Final
row,
which
is
great
and
also
one
of
the
major
things
that
we
have
agreed
to
is
to
over
reorganize
the
previous
text.
I
believe,
that's
that's
this
revision
five
and
we
try
to
consolidate
all
the
hit
visual
text
to
section
four
and
five
and
also
section
six
introduced
the
crowdsourced
rate,
which
is
trying
to
bridge
broadcast
study
to
network
id
and
then
stools
test.
We
need
to
align
with
the
requirement
with
some
the
terminologies,
like
things
like,
replace
us
us
id
with
the
drip
entity.
E
Identifier,
which
is,
I
think,
is
really
reasonable.
I
mean
the
the
principle
here
and
the
reason
for
the
changes,
because
we
understand
the
us
read:
stand
for
remote
identification,
it's
really
not
a
identifier,
and
but
in
drip
we
actually
working
on
define
a
ues
identifier.
So
I
think
the
change
is
aligned
with
the
requirement.
E
It's
it's
a
it's
good
as
testing
another
decision
we
recently
made
since
revision
10
is
to
remove
the
original
section,
7
drip
transaction
trustworthiness,
since
we
all
believe
that
actually
belonged
to
a
solution
space
in
terms
of
where
it
goes.
I
think
the
co-author
will
decide
later
and
the
last
point
I
want
to
mention
since
the
architecture
draft
we
all
agree
that
is
informational,
nature
and
also
for
80
recent
comments.
E
We
replace
all
the
normative
wording,
specifically
in
section
four
for
item
six.
I
think
that's
kind
of
like
quick
fix,
not
sure
this
is
actually
100
percent
right
thing
to
do
in
the
ietf.
So
do
provide
your
comments
on
that
next
slide,
please!
E
So
here's
the
things
one
of
the
topics
really
to
get
some
decision
in
these
meetings
right.
We
want
to
get
the
working
group
consensus,
which
is
regarding
you
know
what
level
specification
details
right.
Drip
architecture
should
include
right.
That
means
more
specifically,
we
should
ask
the
working
group.
It
is
the
right
thing
to
include
enough
specification
details
to
justify
the
proposed
solutions.
E
The
discussion
is
really
around
the
new
sections
we
organized
like
four
five
six,
but
just
to
recap
right
for
what
other
sections
are
right.
Section
four
is
really
just
the
usage
of
hhit
as
a
trustable
drip
identifier
and
which
is
also
including
some
of
the
hid,
background
definition
and
some
level
of
the
details
regarding
how
hrt
can
satisfy
triple
requirement.
E
E
E
See
if
this
is
really
acceptable
to
including
some
level
of
the
details
to
like
justify
why
we
propose
hhit
and
some
other
ritual
matters.
So
the
first
comment
from
danielle
mentioned:
it's
really
reasonable
to
have
additional
background
information
around
the
rip
ecosystem,
but
of
course
some
texts
need
to
be
simplified
needed
to
be
cleared,
so
that
is
actually
aligned
with
our
current
understanding
right.
E
So
I
don't
want
to
read
it
through,
whatever
danny
has
said,
but
I
think
that's
actually
aligned
with
our
current
understanding
next
slide,
but
the
recent
review
we
get
from
eric
and
stefan
kind
of
posed
a
different
direction.
E
I
recommend
that
it
it's
kind
of
unusual
to
have
normative
wording
in
architecture
draft
right,
but
that's
all
that,
but
that
is
not
a
prohibited,
which
is
great
news,
but
we
need
to
deal
with
the
normative
word
to
reply
to
that,
since
revision
10,
we
already
removed
or
replaced
the
number
two
wording
into
sections,
as
I
mentioned
earlier,
not
sure
this
is
acceptable
yet
because
I
don't
receive
any
comments
so
hopefully
can
get
some
comment
later
then.
Stefan
also
expected
similar
concerns
regarding
the
mixed
solution
with
the
architecture
discussion.
E
E
So
in
order
to
move
us
forward,
I
think
we
need
to
make
a
decision
in
this
meeting
so
how
to
fix
this
issue
right.
This
is
so,
but
I
want
to
remind
us
right
before
I
think
some
time
I
run
10801
to
9.
E
We
have
agreed
that
drip
architecture
is
actually
a
solution
based
adaptation
right,
it's
kind
of
informational,
but
it's
also
proposed
as
some
sort
of
a
solutions
right.
Not
only.
We
should
include
the
entities
around
the
drip
and
explain
how
each
components
talk
to
each
other
as
that's
we
have
done
in
section
one
now
also.
We
need
to
propose
a
solution:
how
to
fulfill
the
requirements
specified
in
a
requirement
draft
a
result
without
repeating
the
text
in
the
requirement.
E
So
that
being
said,
in
my
opinion,
I
think,
including
all
the
co-authors
from
offline
discussion,
we
believe
all
those
sections
should
stay
but
of
course,
some
of
the
details.
As
daniel
mentioned
already,
we
needed
to
be
justified
that
can
be
done,
but
before
we
even
move
forward
to
fixing
stuff-
and
we
need
a
working
group
of
consensus
if
we
should
keep
these
sections
right
now,
we
are
in
the
working
group.
Last
call.
E
I
think
we
needed
to
ask
specifically
that
if
we
want
to
and
keep
section
415,
then
I
guess
things
are
much
easier.
We
just
follow
up
with
the
comments
and
fix
the
problem
about
point
by
point,
but
on
the
contrary,
we
don't
want
this
in
the
architecture
draft
and
just
to
be
honest,
I
have
I
don't
know
how
to
how
to
move
forward
without
even
talking
about
hijacking,
because
I
think
that's
a
decision
we
made
a
long
time
ago.
E
In
the
end,
in
my
home
opinion,
I
think
it's
those
sections
actually
provide
a
really
great
introduction
information
for
this
solutions
base
right
things
like
us,
id
or
authentication
draft
right
with
all
the.
Without
this
background
information,
the
the
solution
seems
like
a
weight.
It's
a
little
weak
to
me.
E
E
E
I
think
that's,
that's
the
that's
all
I
wanted
to
discuss
this
last.
This
is
the
last
slide.
That's
all
well,
this
is
all
I
wanted
to
discuss
and
so
floyd
is
open
and
I
want
us
to
I
want
to
hear.
I
want
to
hear
the
other
comments
regarding
how
to
how
we
deal.
C
C
Draft
project
presents
ietf
as
an
sdo
as
a
player
in
this
whole
arena
and
what
we
are
bringing
to
the
uas
commun
arena
so
having
the
the
architecture
structure
as
it
is,
is
very
powerful
for
outside
our
community,
and
that
is
a
document
that
I've
been
first
pointing
people
to
see
what
we
are
doing,
what
we
are
bringing
things
which
are
probably
going
to
be
coming
up
in
the
next
number
of
months.
C
In
discussions,
this
kind
of
becomes
a
key
document,
and
then
I
point
them
to
the
other
document
I
just
had
just
yesterday,
another
individual
representing
yet
another
community,
pointing
them
to
work
that
we're
doing
here.
So
we
need
to
look
at
this
that
we
are
with
this
document.
C
We
are
not
just
here
in
the
ietf,
we
are
the
ietf
as
an
sdo
participating
in
a
much
larger
community,
and
and
that's
why.
I
feel
that
the
structure
we
have
achieved
is
important
and
that's
all
I
want
to
say,
with
regard
to
keeping
four
five
and
six
as
they
are,
because
all
three
of
them
I've
been
pointing
people
too.
A
No,
who
is
next.
F
A
G
Okay,
okay,
I
only
have
one
quick
comment.
The
only
new
architectural
entities
that
drip
has
introduced
into
the
uas
rid
pre-existing
architecture
are
the
finders
and
the
crowdsourced
rid
supplemental
data
service
provider.
So
I
feel
that
if
the
architecture
document
is
supposed
to
talk
about
all
of
the
entities
in
our
architecture,
section
6
needs
to
remain
in
the
architecture
document.
D
So
first,
my
major
concern
was
about
this
using
this
mass
uppercase
and
basically
using
normative
language
into
an
informational
document.
You
have
put
everything
in
lowercase
you
may
want
to
remove
as
well.
Then
I
think
in
section
two,
you
still
keep.
Oh,
we
are
using,
what's
called
bcp14,
so
remove
it
now.
D
My
major
concern
is
is
solved.
Okay,
we
could
go
over,
but
I
think
it's
all
the
nice
way
about
the
section
four,
five
and
six.
I
think
they
describe
the
hit
a
hit
and
the
release
three
I'm
a
little
bit
concerned.
If
we
put
some
content
here
and
we
got
other
draft,
the
two
document
must
be
con
consistent.
D
If,
specifically,
if
the
akita
draft
is
going
forward
quickly-
and
it
seems
pretty
close,
a
good
last
call
and
then
we
change
our
mind
and
we
change
the
other
document
and
we
change
what
it
is
right.
That's
a
major
concern
now
if
we
are
confident
that
it's
not
a
concern,
we
can
go
forward,
but
my
major
concern
and
again
it's
simply
my
concern
and
my
concern
simply
to
ease
the
itf
last
call
and
the
isd
evaluation.
Yeah
right
just
provide
you
guidance,
nothing
was
blocking
there.
F
Thank
you
eric.
This
is
really,
I
would
say,
an
important
comment
to
to
hear
from
the
I
would
say,
the
editors
and
yeah,
and
we
need
to
work
out
how
to
better
articulate
with
it
tanks
in
the
architectural
document
so
that
it
can
be,
I
would
say,
a
base
reference
for
the
work
we
are
doing
and
I
would
say
the
the
main
entry
point
for
the
the
contribution
from
from
the
working
group
yeah.
F
So
so
should
I
please
please
take
note
of
that
and
we
will
discuss
how
to
to
handle
that
daniel.
Please.
A
Yeah,
so
I
I
think
we
are
all
heading
in
the
same
direction,
so
I
I
think
the
sections
four
five
six
are
important
and
as
mentioned
eric,
I
mean
we
should
not
have
something.
The
architecture
document
should
not
be
completely
abstracted
from
the
solution
space,
but
it
should
not
tie
too
much
to
it.
So
we
need
to
find
the
middle
ground
so
that
when
someone
reads
the
document
he
understands
what
an
h-I-t-e-h-I-I-t
is.
Why
we
use
that.
A
What
is
the
main
principle,
the
main
motivation
for
that,
how
it's
going
to
articulate
with
the
registries
and
so
on-
and
I
mean
the
the
discussion-
should
be
sufficiently
high
level
so
that
if
we
have
some
changes
in
six
months,
one
year,
it's
not
going
to
affect
the
full
architecture
so
which
I
mean
the
way
we
described
should
give
a
good
insight,
a
good
understanding
of
the
reader,
how
the
documents
are
going
to
be
articulated
and
not
the
document,
the
concepts.
A
And
then
I
mean
the
the
solution.
Space
documents
are
here
to
to
say:
oh
yeah,
in
the
architecture
dimension
yeah
we
have
a
smaller
radio
frame,
so
we
need
that
authentication
framework.
So
that
we
can
carry
the
rads,
oh
yeah,
we
we
need
to
be
able
to
bind
that
ri
er
er
our
id
to
to
some
information
in
registries
and
so
on
and
and
then,
if
we
use
dns
and
we
use
or
something
else,
it's
not
a
big.
I
mean
it
should
not
be
a
big
a
big
problem
later.
So
I.
D
D
D
F
Thank
you
eric.
Thank
you.
Thank
you
daniel,
so
the
next
one
is
adam,
and
then
we
will
close
the
the
line
just
right
after.
B
Yeah,
so
I'm
I'm
going
to
agree
with
daniel
on
this
and
eric
as
well
on
everything
that
was
just
stated.
B
I
think
the
best
frame
of
mind
to
walk
into
this
is
when
you
look
and
read
the
architecture
document,
it
is
general
enough
that
in
six
months
we
don't
have
to
completely
rewrite
it
we're
using
uas
remote
id
as
the
first
use
case.
There
might
be
a
use
case
in
a
year
from
now,
which
is
like
oh
we're
going
to
do
microcontrollers
and
iotx
devices.
B
B
I
I'm
torn
on
section
6
being
either
an
architecture
in
the
uas
right
draft,
I'm
not
quite
sure
where
it
should
go,
because
it
is
very
much
a
uas
remote
id
thing
because
that's
where
it
applies,
but
as
a
general
concept,
it's
good.
So
I
think
if
we
find
that
balance
that
daniel
and
eric
were
pointing
out,
I
think
that
would
be
the
best
way
forward.
C
Bob,
if
we
can
maybe
do
a
little
softing
of
the
words
and
say
this
is
technology,
which
is
the
way
that
you
want
to
look
at
solving
this
architecture
and
then
three
years
from
now,
there
may
be
other
technology
which
will
be
invented,
which
also
will
solve
this
fit
into
this
architecture.
C
Maybe
that's
a
way
to
address
some
of
these
things.
That's
saying
that,
well,
what
about
something
else?
Make
the
architecture,
use,
hhit
and
use.
The
registry,
as
here
are
ways
that
no
look
at
this
technology
technology,
as
this
is
how
it
can
be
done,
and
then
the
later
documents
say-
and
this
is
how
we
are
doing
it,
but
then
make
it
so
it's
it's
also
an
example
of
what
any
technology
needs
to
do
to
meet
the
requirements
or
meet
the
design.
A
Yeah,
for
example,
section
four
is
h8
it
for
the
rip
entity
identifiers.
I
think
the
section
should
be.
I
mean
the
title:
just
the
title
should
be
more
the
ident,
the
direct
identifiers,
and
then
we
explain
why
h,
I
t
is
a
valid
choice
and
how
it's
going
to
be
is
going
to
be
articulated
with.
C
Of
it,
I
don't
think
it
take
much
of
a
change
to
do
that,
we're
pretty
close
to
that
in
terms
of
wording
changing
the
title,
adding
maybe
another
sentence
and
something
another
another
sentence.
We
can
show
that
and
then
dive
into-
and
this
is
what
this
brings
to
the
architecture.
So
somebody
comes
with
some
other
great
idea
say:
well,
you
got
to
be
able
to
do
as
well
as
this,
if
not
better,.
C
E
All
right
I
mean
it's
that's
great,
so
this
is
a
great
comment.
I'm
I'm
so
glad
we
talked
about
that
and
this
is
really
make
things
clear.
I
just
want
to
echo
stool
first,
the
the
crossover
is
actually
not
new.
It
has
been
there.
We
just
you
know,
try
to
pick
it
and
put
in
the
66,
and
I
think
this
weird
piece
of
information
we
needed
to
put
in
the
architecture.
Thank
you.
This
is
only
one
piece,
information
bridged
broadcast
id
to
network
id,
and
this
is
for
future.
E
Looking
right,
even
though
fa
said
okay,
we
only
need
to
broadcast
it
right
now,
right
and
but
network
id
should
be
the
future.
So
that's
one
of
the
piece
information
we
should
keep
it
then,
regarding
the
section
two,
definitely
we
should
remove
the
nominal
working
and
wording,
and
I
think
I
I
thought
I
did
some
good
work
but
looks
like
I
missed
something
and
now
face
section
four
and
a
five
being
consistent
with
the
the
solution
speech.
That's
something
we
really
need
to
be
careful.
E
I
think
daniel
put
a
greedy
comment
in
the
in
in
the
github,
which
is
really
regarding
what
kind
of
level
the
details
we
should
put
in
there.
I
think
we
needed
to
the
co-authors
needed
to
actually
work
closely
to
identify
things.
We
are
for
sure
won't
be
changed
at
least
in
half
a
year
a
couple
years,
but
if
astm
decided
okay,
we
want
to
change
the
buy
numbers
right.
E
Yes,
change
from
200
to
300,
then
that's
something
we
are
actually
out
of
control,
section
four
and
then
I
think
the
last
we
have
just
said
we
should
change
the
title
right
and
then
you
missed
that
already
we
should
be.
I
think
we
should
be
good
to
go,
as
I
think
last
email
we
see
from
daniel.
A
So,
just
to
sum
up,
we
we're
I
mean
we're
quite
close
to
the
end
of
the
tunnel,
so
no
big
mess
in
that
document,
but
that's
the
discussion
worth
with
it.
So,
let's
move
to
the
next
presentation.
Unless
we
have
any
other
comments.
H
H
So
first,
a
few
slides
are
going
to
be
the
same
as
last
time,
except
that
they
have
a
bit
more
time
to
to
go
from
from
them
in
the
detail
and
then
I'm
going
to
jump
to
the
actual
implementation
status.
H
So
the
starting
point
for
our
code
was
open,
hip
version
two
which
is
available
on
bitbucket
and
the
the
draft
was
a
new
crypto06
from
bob.
H
But
I
think
current
one
is
zero
and
nine,
but
I
think
actual
code
now
supports
zero,
seven,
zero,
eight
something
in
in
in
between
and
for
students
working
on
that
code,
because
it's
prerequisite
for
for
drone
id
and
then
the
actual
code
for
for
for
drone
id
was
with
open
drone
id
project,
and
this
was
part
of
an
advanced
project
on
secure
systems
in
in
shopping
university,
which
was
run
last
fall
and
we
are
going
to
have
it
again
next
year.
So
in
next
fall
we
can
hope
to
get
to
another
level.
H
So
if
you
know
the
part
of
ipv6
address,
space
is
reserved
for
things
like
host
identity
protocol
and
also
mobile
ip,
so
it's
needed
to
implement
an
extension
to
orchid
to
have
so-called
hierarchical
hits.
H
So
in
the
beginning,
the
heat
was
just
sort
of
128
bit
hash
when
there
was
some
prefix
added
to
it
and
later
there
is
a
need
to
have
a
few
other
intentions
to
have
some
hierarchy
so
like
who
actually
issued
this
hit
host
identity
tag,
which
is
a
hash
of
a
host
public
key
next
slide.
Please.
H
So
orcid
is
essentially
a
combination
of
a
few
things
like
in
traditional
orchid.
It
was
a
prefix,
then
orchid
generation
algorithm
id.
So
it's
a
type
of
hash
function
which
is
used
to
generate
it
like
sha2
or
256
or
shafri,
and
so
on,
and
then
the
prefix,
which
is
a
part
of
ipv6,
address
space
which
is
reserved
by
ayana
and
the
final
hash,
which
is
96
bit
long,
which
is
a
hash
of
a
public
key
of
a
host
identity.
H
D
H
Of
course,
it
makes
itself
shorter,
which
increases
probability
of
collisions,
so
it's
not
maybe
anymore
guaranteed
that
two
hosts
can
have
statistically
unique
hits,
so
they
need
maybe
increased
need
for
heat
collision
detection
and
resolution
next,
please.
H
H
So
for
some
lightweight
things
such
as
iot
and
drones,
we
need
the
algorithm,
which
is
fast
but
also
has
very
small
signature
size
and
also
that's
the
reason
why
we
need
something
better
than
advanced
encryption
standard
for
symmetric
crypto,
because
this
chaor
cipher
is
a
more
lightweight
alternative
and
in
the
current
code
we
also
starting
to
move
to
so-called
hoodiac
cipher,
which
is
a
candidate
for
lightweight
crypto
competition
in
nist.
H
But
it's
not
kind
of
clear
if
it's
gonna
win
or
not.
So
we
are
following
the
situation
and
we're
updating
the
code
to
to
hudiak
support
at
the
moment
and
final
thing:
is
this
key
derivation
function
because
it's
this
key
key
ddf
is
more
efficient
than
the
other
previous
functions
so
again
for
lightweight
devices
such
as
drones
running
on
battery?
It's
important
factor
next.
H
Also,
the
new
arcades
based
on
hierarchical
hits
I
implemented,
and
the
key
derivation
function
based
on
kayak,
so
kkdf
and
kiak
is
implemented
as
a
cipher
lightweight
site
for
alternative,
as
well
as
codyak
is
in
the
process,
and
it
was
tested
in
the
core
simulator
that
this
basic
change
with
heap
version
2
works
fine
and
succeeds.
So
the
code
is
kind
of
ready
for
use
next
slide.
Please.
H
So,
as
we
know,
drip
is
based
on
using
hip
version,
2
heat,
so
basically
broadcasting
it
over
bluetooth
or
four
and
five
or
wifi,
like
maybe
in
traditional
mode
or
in
so-called
nein
board.
Neighborhood
awareness
node.
H
H
But
the
latest
drone
id
seminar
in
in
europe
I
attended.
There
is
actually
already
several
companies
implementing
those
add-on
boxes
to
comply
with
eu
laws.
So
it
seems
to
be
also
an
option
just
to
buy
this
small
box
and
stick
it
to
your
old
drone
and
you
become
kind
of
compatible
and
legal
and
the
challenge
is
the
size.
H
So
in
bluetooth
forward
is
like
only
20
bytes,
so
it's
kind
of
hard
to
to
fit
things
into
it,
and
also
there
was
a
receiver
android
in
android
application
developed
so
that
the
observer
and
the
policemen
can
see
which
drones
are
flying
nearby.
So
it
receives
this
broadcast
information
and
and
displays
it
on
the
screen.
H
H
So
the
current
proposals
like
in
eu
and
sdsm,
we
have
no
no
trust
in
broadcast
messages,
so
if
drip
is
supported
when
we
basically
broadcast
this
hierarchical
hit,
which
is
generated
based
on
our
orchid
from
a
public
key
of
a
drone
and
the
implementation
that
I'm
going
to
talk
about
is
based
on
this
draft
version
from
I
think
end
of
october.
So
we
have
the
architecture
and
the
protocol
draft.
So
now
I
guess
we
are
a
bit
forward,
but
the
basic
things
should
still
work.
H
H
Slide
so
the
current
status
is
that
the
application
is
implemented
on
android
phone
and
tested,
and
it
can
use
like
google
maps
to
display
the
drone
coordinates
on
a
map,
and
this
information
is
fetched
from
a
back-end
server,
which
is
a
kind
of
alpha
implementation
of
heap
registry
which
drip
registry,
which
things
has
things
like
operator,
id
position
and
so
on,
and
on
raspberry
pi.
So
rpa
is
raspberry
pi.
There
is
a
script
that
basically
implements
drip
in
a
very
simple
way.
H
So
what
it
does
it
downloads
latest
open,
hip
version,
installs
generates
public
key
heat
and
then
uses
another
script
to
broadcast
it
over
bluetooth,
so
kind
of
very
simple
way
to
implement
drip
next
slide.
H
So
basically,
there
are
two
scripts,
one
that
configures
raspberry
pi
to
have
a
correct
version
of
openssl
because
well
actually
it
was
a
huge
job
to
upgrade
open
heap
to
support
open,
ssl
reasonably
late
version,
because
there
were
a
lot
of
changes,
but
still
it's
not
like
the
latest
version.
So
we
need
to
use
configure
that
and
the
other
script
actually
takes
a
hit
from
the
file
system
and
then
broadcast
it
using
bluetooth
frames
next
slide.
H
So
this
is
a
format
of
bluetooth,
4
standard.
So,
as
you
you
can
see,
the
payload
is
not
big,
something
like
20,
bytes
and
well.
Heat
is
four
bytes,
so
it's
not
like
that
big.
But
if
you
need
additional
things
like
signature
and
so
on,
which
are
part
of
the
drip
it's
like,
maybe
not
enough,
but
for
bluetooth
5,
there
is
going
to
be
increased
sizes
so
and
hopefully
it's
going
to
work,
but
even
for
raspberry
pi,
it
says
it
implements
bluetooth.
Five.
H
It's
actually
only
implements
limited
features,
part
of
it,
but
now
we
acquired
also
usb
sticks
that
have
full
bluetooth,
five
support.
So
hopefully,
in
autumn
we
can
experiment
more
with
that
part.
Next
slide.
H
And
since
bluetooth
standards
were
reasonably
compatible,
it's
kind
of
easy
to
move
between
versions
and
easy
to
use
with
just
few
lines
of
code,
and
there
is
a
good
support
for
android
for
scanning
scanning,
bluetooth
bacon.
So
it
was
relatively
easy
to
implement
this
application
and
some
preliminary
tests
that
indicate
that
bluetooth,
4
is
roughly
100
meters
and
0.25.
Maybe
half
a
kilometer,
so
much
larger
range
and
also
the
payload
is
much
bigger
in
bluetooth
5.
H
So
it's
going
to
be
more
compatible
and
better
implementation
next
slide,
and
this
is
actually
the
funny
script
which
implements
drip
in
in
raspberry
pi.
So,
basically,
you
first
install
write
libraries
when
you
install
write,
open,
ssl
version
when
you
install
openhip
and
in
the
end
you
call
a
navascript
which
just
takes
hit
from
a
file
system
and
sends
it
via
bluetooth.
So
it's
kind
of
simple
and
hacky
way
to
implement
basic
drip
next
slide
and
the
backend
kind
of
alpha
version
of
heap
registry.
H
It
was
based
on
this
web
technologies,
which
has
some
database,
which
maps
hit
to
things
like
position
of
operator,
owner
name,
model
of
aircraft
and
so
on.
So
it's
kind
of
a
back-end
which
is
used
for
database
next
slide.
H
So
this
application
was
based
on
on
java
and
cradle
and
using
google
maps
api
to
actually
place
the
markers
and
download
the
geographical
map
picture
office
area
next
slide,
so
this
screenshots
from
application,
so
it
for
example.
Here
it
shows
like
six
drones
in
this
area
with
certain
ids
which
are
being
broadcast,
and
you
see
them
also
mapped
on
the
geographical
map
of
lean
shopping
being
placed
somewhere
on
the
map
next
slide.
H
H
So
it's
a
bit
limited
because
it's
based
on
bluetooth
four
but,
as
I
said
we
plan
to
go
to
five.
Also,
we
didn't
have
like
a
device
which
is
supporting
wi-fi,
neighborhood
awareness,
node
mode.
I
think
raspberry
pi's.
Don't
don't
do
that
so,
but
I
think
latest
devices
from
like
galaxy,
maybe
s10
already
does
it
so
maybe
in
the
autumn
we
try
to
get
this
kind
of
phone
and
then
experiment
with
nin
mode,
because
it's
much
better
way,
you
don't
need
to
root
your
mobile
phone
to
actually
implement
it.
H
So
it's
much
easier
for,
like
a
normal
user,
just
to
download
the
application
from
a
store
and
then
started
using
it
instead
of
turning
turning
phone
into
a
hacking
mode,
next
yeah,
so
bluetooth,
five
is
on
our
to-do
list
and
nin
mode
and
then
adding
certificates
to
a
broadcast
so
that
the
hit
is
authenticated
and
maybe
provide
some
better
visualization
in
the
application.
For
example,
maybe
have
a
marker
reflect
somehow
size
of
a
drone
and
speed
or
some
how
dangerous?
H
Is
it
or
something
like
that,
so
that,
for
example,
users
can
have
a
better
view
of
what's
happening
in
the
area
and
what
kind
of
drones
are
flying
there
next.
So
to
conclude,
there
was
some
simple
version
of
drip
implemented
based
on
heap
version,
2
extensions
and
yeah.
Some
problems
were
faced
with
limited
support
of
bluetooth,
5
and
wi-fi
mode
in
inside
the
raspberry
pi,
but
in
future
we
hope
to
to
use
the
latest
sticks
and
usb
sticks
and
maybe
latest
smartphones,
to
have
better
and
more
more
full
implementation.
H
So
if
there
are
any
any
questions,
I
can
answer
yeah
and
just
to
add
that
I
went
for
the
trouble
to
become
official
drone
pilot
in
eu.
So
I
had
to
read
like
200
pages
of
documentation
and
pass
two
hours
online
exam,
so
it
was
actually
pretty
complex.
I
mean
like
more
difficult
than
getting
in
driving
license.
It
seems
so
not
not
too
many
people
are
flying
drones
nowadays
in
europe.
I
guess.
A
I'm
hearing
stuart.
G
I'm
wondering
when
we
might
consider
doing
a
plug
fest
to
check
interoperability
of
the
independently
developed
implementations
that
you
have
there
at
the
university
and
that
we
have
at
the
new
york
uas
test
site,
I'm
not
in
a
big
rush
to
do
that.
I
think
we've
got
some
work
to
do
before
we
can,
but
I
think
it
would
be
a
great
thing
to
do
at
some
point.
H
A
I
mean
the
draft
I
mean,
I
mean
it
would
provide
some
additional
confidentiality
that
the
specifications
are
working
and
I
think
that's
gonna
ease
the
process.
I
don't
know
how
much
is
required,
but
I'm
letting
eric
to
jump
on.
A
D
Yeah
I
was
there,
I
was
six
simply
taking
time
to
find
back
the
right
button
on
the
right
window.
Yeah.
Sometimes
it
takes
so
now.
I
understand
you
asked
me
a
question
like:
can
you
rephrase
it
danielle.
A
Okay,
so
we
were
talking
about
interoperability
because
we
currently
we
know
there
are
two
ongoing
implementations
of
the
specifications
and
the
question
that
I
was
asking
is
how
much
having
an
interoperability
before
moving
the
specification
forward
is
helpful
or
not.
D
It's
always
helpful
right,
so
in
some
working
groups
within
routing
area
there
we
try
absolutely
to
have
two
interpretable
implementation
before
going
to
working
group
class
call.
Now
we
are
not
that
far,
but
proving
that
we
have
two
interval
implementation
of
usually
makes
things
way
easier
after
right.
So
you
don't
need
to
document
it.
D
H
C
This
makes
it
with
our
virtual
work,
challenging
I
mean
potentially
andre.
Could
ship
code
to
adam
and
adam
could
put
some
stuff
together
and
and
see,
maybe
that
he
can
actually
run
andre's
code
there
at
the
new
york
test
site,
but
until
we
get
to
san
francisco
or
potentially,
we
could
have
a
physical
hackathon
it's.
F
Yeah
that
that
was
the
assumption
bob
thank
you
should
I
please.
E
Yeah,
I
think
netapp
is
great,
but
we're
really
not
there
yet
right
I
mean
I
think
the
first
step
is
really
to
find
the
people
find
the
time
come
up
with
a
open
platform
that
everybody
can
experience.
What
triple
looks
like
right.
I
mean
everybody
can
get
a
drone.
Now,
it's
not
that
hard
to
to
to
to
experience
right,
there's
so
many
open
source
code
bases
like
open,
jewel
id
and
some
cheap
bluetooth
module
module.
People
can
buy.
E
I
mean
if
we
can
come
up
with
some
really
basic
code
base
for
people
to
start
with.
That
would
be
fantastic
to
help
drip
move
forward
all
right.
Thank
you.
C
C
C
These
are
supposed
to
be
performance-based
rules.
They
expect
sdos
to
now
actually
say
how
it's
done
and
then
produce
an
mlc.
That's
not
what
itf
to
do
but
are
kind
of
like
inserting
into
the
process.
C
An
important
thing
which
is
part
of
these
final
rules
is
that
they
remove
the
network
remote
id
and
I
won't
go
into
the
whys
and
where
force
and
where
and
how
that
may
change
in
the
future,
but
they
added
something
they
called
the
remote
id
module
for
upgrading
non-compliant
existing
uas,
and
this
is
going
to
be
important,
and
particularly
what
andre
has
built
is
a
remote
id
module.
It's
not
what's
controlling
the
plane,
and
this
is
an
important
distinction
in
the
faa
and
also
is
coming
out
in
in
europe.
C
Potentially
as
well
point
is.
We
now
have
a
target
deadline,
supposedly
when
vendors
are
going
to
be
wanting
to
put
create
code
and
put
product
out,
we
just
received
a
two
and
a
half
month,
extension
of
the
deadline
that
came
out
just
a
couple
days
ago
from
the
faa,
where
they,
because
of
reviews,
because
of
the
new
administration,
everything
being
extended,
so
there's
been
a
two
and
a
half
month,
push
out
of
deadlines,
but
nonetheless
we
now
have
final
rules.
We
now
have
deadlines.
C
We
now
have
things
to
target
too
and
and,
like
I
said,
the
big
thing
now
is
these
remote
id
modules
are
firmly
defined
in
the
final
rules.
You
can
see
them.
If
you
just
go
to
the
summary
you
don't
have
to
go
through
the
whole
rules.
Just
look
at
the
executive,
summary
and
you'll
see
what
the
what's
about
these
modules
being
able
to
snap
them
onto
a
ua
to
be
able
to
participate
in
mode
id
next
slide.
Please.
C
And
because
of
these
modules,
the
faa
says
that
these
modules
cannot
use
session
ids,
whereas
inside
the
ua
you
can
have
session
ids.
They
can
only
use
the
serial
number
of
the
remote
module,
not
the
serial
number
the
ua,
but
these
remote
modules
have
their
own
serial
number,
which
is
encoded
in
the
cta
2063a
format.
Only
that's
the
way
the
rules
written
right
now
and
doesn't
leave
much
wiggle
room
from
discussions
that
I've
been
participating
in,
don't
expect
faa
to
change
the
rules.
Regarding
that,
so
what
did
I
do?
C
I
came
with
the
way
that
hierarchical
hits
can
be
encoded
in
the
cta
2063
format,
so
that
these
modules
could
benefit
from
hierarchical,
hits
and
and
be
able
to
use
them
and
to
be
able
to
use
the
auth,
though
this
would
not
be
a
session
id
that
can
be
in
theory,
change
for
each
operation,
but
it's
rather
set
in
it,
but
the
potentially
the
manufacturer
could
use
this
for
for
creating
their
serial
numbers
and
then
benefit
from
the
rest
of
it.
So
that
is
an
important
addition
and
why
I
added
it.
C
I
added
to
meet
the
faa
definition
of
the
rid
modules,
so
take
a
look
at
that
and
understand
why
I
did
that.
The
other
thing
that
I
did
per
our
chairs
is.
I
took
the
off
that
the
auth
draft
and
I
merged
it
in
creating
some
new
definitions
and
creating
appendix
f.
So
I
pretty
much
lifted
what
adam
had
in
off
zero
six
did
some
massaging
of
it
and
there's
appendix
f,
so
you
should
see
that
I've
now
completed.
I've
done
things
there
and
I've
done
that
merge
in
those
are
the.
C
If
you
do
a
diff
between
version
four
and
version
nine,
I
think
our
current
version
is
nine
or
six
seven
or
whatever
you'll
see
that
those
are
the
only
substantial
changes
in
the
document
move
on
to
the
next
slide.
Please.
C
We
can
expect
to
have
an
id
type
allocated
for
hierarchical,
hits,
that's
pretty
much
a
no-brainer
and
agreed
upon
our
requesting
changes,
the
authentication
menu.
Those
changes
are
in.
The
queue
thing
to
note
is
that
the
faa
removed
all
cyber
security
from
the
final
rules.
The
preliminary
rules
had
a
lot
of
cyber
security,
there's
nothing
about
cyber
security
in
the
final
rules,
so
people
are
asked.
Why
is
this
message
now
needed?
I
won't
go
into
why
it's
kind
of
like
future
proofing
areas.
C
We
can
expect
a
resolution
on
all
this
from
astm
f3802
by
the
end
of
march
and
having
to
know
know
what
our
in
a
sense
marching
orders.
What
can
we
work
with,
but
we
can
we
can
work
with.
Yes,
there
will
be
an
id
type
allocated
particle
hits
and
there
will
be
some
changes
made
authentication
method
message
such
that
what
we
are
doing
will
fit
into
the
our
proposal
will
be
accepted
in
some
form.
A
Okay,
I
was
just
wondering
I
didn't
get
why
cyber
security
has
been
removed.
C
You'd
have
to
go
through
all
the
comments
that
and
the
responses
that
in
the
the
final
rules,
the
faa
has
a
whole
set
of
discussions.
Here
are
the
comments
that
you
received
and
here's
a
various
companies
positions
on
this,
some
foreign,
some
against
and
here's
the
action
we
took.
So
if
you
want
to
understand
why
they
removed
it,
it
was
basically
either
vendors
saying
that
this
is
too
hard
too
much
of
a
demand
on
our
processor.
Oh,
this
is
only
within
radial
range.
C
Why
are
you
concerned
about
this
because
you
have
security,
because
the
range
is
only
a
half
per
kilometer
per
the
the
bluetooth
five,
so
it's
kind
of
cute
stuff
and
and
there's
like-
I
said
if
you
want
to
understand-
why
look
at
the
final
rules
and
look
at
at
the
fa
was
pretty
good
at
including
that
debate,
and
I
don't
want
it
to
spend
any
time.
It
is
what.
B
C
We're
we're
kind
of
like
kind
of
coming
out,
saying
that
all
this
discu
I'll
make,
I
will
say
something.
What
we're
doing
here
is
saying
that,
yes,
we
had
all
this
discussion,
because
people
didn't
know
how
to
do
it
here.
We
are
an
sdo
saying,
and
here
is
how
it
can
be
done
with
inside
the
capacity
of
these
devices.
C
So
we're
sticking
our
foot
in
the
in
the
fire
here
in
a
sense.
C
As
an
aside-
and
I
posted
this
to
the
110
mailing
list
that
you
may
have
seen-
is
that
anyone
here
involved
in
e
911
location
accuracy,
because
the
faa
wants
a
15
foot
altitude
at
95
percent
for
the
ground,
control
stations,
and
discussion
has
been
that
this
may
not
be
possible
with
current,
affordable
tech.
C
I
won't
go
for
you
with
other
debate
items,
but
this
particular
one.
I
I've
been
asked
to
go
out
on
a
fishing
expedition
to
find
out
if
there's
people
elsewhere,
who
are
familiar
with
with
e911
and
how
the
the
toggle
community
is
meeting
the
elevation
altitude
requirement
of
e911,
because
faa
wants
to
do
the
same
for
remote
id
so
that,
for
instance,
the
first
responder
can
know
that
the
operator
is
on
the
third
floor
balcony,
not
that
operator
on
the
fourth
floor
balcony.
C
So
it
looks
like
that,
if
you
look
at
the
gcs
documents
themselves
at
jesus.gov,
they're,
saying
plus
or
minus
13
meters,
so
it's
I'm
reaching
out
for
people
who
are
who
have
any
experience
in
this,
and
I
have
gotten
a
couple
of
you
respond.
C
So
this
just
just
aside
so,
but
what
still
needs
to
be
done
and
the
big
thing
is,
I
need
to
merge
what
I
did
in
appendix
e
into
what's
in
appendix
f,
and
I'm
pretty
much
done
that,
but
I
need
to
resolve
this
replay
issue
and
I'm
interested
in
inputting
into
this
ideal,
appendix
e
and
replay
one
way
appendix
f
really
does
not
address
replay.
C
What
is
I
have
the
time
stamp
I
have
in
appendix
e
is
the
time
that
it
was
created
and
expecting
the
the
observer
to
not
trust
to
be
on
a
certain
period
of
time.
Appendix
f
has
a
use
by
date,
but
it
allows
that
to
be
too
large.
C
I
will
be
posting
to
the
list.
My
thoughts
on
this.
I
think
what
it's
going
to
be
is
that
in
the
product
using
an
attestation
in
a
broadcast
mode,
I
use
an
authentication
message
that
that
that
the
time
stamp
is
a
future
out,
no
more
than
one
minute
that
this
particular
attestation
could
only
be
trusted
for
for
a
very
short
period
of
time,
and
that's
why
I
need
to
say
to
protect
against
the
replay
attack.
Now.
C
We've
posted
to
the
list-
and
this
is
this
important
discussion,
because
if
we
are
supplying
trust
here
and
we
want
to
protect
against
fraudulent
use,
we
need
to
be
able
to
constrain
this.
I
see
russ
is
in
the
queue
and
and
stewart.
So,
let's,
let's
press
this,
the
q
before
I
go.
I
Hi,
so
the
cell
phone
industry
association
did
a
911
accuracy
test
for
location
and
if
you
want
to
get
in
touch
with
those
people,
I
can
introduce
you.
C
Thank
you.
Please
do
because
it'd
be
interesting
to
see
how
they
achieved
it.
The
the
accuracy
that
is
required,
which
is
three
meters
and
the
faa,
is
kind
of
willing
to
settle
for
four
meters.
Do
they
do
after
three
meters.
I
Well,
the
concern
here
was,
if
you're
dealing
with
a
metropolis,
you
know
the
first
responders
go
to
the
wrong
floor.
You've
cost
a
lot
of
time.
C
Exactly
it
is
a
different
use
case.
I
understand
the
use
case
and
our
use
case
is
the
they're
trying
to
find
the
operator
for
whatever
reason.
G
Is
my
audio
working
this
time?
Yeah?
Okay?
So
just
a
quick
answer
on
the:
why
did
cyber
security
get
entirely
dropped
because
they
dropped
for
the
moment,
and
they
were
clear
that
this
was
only
for
the
moment
network
remote
id
and
retained
only
broadcast
remote
id
and
further,
because
broadcast
remote
id
does
not
even
require
a
two-way
bluetooth
or
wi-fi
link.
It
can
be
simply
a
broadcast
from
a
transmitter
on
the
aircraft
that
doesn't
even
have
receive
capability.
G
C
So
the
last,
the
next
point
is,
is
I
need
to
fix
the
dns
examples.
Stuart
and
adam
have
given
me
some
some
content
of
what
I
need
to
put
in
there.
So
I
need
to
work
on
that.
One
of
the
things
here
is
that
we
are
looking
at
some
of
the
stuff
putting
it
under
some
aerial
tld,
but
I
have
talked
to
two
different
companies
that
say
these
hardcore
hits
are
so
interesting.
C
We
can
use
them
what
we
want
to
do
and
they're
not
aviation
related
at
all.
So
that's
why
I
kind
of
would
like
to
see
if
we
can
find
a
home
for
the
dns
structure
for
hydro
hits,
which
is
not
in
aeronautics
and
allows
that
some
of
the
space
to
be
used
in
aeronautics,
but
some
of
the
space
potentially
otherwise
each
group
is
going
to
want
going
to
anna
for
their
own
prefix,
which
is
not
good.
A
C
Happy
to
be
able
to
have
dns
be
able
to
look
up
for
fine
stuff
and
they
and
and
they
want
to
be
able
to
have
harco,
hits
work
for
them
as
well
and
that
they
have
their
raa
that
they
will
have
for
the
their
for
for
their
community.
C
So
it's
like
that's
what
I
hope
would
have
happened
and
and
and
that
I'm
in
conversation
with
with
one
company
about
about
what
they're
they're
doing
so
that's
I
need
to
allow
for
that.
I'll
be
posting
some
of
this
to
the
list
coming
up,
but
those
are
the
only
two
areas
left
right
now,
based
on
any
changes
that
discussion
that
on
architecture
will
get
to
me
that
those
are
two
things
I
need
to
fix,
based
on
comments,
I've
received
and
then
we're
open
for
last
call,
and
I
need
to
get
that
out.
C
A
B
Hello,
you
guys
hear
me
all
right,
yeah,
okay
cool,
so
I
will
actually
have
michael
pelage
also
be
joining
me
today,
midway
through
this
presentation,
but
I'll
be
starting
it
off.
So
these
are
some
draft
updates
on
the
drafts
that
are
not
on
working
group
last
call,
so
the
authentication
format
was
adopted
on
the
18th
of
december
last
year.
B
The
document
is
up
on
github.
I
moved
it
up
there
so
that
it
was
publicly
available
it's
in
markdown,
so
you
guys
can
go
and
look
at
it
post
comments
pr
requests
issues.
However,
you
feel
comfortable
once
we
get
a
rhythm
with
architecture,
I
think
it'll
just
naturally
flow
into
this
document.
The
document
is
currently
on
a
hold.
B
This
is
just
due
to,
as
bob
was
mentioning
in
his
last
presentation.
There
are
some
f-3411
updates
that
are
coming
up
in
the
authentication
message.
Two
of
them
are
the
fec.
Some
of
our
fec
work
might
be
integrated
into
astm
itself,
which
negates
the
need
for
it
to
be
in
our
draft.
So
I
need
to
see
where
that
ends
up
landing
and
also
to
harmonize
a
bit
better.
B
There's
some
discussion
on
an
auth
type
and
a
pseudo
drip
header
of
some
capacity
in
the
message
itself,
but
those
are
more
astm
specific,
but
in
general
we
are
in
queue
and
once
the
changes
are
confirmed,
we
will
update
the
draft
accordingly.
So
the
also
uploaded
actually
earlier
last
month
was
the
registries
draft.
So
this
is
just
a
stump
of
previous
work.
When
we
broke
apart,
some
other
drafts,
the
identity
claims
draft
was
ripped
apart
and
some
bits
went
to
bob's
draft
some
bits
went
to
architecture.
B
So
this
is
where
it
now
sitting
the
main
thing
that
this
draft
does
hold
is
the
provisioning
process
a
somewhat
detailed
one
for
interactions
between
the
parties
to
get
ids
registered
into
dns,
at
least
on
a
high
level.
So
we
just
need
some
word
spitting.
I
think
on
that
side
and
figure
out
where
it
fits
in
the
draft.
B
I
do
have
some
missing
sections:
how
the
id
is
added
and
updated
to
link
to
a
serial
number
via
probably
epp,
how
the
id
is
looked
up
to
get
a
serial
number,
but
only
by
authorized
parties
using
rdap
and
exactly
what
fields
need
to
be
in
the
registry
and
me
and
michael
palace
have
been
working
in
the
background.
B
J
Hear
me
we
can
hear
you
yeah
excellent,
so
what
I'd
like
to
do
is
give
an
update
on
what
we've
been
up
to
since
ietf
108,
when
I
first
presented
the
concept
of
what
the
regis
potential
registry
would
look
like,
and
what
what
I've
done
is
in
the
context
of
another
gtld.music
that
we'll
be
launching.
J
We
have
actually
come
up
with
a
full
registry
sandbox
that
deals
with
authenticated
users
and
also
uses
rdap
to
provide
differentiated
access
to
data
about
the
contact
droids
stored
in
the
registry.
So
these
are
important
concepts
that
we
think
are
important
in
the
broader
drips
discussion,
because
there
clearly
needs
to
be
some
account
of
data
privacy
considerations,
particularly
in
europe
with
the
gdpr,
as
well
as
balancing
those
with
the
needs
of
first
responders
and
law
enforcement.
J
We
will
be
going
public
with
this
within
icann
and
one
later
this
month,
and
then
what
adam
and
I
bob
and
still
be
doing,
is
looking
at
how
to
take
that
code
and
begin
to
modify
that
to
handle
a
potential
mock-up
of
what
a
drone
registry
would
look
like,
as
adam
had
just
mentioned
before,
we
are
beginning
to
look
at
what
those
different
con
profiles
of
those
contact
droids
or
grips
what
they,
what
the
individual
data
elements
would
need
to
contain
and
the
potential
access
those
data
elements
by
different
requesters
within
the
ecosystem.
J
One
other
thing
that
I
will
be
touching
upon
in
this
presentation
today
is
some
recent
lego
and
regulatory
update
from
europe,
particularly
in
the
area
of
cyber
security.
So
while
the
faa
has
temporarily
hunt
it
on
cyber
security,
interestingly
enough,
the
european
union,
as
part
of
their
nis
ii
initiative,
actually
has
become
specific
enumerated
areas
which
potentially
intersect
with
the
work
of
of
this
group.
Next
slide.
Please.
J
So
my
apologies
for
the
non-ascii
diagram
here,
but
in
the
interest
of
time
it
was,
it
was
a
lot
easier.
This
was
this
was
the
original
concept
that
I
shared
with
the
group
during
ietf
108
and
the
concept
that
the
main
concept
that
we
were
trying
to
convey
with
this
diagram
here
was
the
idea
of
a
registry
providing
different,
authoritative
sources
in
a
distributed
manner.
Recognizing
that
there
were,
we
were
not
trying
to
create
a
single
mother
of
all
databases
to
hold
everything.
J
We
realized
that,
in
order
to
meet
a
certain
national
requirements,
there
would
need
to
be
a
distribution
of
certain
data
sources
and
how
that
data
would
be
done
in
an
authoritative
way,
but
then
allow
the
contact
droids
move
seamlessly
between
the
the
different
participants
within
the
ecosystem.
J
J
J
J
What
we
were
proposing
here
was
using
sort
of
mimicking
the
iso
3166
ii
character
using
that
as
a
second
level
and
then
allowing
a
local
registry
in
country
to
store
that
particular
operator
pilot
information,
but
then
because
the
drones
would
be
written
globally,
that
local
registry
would
be
able
to
have
the
ability
to
take
that
contact
droid
and
have
access
to
it.
So
that
is
what
we
are
trying
to
convey
and
if
I
could
go
to
the
next
slide,
real
quick
I'll
I've
actually
come
up
with.
J
What
I
think
is
a
good
example
to
explain
how
this
differentiation
could
potentially
work.
So
for
those
people
that
may
be
familiar,
the
dot
co
is
the
colombian
country
code
top
level
domain.
Although
this
is
a
cctld
dot,
co
has
been
marketed
as
a
hybrid.
J
It
is
recognized
in
country
as
serving
the
colombian
local
internet
community,
but
then
is
also
broadly
marketed
as
an
alternate
gtld
to
the
broader
internet
community
and
what's
interesting
here
when
you
look
at
the
particular
ecosystem
of
what
they've
done
is
there
are
one
set
of
registrars
and
resellers
that
sell
second
level
domains
at
the
at
the
second
level
in
dot
co.
But
then
there
are,
if
you
will
third
level
registrations
occurring
in
dot,
org
and
dot
com
in
the
dot
co
space.
J
Those
those
registrations
are
again
done
through
a
series
of
competitive
registrars
and
resellers.
So
this
is
where
you
would
see
the
uss
and
other
people
providing
a
dynamic
marketplace
to
provide
these
services.
The
other
example
I
use
here
to
show
a
perhaps
more
restrictive
access
within
this
hierarchical
namespace
is
mill
dot
co.
This
is
there
are.
There
are
third
level
domains
registered
in
mill,
dot,
co
that
are
not
available
to
commercial
registrar
to
provide
those
services.
J
Those
are
in
fact
issued
by
the
government
itself,
so
this
was
kind
of
an
example
of
a
parallel
when
some
people
were
trying
to
wrap
their
arms
around
the
the
ability
of
having
different,
authoritative
data,
if
you
will
federated
but
interoperable
within
an
ecosystem.
We
thought
that
this
is
again
one
example:
a
a
potential
analog
moving
forward
to
the
next
slide,
and
this
was
also
something
that
was.
This
was
a
table
that
was
presented
in
iatf
108.
J
this
in
trying
to,
if
you
will
explain
to
regulators
and
other
people
that
may
not
be
as
intimately
involved
in
the
the
drone
space.
We
have
talked
about
this
concept
of
vin
numbers.
I
believe
one
question
about
globally
unique,
so
much
like
vin
numbers
are,
you
know
when
a
when
a
car
rolls
off
of
a
manufacturing
assembly
line
somewhere
in
the
world?
There
is
a
vin
number
associated
with
that
vehicle
and
that
vin
number
is
globally
accessible.
So
that
would
be
something
that
every
participant
in
the
ecosystem.
J
They
would
require
uniqueness,
as
well
as
access
to
that
identifier.
Then,
when
you
get
into
the
individual
license
plates,
which
would
be
affixed
to
the
drones
or
the
individual
drone
operator
of
credentialing
or
licensing,
that
would
be
the
equivalent
of
the
driver's
license.
So
these
are
some
of
the
authoritative
data
sources
that
a
registry
will
need
to
track
to
allow
the
utms
and
the
uss
to
begin
to
exchange
data
in
an
interoperable
and
in
a
unique
manner.
J
So
what
I,
as
again
adam,
had
touched
on
this
before
we
have
begun
to
go
through
some
of
the
different
data
elements
and
what
are
what
are
those
profiles
of
of
the
different
voids
within
the
ecosystem?
So
what
are
those
repository
object,
identifiers
that
are
going
to
be
exchanged
within
the
epp
registry
and
what
are
the
underlying
data
elements
that
are
going
to
be
necessary?
J
So
this
is
some
of
the,
as
adam
had
alluded
to
what
we've
been
doing
over
the
past
three
weeks
is
trying
to
sketch
out
what
the
minimum
voids
would
be
for
an
initial
registry
deployment
and
then
expanding
that
as
we
begin
to
get
a
fuller
view
of
who
all
the
participants
are
in
the
ecosystem.
B
Okay,
so
I'm
actually
going
to
take
over
for
here.
So
this
is
a
diagram,
I'm
I'm
using
ascii
art.
So
sorry,
michael,
I'm
stepping
above
you
a
little
bit
in
classiness.
So
this
is
a
registry
key
value
and
lookup
link
diagram.
So
there
are
five
main
registry
key
values
that
we
need
to
keep
track
of
in
a
drip
ecosystem.
B
So,
for
example,
authorized
personnel
only
could
look
up
from
a
ua
session
id
to
a
serial
number,
which
a
serial
number
in
this
case
and
in
any
subsequent
slide
is
an
ansi
cta
2063-a,
and
it's
either
the
serial
number
of
the
aircraft
itself
or
of
the
ua's
remote
id
module,
something
that
bob
touched
on
in
his
presentation
that
these
modules
were
introduced
and
are
required
to
use
these
serial
numbers.
They
cannot
use
a
session
id
and
only
a
caa
can
look
up
from
a
serial
number
to
a
registration
number.
B
A
ua's
registration
number
that's
in
their
database
of
registered
aircraft
in
their
flight
area
and
from
that
get
an
operator
registration
number,
which
is
considered
pii,
we're
not
quite
sure
on
the
link
between
ua
session
id
and
the
uss
account
id
that's
kind
of
a
gray
area.
It
might
depend
on
the
uss,
but
we
imagine
it
is
something
more
akin
to
like
an
authorized
personnel,
an
admin
of
the
uss
or
a
law
enforcement
agent
who
wants
wants.
It
can
look
it
up
next
slide.
Please.
B
B
Note
that
it's
not
that
the
uss
has
to
go
to
a
manufacturer
maintain
registry
to
get
the
serial
number
it's
more,
the
authoritative
location
where
the
serial
number
would
originate
from
and
its
main
unit.
B
B
So
this
is
an
example
of
the
of
possible
registries
and
their
interlinks,
so
you'll
notice
that
in
all
three
of
these,
the
serial
number
is
in
each
one,
but
certain
pii
is
in
the
caa
maintained
registry
and
only
the
set
and
the
session
id
can
be
looked
up
to
the
serial
number
through
an
rdap
lookup.
The
dns
lookups
to
like
get
the
certificates
and
the
public
keys
for
the
hierarchical
hit
would
be
a
dns
entry,
but
all
the
other
information
would
not
probably
be
dns
entries
because
they
could
be
considered
public
information.
B
B
All
of
these
things
could
be
publicly
accessible
so
that
other
registries
can
go
and
look
this
information
up
based
on
the
serial
number
and
before
michael
comes
in
stu,
is
on
the
queue,
and
I
imagine
he's
going
to
elaborate
or
correct
me
because
I
might
have
said
something
wrong.
G
Yeah,
I
just
wanted
to
give
danielle
a
bit
more
of
an
answer
on
dns.
G
We
don't
anticipate
that
dns
will
be
extensively
used
to
do
lookups
of
the
actual
information
that
is
associated
in
these
different
registries
with
these
different
identifiers
that
are
used
as
unique
keys
in
those
registries.
G
We
do
see
a
role
for
dns,
however,
in
getting
you
from
one
registry
to
another,
so,
for
instance,
given
a
hierarchical
hit,
which
is
the
registry
in
which
you
should
be
able
to
look
up
other
information
that
could
come
from
dns.
B
Yeah
schwein.
E
Adam
just
one
of
the
things
I
wanted
to
bring
up
in
general
aviation
right
now,
if
you
actually
file
a
flight
plan,
faa
does
give
you
a
general
warning
where
the
ues
activities
is,
but
there's
no
way
they
are
showing
either
the
june
number
or
whatever
the
number
it
is,
but
it
generally
basically
details
tail
number.
E
E
Somebody
just
give
the
the
the
thing
about
satisfaction.
You
know.
Okay,
if
you
type
a
tail
number,
you
will
see
where
this
aircraft
is
flying
right
now.
Currently,
what's
the.
D
E
B
So
I'll
fire
a
shot,
I
see
stu
and
rob
in
the
queue.
I
think
stu
has
the
answer.
The
direct
answer
for
you,
but
I'll
say
that
the
tail
number
that
you're
talking
about
is
what
I
think
I've
marked
as
the
ua
registration
number.
I
think
the
faa
is
moving
away
from
that
and
going
to
be
using
session
ids
and
not
so
much
the
ua
registration
number
to
be
passed
around.
I'm
not
a
hundred
percent
sure.
E
For
the
room,
don't
even
have
a
networking
connection,
they
don't
have
broadcast
id.
That
means
the
fa.
There's
no
way
to
tell
to
know
like
the
details,
information
for
that
group
right,
it
has
to
be
like
either
I
mean
for
everyone
to
know
it
has
to
be.
As
I
said,
if
it's
a
session
id,
it
has
to
be
networking
connections
right,
but
broadcasts
that
they
don't
even
have
one.
E
G
G
C
And
I
will
add
to
that
that
there
is
a
conflict
that
we
are
seeing,
that
stu
is
pointed
out
between
what
the
faa
says
they
want
to
do
versus
what
another
section
of
faa
is
doing,
but
the
there
is
going
to
be
a
separation.
C
We
don't
know
what
what
tail
numbers
are
going
to
be
moving
forward.
Will
the
tail
numbers
for
uas
be
the
same
as
tail
numbers
currently
and
for
manned
aircraft,
or
they
can
be
something
else?
This
is
a
gray
area
and
we
have
an
ability
here
again
to
be
inserting
some
rational
thought
processes
into
it.
B
Right
yeah,
so
by
no
means
trying
anybody
that
the
diagrams
I've,
just
the
three
diagrams
that
I've
just
presented
over
are
set
in
stone.
They
are
not.
They
are
basically
throw
it
at
the
wall
for
what
we
know
right
now,
try
to
make
it
all
fit
and
see
where
we
landed
and
then
try
to
feed
back
to
michael
and
work
with
michael
to
get
something
that
is
actually
usable
and
in
a
registry
up
somewhere.
So
that
was
this
is
all
nebulous
at
the
end
of
the
day.
B
J
Yes,
thank
you
adam,
so
we
had
heard
earlier
about
the
faa
punting
on
cyber
security,
as
as
I
referenced,
the
eu
has
actually
dove
into
the
deep
end
of
the
pool
rather
aggressively
as
I
referenced.
This
is
this.
Is
the
eu
cyber
security
directive
it's
nis
ii?
A
draft
of
this
was
released
late
in
december.
J
There
is
talk
about
trying
to
finalize
this
sometime
this
year,
although
there
needs
to
be
through
negotiations
between
the
parliament
and
other
aspects.
This
is
probably
something
that's
going
to
be,
probably
18
months
to
24
months
before
it
finalized.
There
is
still
some
drafting
going
on,
but
there
are
two.
There
are
two
things
that
I
I
want
to
key
on
regarding
this
cyber
security
directive,
annex
one
and
annex
annex
one
identifies
what
they
call
essential
services
and
there
are
two
essential
services
that
are
specifically
enumerated.
J
That
really
requires
our
attention.
The
first
is
number
two
a
and
it
talks
about
air
transport
now
right
now.
The
way
this
is
written
is
this
is
primarily
about
manned
aircraft,
but
they
do
talk
about
air
traffic
management
and,
given
that
the
europeans
are
a
little
behind
the
u.s
and
where
they
are
with
their
drone
regulations.
J
It
would
not
surprise
me
if
this,
this
definition
or
annex
was
potentially
expanded,
to
include
drones,
independent
of
that
specific
enumerated
provision
in
annex
one.
There
is
another
paragraph,
it's
number
eight
and
it
talks
about
digital
infrastructure,
and
here
digital
infrastructure
talks
about
pld
registries,
which,
if,
in
fact,
we
were
to
use
a
tld
for
this,
would
potentially
fall
within
this
definition.
J
There
are
also
other
provisions.
There
are
other
enumerated
services
under
the
digital
infrastructure,
a
bucket
in
annex
one
that
talks
about
trust
services,
cloud
providers.
So
I
think
when
you
look
at
where
we
are
going
with
the
broader
drone
program,
the
utm.
I
think
that
the
ability
for
this
to
fall
under
either
2a
or
under
8,
I
would
say,
is-
is
prob
probable
possible
highly
likely
again
still
in
flux,
but
this
is
something
you
definitely
want
to
keep
your
eye
on
next
slide.
J
Please
there
are
so
what
I
want
to
do
real
quickly
is.
I
want
to
touch
on
a
number
of
articles
contained
in
the
current
nis2
directive
to
call
your
attention
on,
because
I
do
believe
that
they
directly,
inter
potentially
directly
intercept
with
the
work
that
we're
doing
so.
J
Article
23
was
originally
inserted
by
by
dg
connect
as
a
result
of
some
of
the
pushback
that
has
resulted
as
a
result
of
the
gdpr
and
the
blocking
of
underlying
who
is
information
and
associate
and
associated
in
association
with
certain
domain
names.
So
if
you
go
here,
they
talk
about
imposing
upon
the
tld
registry
operator,
the
ability
to
collect
and
maintain
accurate
information,
the
ability
to
make
this
available
upon
a
lawful
and
dually
justified
request
from
legitimate
access
providers.
J
So
I
think,
as
we
look
at
how
privacy
intersects
with
potential
access
to
pii
associated
with
certain
drone
operators,
article
23
is
something
that
you
definitely
want
to
potentially
keep
an
eye
on.
As
as
this
moves
towards
final
implementation,
next
slide,
please.
J
Now,
article
18,
this
is
where
cyber
security,
cyber
security
risk
management
measures
are
specifically
enumerated
and
paragraph
one
basically
imposes
an
obligation
on
all
of
those
essential
services.
So
remember,
I
was
telling
you
about
annex
one
number,
two
and
number
eight.
If,
in
fact,
you
fall
within
this
requirement,
you
have
to
take
appropriate
and
proportional
technical
measures
to
manage
your
cybersecurity
risks
regarding
the
security
of
your
networks
and
the
informations
therein.
J
So
again,
I
think
this
is
is
potentially
relevant
in
what
we're
doing
paragraph
two
talks
about
a
number
of
non-exhaustive
measure,
measures
that
need
to
be
included
and
what
caught
my
attention
specifically
was
was
2d
of
article
18,
where
they
talk
about
the
supply
chain,
and
I
know
some
of
the
discussions
that
we
have
talked
about
within
within
this
group.
Is
this
this
whole
idea
of
a
supply
chain
being
able
to
track
things?
J
J
So
this
is
so.
Why
is
this
important,
why
it's
important
is
administrative
fines,
so,
as
everyone
knows
through
the
gdpr,
you
know
part
of
the
paranoia
within
businesses
and
and
why
that
garnered,
so
much
attention
was
under
the
gdpr,
there
was
the
potential
of
a
20
million
euro,
fine
or
four
percent
of
global
turnover.
J
What
is
interesting
is
under
article
31
if
the
paragraph
four
talks
about
a
violation
of
article
18,
which
was
that
cyber
security,
a
component
that
I
was
just
referencing
if
there
is
a
violation
of
article
18
by
one
of
those
essential
service
providers,
they
are
potentially
looking
at
an
administrative,
fine
of
10
million
euros
or
up
to
two
percent
of
total
worldwide
annual
or
turnover,
whichever
is
higher
so
again,
if
in
fact
we
are
going
to
be,
you
know
looking
at
this,
I
think
that
this
is
something
that
we
need
to
be
paying
attention
next
slide.
J
And
for
those
that
are
like
I'm
not
in
europe,
I'm
only
I
am
a
non-european
entity
or
I'm
only
selling,
or
you
know,
I'm
just
going
to
focus
in
the
u.s
or
in
asia,
pacific
south
america,
much
like
the
gdpr
had
an
extra
territorial
provision.
The
nis
directive
has
a
similar
extra
territorial
provision.
J
So,
if
you
are,
if
the
goods
and
services
that
you
are
providing
are
targeted
to
the
european
union
or
businesses
or
individuals,
they
will
in
fact
exert
jurisdiction
over
you,
your
company,
under
this
provision.
So
I
think
this
again
is
is
interesting.
It's
something
to
closely
follow
as
over
the
course
of
the
the
next
say,
12
to
18
months.
This
is
finalized,
perhaps
in
parallel
with
some
of
the
the
the
drone
regulation.
So
I
think
that
is
it
for
my
presentation.
B
The
registry
stuff
it
has
been
in
flux
for
a
long
time,
it's
finally
coming
back
to
the
forefront
and
we're
actively
trying
to
pursue
it.
There's
a
lot
of
legal
stuff
going
on
in
europe
that
we're
now
aware
of,
and
we're
fine
we're
trying
to
get
rubber
onto
the
road
and
get
it
really
going
now
now
that
some
other
things
are
starting
to
clear
up
so
help
drafting
appreciated
conversation
on
the
list,
because
we
need
it.
A
J
Independent,
no,
there
are
actually
provisions
under
the
I.
This
is
the
iatf
I
I
was
trying
to
wear
more
of
my
engineering
hat
instead
of
the
lawyer
hat.
But
yes,
if
you
want,
there
are
implications
with
the
dsa
as
well
as
the
dma
under
the
dsa.
There
are
provisions
that
talk
about
safe
harbors,
where
the
tld
or
dns
operators
can
potentially
qualify
for
immune.
If
you
will
a
liability
exemption,
so
if
you
want,
I
could
perhaps
provide
that
information
to
the
list.
J
But,
yes,
I
think
where
you
look
at
the
european
union
right
now,
much
like
they
did
with
the
gdpr
in
providing
thought
leadership
in
the
area
of
data
privacy
and
establishing
global
standards.
I
think
you
see
the
same
thing
here
which,
with
the
dsa
the
dma,
as
well
as
the
nis
nist
ii,
cyber
security
directive,
so
I
do
look
at
them
all
as
complementary
and
interconnected.
A
I'm
not
seeing
any
comments,
so
I
think
we
have
been
through
all
the
presentations
we've
addressed
all
the
questions.
We
have
no
more
questions.
Maybe
one
question
we
may
go
back
is
how
we
move
on
the
work
further.
A
Do
we
think
that
interim
meetings
were
useful
and
do
we
want
to
have
those
for
the
next
months
coming
months,
a
basis
of
one
per
month
might
be
a
nice
try.
So
that's
a
question.
I'm
raising
for
the
five
minutes
that
are.
B
So
interims
one
a
month
would
not
be
a
terrible
plan.
I
think
when
we
had
the
interims
going,
we
were
making
good
progress
on
a
lot
of
the
documents
it
kept
us
all
on
our
toes
and
kept
us
pushing
forward,
and
I
think,
with
the
registry
draft
to
get
it
up.
Just
you
know
up
to
speed
where
everybody
else
is.
I
think
that
would
be
useful
and
you
know,
instead
of
four
of
us
sitting
in
a
room
yelling
at
each
other,
maybe
having
20
of
us
in
a
room
virtually
yelling
at
each
other.
B
We
might
get
a
little
bit
more.
It
might
be
a
little
bit
more
annoying,
but
I
think
it
would
be
good,
then
keeping
it
enclosed
within
because
right
now
it's
been
me
michael
stew
and
bob
back
and
forth
to
get
where
we
are
now.
I
think
opening
it
up
and
getting
the
arc.
We
think
we
have
a
good
architecture,
we
start
iterating
over
expanding
it
contracting
it,
etc.
A
Okay,
anyone!
So
so
you
you,
if
I
try
to
sum
up
you
thinking
that
having
internet
meetings
is
going
to
be
useful.
B
F
Not
specific
on
this,
on
the
on
this
point
about
the
the
interims
but
yeah
we
just
a
comment
is
that
we
recently
we
we
stopped
to
organize
the
interims,
because
we
wanted
to,
I
would
say,
to
put
more
focus
on
the
on
the
writing
of
the
documents,
basically,
the
requirement
and
the
architecture.
F
So
now,
if
we
are
confident
that
we
are
making
progress
on
this
one,
so
yeah
we
can,
we
can,
we
can
can
go
back
and
yeah
and
organize
these
meetings.
Yeah.
F
Thank
yeah
just
before
we
adjourn
daniel,
just
a
reminder
that
we
have
the
the
working
of
lascal,
which
is
still
running
for
the
the
architecture
document.
This
is
ryan,
tell
the
23.,
so
please
read
the
draft
and
we
need
more
eyes
and
more
reviewers
on
the
document,
so
to
make
sure
that
the
document
you
will
send
to
eric
will
be
reflecting
what
we
want
to
have.
So
please
share
your
comments
on
the
list.
A
Thank
you
yeah
sure,
okay,
so
I
think
we
can
return
the
meeting
now
and
please
remain
active
on
the
mailing
list.
Thank
you,
everyone
for
attending
that
meeting
and
for
moving
that
work
forward.
Thank
you.