►
From YouTube: IETF98-MILE-20170327-1520
Description
MILE meeting session at IETF98
2017/03/27 1520
A
B
B
B
E
D
D
D
A
D
D
We're
putting
all
our
faith
in
you
Roman,
okay,
so
with
that,
let's
go
ahead
and
get
started
so
for
those
this
is
the
mile
working
group
and
I'm
not
going
to
go
through
next
slide.
Please
just
keep
in
mind
the
note.
Well
I'm
not
going
to
go
through
it.
I
think
most
of
you
are
I've
been
here
before
so
next
slide,
please!
Okay!
D
So
on
the
agenda
today,
we'll
do
a
quick
status
update
and
then
de
ver
Stephen
you'll
present
the
updates
to
the
Rowley,
and
then
I
will
present
the
guidance
draft
okay
and
then
we'll
close
with
the
last
order,
business.
Okay.
So
with
that
we
have
finished
the
io
def
on
drats.
We
are
currently
have
two
working
drafts
in
order,
and
hopefully
we
can
do
another
last
call
jumping
ahead
for
the
rolly
one
speaking
quickly
on
the
XMPP,
not
as
the
chair,
I
still
other
than
Chris's
feedback
haven't
gotten
anything.
D
So
when
Kathleen
comes
back,
we'll
ask
because
it's
been
on
last
call.
So
those
are
the
two
remaining
transport
drafts.
There's
one
more
draft
I
think
on
the
guideline
right,
because
the
other
one
has
RF
seen
the
application
implementation.
So
we
basically
have
three
things,
hopefully
in
the
home
stretch
for
a
mile,
maybe
yeah,
and
so
with
that
let's
go
ahead
and
do
the
update.
So
should
we?
Oh?
Let's
should
we
do
the
milestones
real,
quick,
yeah,
I
think
I
kind
of
covered
this
real
quick.
D
So
we
had
targeted
to
do
the
working
last
call
for
the
guidance
and
the
rolly
and
I
think
we
did
for
both,
but
we
were
expecting
more
feedback
and
an
update
for
the
rolly,
as
well
as
the
guidance
which
we
now
are
going
to
get
updates
for.
I
have
not
gotten
any
feedback
for
the
XMPP,
so
I
don't
know
if
that's
good
or
bad
and
then
we're
targeting
to
do
another
working
group
last
call
for
the
sea.
D
D
F
In
trouble
because
there's
no
pink
box,
the
camera
can
see
me,
though
I
don't
think,
there's
there's
nobody
on
there
anyway.
So
the
recording
okay.
So
for
those
who
do
not
know
me,
I'm
Stephen,
Banghart
from
NIST
and
I'm
going
to
present
some
updates
on
Rolly.
So
next
slide
please.
So
this
is
a
brief
overview
of
the
changes
that
we
made
since
the
last
version
that
we
presented
at
the
last
hi
ETF
meeting,
there
have
been
no
new
major
changes.
F
The
changes
that
we
made
were
the
ones
that
were
decided
on
at
Mile
in
Korea,
so
we
implemented
those
changes
that
were
decided
on.
We
made
a
few
minor
fixes
throughout
the
document
and
those
changes
all
together
constitute
the
06
version.
So
we're
going
to
go
ahead
and
go
through
really
quick
talk
about
the
changes,
try
and
figure
out
what
all
changed.
So
we
can
pin
down
all
of
that.
So
the
first
major
change
that'll
be
better
if
I
just
go
through
slides
from
you
next
slide,
please.
F
So
the
first
change
was
adding
the
rolly
property
element.
This
was
issue
number
51
on
the
github
for
anyone
who
is
following
along
there
and
again,
this
was
decided
on
at
the
last
IETF.
We
added
a
new
element
to
the
rolly
schema
and
defined
it
in
the
document.
Called
Rollie
property
properties
are
registered
and
I
in
a
table
that
we
have
established
in
the
main
document,
and
extension
documents
may
register
new
properties
into
this
table.
F
So
every
time
you
want
to
add
new
properties
doesn't
mean
you
have
to
publish
a
new
document
that
has
those
properties
registered
in
the
eye
on
a
table.
These
properties
are
for
arbitrary
exposure
of
information.
That's
inside
of
the
content,
for
example,
if
you
have
an
io
def
document,
and
you
want
to
be
able
to
expose
the
ID
of
that
I,
oh
def
documents
that
it's
easier
to
search
so
that
somebody
doesn't
have
to
download
the
entire
I.
F
F
This
is
all
explained
more
in
the
document
as
well.
I
don't
have
any
more
questions
there
next
slide,
please.
The
second
change
that
we
made
that
was
agreed
on
in
Korea
was
the
standardization
of
the
URI
template,
URL
template
for
service
and
category
documents.
We
decided
that
there
is
this
template
here:
the
host
port,
rowley
service
document
and
rowley
category
document.
F
This
is
a
requirement
in
the
draft
such
the
dereferencing.
This
URL
results
in
the
retrieval
of
the
service
document
or
the
category
documents
respectively.
This
does
not
mean
that
the
server
has
to
store
the
document's
in
these
locations.
It
doesn't
mean
that
actually
has
to
be
the
the
file
path.
If
you
will,
it
means
the
dereferencing.
F
This
URL
results
in
that
document
being
sent
back
so
you're
free
to
redirect
that,
however,
you
will
as
long
as
it
does
return
the
service
or
category
document,
and
we
made
this
change
as
discussed
in
Korea,
because
it
allows
for
discovery
of
rowley
repositories
and
of
these
service
documents
in
a
standardized
way.
Next
slide,
please,
after
a
brief
poll
at
Korea
and
the
ABS
approval,
we
switched
rowley
to
standards
track.
There
were
three
groups
indicating
support
for
implementation,
so
that
was,
it
was
experimental.
It
is
now
standards
next
slide.
F
We
went
through
and
made
sure
that
the
references
were
categorized
by
normative
and
informative
categories,
as
well
as
making
sure
they
were
accurate,
and
we
also
ran
our
XML
examples
through
a
validation
and
went
through
them
by
hand
as
well
to
make
sure
that
all
the
examples
were
correct,
as
well
after
some
of
the
changes
that
we
made
next
slide,
please
I
should
mention
as
well
that
we
made
a
number
of
grammar
and
phrasing
changes
throughout
the
document.
No,
no
requirement
changes,
no
normative
changes.
F
It
was
just
clarity,
changes
throughout
the
document
to
make
sure
everything
was
as
clear
as
possible,
so
that
that
was
the
summary
of
changes
that
we
made
in
rolla
who's.
No
changes
knew
beyond
what
we
already
decided
in
Korea.
So
with
that
being
said,
all
of
the
major
work
as
far
as
we
are
as
far
as
we
believe
is
done
on
Rolly.
We
believe
that
this
document
is
content.
Complete
is
all
the
major
work
is
complete.
F
Rolly
was
in
working
group
last
call
a
long
time
ago.
It
has
gone
through
several
major
revisions.
Since
then,
we
believe
that
Rollie
is
ready
for
a
working
group
last
call
to
be
restarted
in
order
to
get
review
on
the
new
major
revisions
of
the
document,
so
for
the
myself
and
Dave
the
primary
co-authors.
At
the
moment,
we
believe
that
working
group
last
call's
is
the
next
step
for
Rowley.
F
Those
extensions
still
need
review
from
people
that
are
knowledgeable
in
those
particular
areas,
especially
with
the
inclusion
of
the
new
Rollie
property
element.
Subject
matter,
experts
can
provide
feedback
as
to
which
properties
are
particularly
valuable
to
register
for
exposure.
So
if
you
are
an
expert
in
any
kind
of
software,
descriptor
formats
such
as
swig
or
if
you're,
an
expert
in
io,
deaf
or
any
other
incident
indicator
formats,
then
your
feedback
would
be
greatly
appreciated
on
these
extension
documents.
F
We're
also
beginning
the
process
of
planning
out
additional
extension
documents
where
those
will
be,
I'm
not
sure,
but
we're
looking
at
writing
an
extension
draft
for
check
lists,
as
well
as
one
for
vulnerabilities
and
definitely
very
very
interested
in
that,
and
that
that
is
something
that
needs
to
get
done.
So,
if
you're
interested
in
that
you
can
contact
the
authors
we'll
be
happy
to
help.
You
write
those
extension
documents
in
order
to
plan
out
what
you
need
to
do:
good
Adam
on
Phyllis
ers.
F
Awesome,
that's
great
news,
so
it's
really
good
to
hear
if
there
are
other
things
that
you
are
interested
in,
creating
more
information
types
for
Rowley
that
are
relevant
to
your
use
cases.
You
can
either
write
the
extension
yourself
or,
if
you
contact
the
authors,
will
be
happy
to
give
you
a
couple
pointers.
You
can
also
reference
the
existing
extension
documents
in
order
to
get
a
template
for
some
some
examples
of
how
to
write
a
rolly
extension
document.
F
F
D
D
B
B
F
That
would
be.
That
would
be
a
good
addition
to
clarify
as
far
as
the
requirements
in
the
document
stand
now,
that
location
of
the
service
document
can
redirect
to
that
other
location.
The
the
only
requirement
is
that
that
particular
URL,
when
you
send
a
restful,
get
request
to
that
URL.
It
can
reach
correct
as
long
as
it
resolves
the
service
document.
I,
don't
know.
If
that
answered
your
question.
Dave
Dave.
H
So
so
I
think
we're
talking
about
two
different
paths
of
discovery
so
and
they're
both
talked
about
in
the
document,
the
the
first
that
hockey
was
talking
about
was,
if
a
home
page,
you
know
for
an
organization.
They
include
a
link
element
in
the
in
the
HTML
head
that
that
link
could
actually
point
to
where
the
rolly
services
and
that's
certainly
one
way
of
actually
identifying
a
rolly
service.
H
You
know
that
are
available,
given
you
know
how
you're
authenticated
and
authorized
on
that
server
and
that's
actually
extremely
valuable
and,
in
our
opinion,
to
be
able
to
do
that
without
having
to
interact
with
any
other
services.
Now
11
usage
example
could
be
using
a
service
record
to
point
to
the
rolly
service.
You
know
you
can
look
it
up
in
DNS,
go
straight
to
it,
start
to
enumerate
the
collections
and
you're
off
and
running.
So
that's
something
that
we
also
wanted
to
be
able
to
support
right.
F
B
D
So,
given
that
there
were
significant
changes,
I
think
it
would
be
worth
it
so
question
for
the
group
who
has
read
the
latest
draft
halfway:
okay,
so
so
maybe
what
I'll
do
not?
Maybe
what
we'll
do
is
I'll
send
out?
Another
email
to
the
group
to
get
reviews
can
I
get
at
least
a
couple
people
to
commit
to
reviewing
the
draft
Adam
and
Roman.
Okay.
D
He's
hiding
there
there
you
go
okay,
so
Roman,
if
you
wouldn't
mind
putting
it
in
a
minute,
so
I
I,
don't
forget
to
solicit,
feels
that
I
got
your
name.
Okay,
so
is
it
appropriate
to
just
put
a
deadline
so
that
I
can
yeah
so
yeah?
So
can
we
do
a
30
day?
Is
that
plenty
of
time?
Okay?
So
we'll
do
a
30
day
last
call
and
then,
depending
on
the
feedback,
I'll
either
do
the
Shepherd
and
move
it
forward
or
request
for
more
for
more
updates.
Okay,
all
right!
Thanks
all.
F
Still
version
00
we're
still
working
on
it
and
I
am
confident
that
we
will
have
a
0
1,
perhaps
a
no
to
depend
on
how
fast
we
kind
of
iterate
the
changes
it
regardless.
We'll
have
an
update
for
the
next
IETF
com.
Have
a
presentation
then,
to
talk
about
the
changes
the
document
is
is
fairly
fleshed
out.
There
definitely
needs
to
be
a
couple
passes
of
content
and
polish
we'd
like
to
see
done
on
it
before
it
gets
put.
But
at
that
point
then
we
could
talk
about
adoption.
H
D
H
F
And
then
there
are
the
one
of
the
big
things
I'd
like
to
see
is:
if
there's
anybody
here,
who
is
very
familiar
with
the
properties
and
usage
of
io
def,
just
an
email
dialogue
with
you
would
be
awesome
so
that
I
can
identify
some
of
the
properties
it
doesn't
have
to
be
long
can
be
one
email,
okay,
we'll
keep
piling
things
on
Roman.
That
sounds
like
a
plan
to
me:
okay,
I'm,
not
gonna,
email
Roman.
I
would
call
roman
and
some
that's
on
the
recording
now
so
you're
in
trouble.
Well,.
D
F
F
I
Sure
well,
I
am
mio
suzuki
yeah.
Now
I
will
present
our
cos
theta
so
kind
of
rough,
oh,
the
Tigers
blockades
to
provide
a
guideline:
Ohio
def
implementation,
for
example,
about
representation,
home
security
indicators
and
about
use
cases,
and
I
will
present
updates
from
the
previous
graph
and
then
I'll
show
to
do
this
yeah.
I
I
D
D
D
D
D
Okay
yeah.
Thank
you.
Thank
you
wow.
This
is
short
and
quick.
So
can
you
put
the
agenda?
Are
the
milestones
again?
Okay?
So
basically
we
are
ready
to
do
the
last
call
for
both
the
guidance
draft
and
the
rolly.
We
had
put
out
a
call
for
the
XMPP
draft,
but
we
haven't
gotten
any
other
than
Chris.
You
reviewed
it
and
found
it.
Okay,
I
haven't
changed
the
draft
on
so
Kathleen.
This
is
a
question
for
the
ad.
Given
that
we've
not
had
any
feedback.
How
do
we
proceed
on
the
XMPP
draft?
J
D
J
G
J
It
gets
wider
community
and
click
on
that
and
add
some
sad
I'd
like
to
skip
over
because
I
right
to
be
and
down
the
capillary
I'd
like
to
get
more
reviews
on
it.
I
thought
we
would
have
had
more
reviews
from
across
the
area,
so
I
guess
we
could
do
one
of
two
things:
I
could
do
an
extended
IETF
last
call,
or
we
could
do
another
working
group
last
call
where
we
reach
out
to
sag.
Do
we.