►
From YouTube: IETF-CBOR-20220209-1500
Description
CBOR meeting session at IETF
2022/02/09 1500
https://datatracker.ietf.org/meeting//proceedings/
A
The
charging
stations
that
I'm
using
here
in
the
office
park
are
level
two
chargers
and
it
takes
four
or
five
hours
to
put
a
full
charge
on.
But
you
know,
if
I
spend
an
hour
every
couple
of
days
at
it
on
a
conference,
call
it's
sufficient.
A
Very,
very,
very
yes,
I
know
christian
told
me
yeah
as
I
was
explaining
I'm
my
microphone
is
on
the
laptop,
but
my,
but
the
audio
is
coming
through
the
car
through
the
car's
bluetooth
system.
So
the
echo
cancellation
is
probably
not
working
right,
I'll
just
make
sure
I'm
muted,
when
I'm
not
talking
during
the
meeting.
A
B
C
C
A
Well,
I
mean
actually
it
used
to
work
better.
When
I
was
working
from
the
car
I
said
I
was
just
sitting
in
the
front
seat
and
had
the
microphone
and
the
audio
both
going
through
the
car,
and
that
seemed
to
be
fine,
but
I
discovered
that,
of
course,
the
steering
wheel
gets
in
the
way.
A
A
B
Well,
that
depends
a
lot
I
I
booked
and
booked
my
flight
yet,
and
I
have
asked
people
who
are
on
the
attendance
list
whether
they
are
coming
and
some
of
them
are
still
waiting
for
the
travel
exception
they
need
from
their
cto
or
somebody
high
up
in
the
in
the
company.
So
I
probably
will
make
a
last
minute
decision
whether
I'm
coming.
A
Or
not,
okay,
yeah,
I'm
I'm
coming
the
my
company
just
requires
that
we
show
them
proof
of
vaccination
to
get
travel
approval
and
I
did
and
they
did
so,
I'm
good.
D
And
generally,
vienna
is
right
in
the
middle
of
the
omicron
wave,
so
chances
are
by
that
by
the
time
we
have
the
itf
here,
we'll
be
past
that
right,
okay,.
D
After
thanks
marco
for
volunteering
to
take
a
few
notes,
have
two
major
points
on
the
agenda
for
today,
one
is
the
status
of
file
magic
and
the
other
is
religion
of
the
the
new
illusion
proposal.
D
I'm
hearing
nothing,
there
was
the
question
of
the
status
of
filematic,
so
yesterday
a
08
was
submitted
and
I've
already
looked
through
that
it's
mainly
document
cleanup
in
terms
of
what
is
normative.
What
is
informative
and
the
tutorial
changes
cast
and
brief
question
for
you.
In
the
last
interim,
you
mentioned
that
there
were
a
few
easily
contested
historical
statements
and
I
didn't
see
any
changed
was
that
just
in
the
editor's
copy,
or
was
that
already
in
the
document.
B
B
D
So
for
me,
so
I'd
like
to
ask
one
question
about
the
document
to
the
group,
and
that
is:
are
you
happy
with
sending
this
as
it
is
off
to
the
iesg,
and
I
know
that
many
of
you
may
not
have
read
the
sh08
but
considering
the
changes?
I
think
that
if
you
read
o7
and
we're
happy
with
that,
you
will
be
happy
with
the
way
too.
D
D
I
suggest
we
start
a
shortened
working
group
last
call
just
one
week
if
I
so
everyone
please,
if
you
have
any
comments,
add
them
or
look
over
the
document
once
more
and
if
there,
if
nothing
comes
up
in
that
period,
I
will
click
the
button
and
in
that
time
also
write
the
shepherd
right
up
with
which
this
can
progress
to
the
isg.
D
D
Let
me
see
whether
this
deck's
ready
to
be
shared.
Yes,
please
try
again.
A
A
B
Yeah,
so
we
have
been
looking
at
diagnostic
notation
in
in
various
environments,
so,
for
instance,
if
you
follow
the
earth
48
for
1952
1953,
the
the
cosy
documents,
there's
tons
of
diagnostic
notation
in
there
for
the
examples-
and
that
will
only
increase
and
michael
noticed
that
often
he
actually
doesn't
want
to
put
in
the
complete
possible
example.
B
But
one
wants
to
only
show
the
important
parts,
and
in
that
case
it
would
be
necessary
but
very
useful
to
have
a
illusion
mechanism
and
michael
mentioned
byte
strings.
I
added
text
strings
to
this
list
and
also
arrays
and
and
maps.
These
often
also
have
entries
that
you
don't
really
need
in
in
the
example.
B
So
the
question
is:
can
we
add
an
illusion
mechanism
to
this
to
the
diagnostic
notation
and
yeah?
I
just
cuffed
up
the
syntax.
That's
maybe
not
the
the
final
syntax,
but
it's
that
to
me
it's
pretty
suggestive.
B
So
we
we
just
use
three
dots
for
a
data
item
that
is
not
supplied
so
syntactically.
The
three
dots
can
occur
everywhere.
A
data
item
can
occur
and
in
a
few
more
spaces
that
will
be
next
slide
and
the
the
semantics
really
is.
This
is
not
just
a
single
data
item,
but
it
actually
can
be
multiple
data
items,
but
we
are
not
trying
to
to
express
this
zero
or
more
semantics
visually.
B
B
And
then
michael
also
had
the
observation
that
we
sometimes
want
to
provide
some
context
in
the
text.
Strings
or
byte
strings
that
that
we
are
using
so
being
able
to
put
in
a
illusion
in
the
middle
of
a
string
would
be
useful.
B
That's
easy
for
most
byte
strings
except
the
ones
that
look
like
text
strings,
because
the
dot
dot
is
otherwise
not
syntactically
allowed
in
in
a
byte
string.
But
in
a
text
string
you
actually
would
have
to
make
use
of
the
maybe
slightly
obscure
syntax.
We
have
for
concatenated
strings
in
our
seed
8610
appendix
g4,
so
we
we
can
always
just
end
a
string
and
start
a
new
one,
and
here
we
we
just
intersperse
an
illusion
to
say
some
some
bytes,
some
some
characters
excuse
me
could
come
between
them.
B
So
this
is
the
syntactical
proposal.
Obviously,
to
to
make
this
happen,
I
would
need
to
actually
implement
appendix
g4
the
currency
where
diagnostic
package
doesn't
implement
that,
and
there
is
an
issue-
has
been
an
issue
on
the
repo
for
more
than
a
year.
Sorry
about
that,
so
that
that
will
need
to
be
done.
But
with
respect
to
what's
already
standardized,
this
is
a
pretty
minimal
edition.
Michael.
C
Hello,
hello,
yes,
okay,
all
right,
just
speaking
louder,
all
right,
so
I
think
that
a
big
question
we
need
to
answer
as
a
group
is
whether
or
not
some
of
these
things
are:
what
happens
when
you
parse
them?
Do
you
get
an
error
or
a
warning?
What
happens
you
get?
Oh,
yes,
this
is
syntactically
correct,
but
I'm
not
going
to
convert
it
to
anything
because
it's
incomplete
or
what
and
I
think
the
other
question
is.
C
C
I
think
that's
not
controversial,
but
I
think
the
question
is
what
happens
when
you
parse?
Something
that's
incomplete
needs
to
be.
B
Answered
yeah,
I
have
a
couple
more
slides
about
pretty
much
that
question
so
when
we
discussed
the
the
previous
extension
to
the
diagnostic
mutation,
this
was
really
just
meant
as
something
that
that
humans
write
and
then
that
is
processed
by
the
diagnostic
notation
parser
into
your
bug,
standard,
sibo
data
item
and
this
one
is
actually
a
little
bit
different,
because
the
examples
we
use
in
rfcs
are
often-
and
I
hope,
more
and
more
machine
generated.
B
So
we
want
to
be
able
to
to
generate
a
seabor
data
item
that
is
then
expressed
as
if
it
had
illusions
and
what
what
I'm
proposing
is
just
to
sing
out
a
tag.
I
use
tag
number
888
because
it
almost
looks
like
an
illusion
and
ellipses
I
mean,
and
we
use
that
tag
to
actually
indicate
illusion
in
cases
when
the
diagnostic
notation
tool
is
configured
to
process
these
illusions.
B
So
when
you
write
your
example
generator
you
could
have
places
where
you
use
this
tag
or
you
could
write
a
little
tool
that
looks
at
all
strings
and
if
they
are
too
too
long,
they
are
turned
into
this
tag.
Number
88,
so
that
there
are
many
ways
to
handle
this,
but
we
don't
need
to
standardize
those.
The
the
important
part
is
that
we
have
a
standard
route
for
converting
between
diagnostic
notation
and
something
that
can
be
processed
as
a
sibo
data
item.
B
B
So
the
what
you
see
on
the
left
side,
what
you
you
might
see
in
an
example
in
an
rfc,
the
the
zebra
diagnostic
parser
in
in
illusion
processing
mode,
would
create
this
sibo
data
item
from
it
and
that
could
be
then
processed
by
by
an
example
verifier
or
whatever
you
have
around
for
for
doing
this,
and
that,
of
course,
is
also
the
inverse
direction.
B
So
if
your
example
generator
wants
to
generate
illusions,
but
actually
is
designed
to
generate
a
sibo
data
item,
then
it
could
include
these
tag
8
in
there
and
if
you,
if
it
then
runs
the
sieber
diagnostic
processor
in
illusion
processing
mode.
You
get
what's
on
the
left
side
of
the
slide.
B
Yeah,
so
one
problem
here
is
that
the
zebra
diagnostic
code
already
has
half
a
dozen
flags
you
can
set.
B
B
So
that's
the
the
proposal.
I
haven't
written
up
the
text
because
I
wanted
to
hear
feedback
first
and
my
my
plan
would
be
to
actually
implement
this.
I
think
it's
always
important
to
have
things
implemented,
so
you,
you
know
whether
you
over
something
and
then
document
it
either
as
an
addition
to
the
the
other
extended
diagnostic,
notation
draft
or
maybe
as
a
separate
specification.
B
B
Check
examples
whether
the
the
illusions
are
acceptable
from
the
point
of
view
of
some
special
city
and
you
cook
up
for
that,
or
something
like
that
right
now
I
haven't
come
up
with
a
particularly.
C
C
I
don't
know
exactly
how
I
would
use
this
888
tag
in
practice
with
example
generation,
but
I'm
happy
to
have
it
one
way
or
the
other,
and
I
guess
we
need
an
argument
somewhere.
Maybe
that
says,
if
you
see
888
then
blow
up
because
it's
this
is
not
correct,
and
so
I
kind
of
like
that
that
that's
okay,
you
can
process
it.
It
turns
into
this
thing:
there's
a
tag
in
it.
People
could
see
that
it's
broken.
C
I'm
am
fearful
that
it'll
wind
up
as
code
somewhere,
but
you
know,
there's
lots
of
other
stupid
things.
People
could
do
so
not
too
worried
about
that.
E
Yeah
wait
wait
my
seconds
for
the
audio
to
kick
in
hey,
so
I
think,
carson.
Your
point
about
you
know
using
this
to
actually
generate
to
generate
these
from
machines,
as
opposed
to
by
hand
is
actually
valuable,
and
that
this
helps,
I
I
think,
I'm
all
for
it,
because
at
the
end
of
the
day,
right
you
with
with
better
machine
generated
stuff,
you
you
you're,
actually
showing
what's
really
going
to
get
used.
It
saves
people
checking
it
later
as
well.
B
Yeah
christian's,
the
point
is
there
will
be
lots
of
ways
to
abuse
this,
and
I
think
that
that's
why
it's
important
to
to
have
a
special
flag.
You
give
to
the
diagnostic
notation
tools
to
indicate
that
you
actually
want
to
use
this.
So
this
is
not
something
that
the
data
can
tell
you
that
this
is
in
use,
but
the
invocation
of
of
the
tool
needs
to
indicate
that
this
is
actually
something
that
is
supposed
to
happen.
D
Okay,
as
you
as
you
mentioned,
the
interaction
with
cddl.
I
wonder
a
bit
whether
the
whether
this
the
the
interior
of
the
888
kind
of
it
looks,
looks
a
bit
like
cat
to
me
and
I'm
not
sure
whether
that
can
be
used
in
any
way.
Here.
D
It
looks
a
little
bit
like
what
like
the
the
cat
like
the
cat
operator.
D
B
But
the
cat
operator
is
in
cddl,
and
here
we
are
talking
about
diagnostic
notation,
which,
by
the
way,
is
one
of
the
greatest
sins
that
we
committed
that
we
let
these
diverge
so
much
so
yeah.
You
would
do
similar
things
in
cda
with
a
cat
but
yeah
there.
You
would
say
what
what
is
actually
allowed
and
not
allowed-
and
this
is
just
an
example.
D
Yeah,
but
but
some
some
somewhere,
basically
some
of
the
rules
for
verifying
that
something
is
that
an
888
tag
matches
that
an
888
tag
with
a
three
lung
array.
A
middle
b
matches
matches
something
that
that
that
matches
something
could
be
checked
by
verifying
whether
it
matches
a
cat
t
straight
cat
b.