►
From YouTube: CBOR WG Interim Meeting, 2021-02-10
Description
CBOR WG Interim Meeting, 2021-02-10
A
This
is
an
official
itf
meeting
and
the
note
applied
yes,
so
I
we
have
an
agenda
for
today
is
you
can
see
it
on
the
code
dmd
link
that
christian
has
posted
in
the
chat
and
michael
in
jabber
chat
if
you're
there
so
for
today
we
wanted
to
first
have
a
short
discussion
or
yeah
a
document
status,
update
and
any
issues
that
might
come
up,
and
then
we
can
talk
about
the
individual
documents
and
michael
has
some
slides
for
those
and
then
any
other
business
anything
else
you
want
to
talk
about
or
discuss
today.
A
Okay,
if
not,
then
we
can
start
with
the
status
update
so
for
your
information
for
the
tags,
oid
draft
we're
still
waiting
for
sean's
reply
about
the
ipr
declaration
and
it's
been
two
weeks
and
one
day.
So
it's
it's
not
it's
not
been
months
or
anything,
but
we
had
really
zero
reply
whatsoever.
So
yeah
we
don't
know
like
how
long
should
we
wait
and
what
to
do?
If,
if
we
don't
get
any
answer
from
sean.
B
Managed
to
release
the
response
from
him
there,
so
I
did
this
again.
A
B
I
don't
know
more
about
that,
but
I
think
the
the
general
observation
is
that
most
of
the
things
that
he
originally
contributed
to
to
this
document
have
gone
in
into
a
different.
B
Yeah
and-
and
so
I
think
it
would
not
be
an
entirely
wrong
to
put
him
into
the
contributor
section
instead
of
the
offer
section.
Of
course,
the
next
question
is
whether
our
ad
will
be
happy
with
a
shepherd
report
that
mentions
this.
A
So
I
also
have
to
contact
him
via
another
channel
linkedin.
I
am
tempted
to
wait
a
little
bit
longer,
maybe
one
week
or
so,
but
yeah
the
shepherd
review
is
ready,
so
yeah.
C
The
the
thing
that
would
need
updating
it
very
least
is
the
is
the
address,
because
the
the
ui
that
is
given
for
his
company
address
doesn't
doesn't
resolve
anymore
or
doesn't
doesn't
just
give
any
response
anymore
or
precisely
so,
I'm
not
sure
whether
penang
is
still
in
business
or
whether
this
can
be
reasonably
determined
in
the
united
states.
But
at
least
it's
not
a
meaningful
context.
Contact
as
it
is.
C
A
A
We
have
a
co-author
in
the
document
that
is
under
shepard
review
that
is
not
answering
to
ipr
and
yeah
and
yeah.
D
D
So
if
the
issue
is
just
the
you
know,
getting
the
ipr
release,
whatever
moving
into
a
to
a
contributor
section,
doesn't
help
it.
Somebody
who
was
a
significant
contributor
does
need
to
confirm
part
of
it
depends
upon
how
well
you
know
him.
If,
if
you
re,
if
you
know
that
he
there's
really
no
ipr
issue,
then
you
can
move
ahead
with
it,
but
if
he
stays
just
engaged,
you'll
have
a
you'll
have
trouble
during
off
48.
right.
D
Okay,
well
so,
and
during
off
48
you,
as
the
responsible
ad,
will
be
able
to
approve
for
him,
but
we
usually
save
that
as
a
last
resort,
I
mean
I
thought
so
I
guess
from
an
off
48
point
of
view,
it's
good
to
move
him
to
the
contributors
section.
If
he's
really
gone
incommunicado.
A
It's
the
first
time
that
he
is
not
easily
reachable.
D
And
the
other
thing
for
for
the
shepherd
write-up,
you
can
always
answer
the
ipr
question.
With
that
we
we
believe
that
he
that
he
has
been
in
compliance
with
bcp
79,
but
we
do
not
have
confirmation
because
he's
got
a
wall
or
something
like
that.
You
can
just
put
that
in
the
shepherd
right
up,
but
yeah.
It's
it's
not
good.
When
you
have
a
co-author
who's
who
doesn't
respond
and
there's
really
little,
you
can
do
about
it
in
the
long
run,
yeah.
A
So
I
think
our
plan
is
to
wait,
maybe
one
more
week
or
so
because
now
we
try
to
contact
him
via
other
channels
and
then,
if
nothing's,
if
we
get
no
answers,
maybe
cast
and
move
him
to
the
contributor
section
and
we
will
post
a
shepard
review
or
christian.
Will
you
say
you're
going
to
be
the
shepherd
and
we'll
move
forward.
D
A
Okay,
then,
I
noticed
that
we
got
a
update
on
the
seaboard
pack
document.
B
So
this
is
partially
handcrafted,
but
I
think
it
should
show
how
the
the
prefix
compression
at
least
would
work
and
the
example
isn't
invented
that
came
from
an
from
a
hackathon
in
2017,
where
we
we
had
some
w3c
weber,
things
stuff
and
and
wanted
to
know
what
what
options
we
had
for
making
this
a
little
bit
more
compact
and
that
that's
actually,
when
I
wrote
the
the
current
implementation
of
of
similar
package.
B
B
I
wrote
a
while
ago
for
lp
when
so
I
think
that
that's
really
its
own
little
can
of
worms,
and
I
think
that
the
the
best
possible
outcome
would
be
that
we
understand
templates,
circumfixed
and
packed
in
in
such
a
way
that
this
can
be
an
optional
additional
ingredient.
B
A
Okay,
so
we're
waiting
for
for
people
to
read
this,
this
one
more
feedback
right.
B
Yeah
not
much
has
changed,
but
but
the
the
numbers
have
been
allocated.
I
stuck
to
the
one-quarter
space
for
for
suffixes
compared
to
prefixes,
and
there
is
an
example
that
that's
worth
looking
at.
C
On
on
the
topic
of
of
making
progress
around
the
circumfix
expansion,
what
what
I
think
would
help
is
if
we
had
a
bit
of
a
better
idea
on
whether
the
dictionary
needs
to
be
agreed
on
about
him
or
whether
that
there
needs
to
be
just
semantic
agreement.
That
is,
for
example,
if
the
if
both
parties
could
say
that
this
particular
variable
expands
to
the
document
request,
uri
whatever
that
is,
and
if
both
parties
are
in
disagreement,
the
whole
protocol
won't
fall
apart
or
whether
there
needs
to
be
explicit
agreement.
C
B
So
I
think
we
can
have
the
best
of
both
worlds.
So
people
who
want
to
do
something
like
security,
where
we
they
have
to
be
absolutely
sure
that
they
exactly
know
what
the
statement
that
they
are
assigning
means
they
can
go
with
a
strict
part
and
those
people
who
don't
actually
need
that,
like
what
thing
description
probably
doesn't
need
that
could
use
more
more
context
in
constructing
these.
B
A
A
I
think
that's
it
for
now
and
we
can
move
to
the
next
agenda
item
yeah.
B
Let
me
just
say
what
my
plan
is:
I
started:
fixing
the
the
sibo
packed
implementation
and
well,
the
the
the
term
is
ending
here,
and
people
find
out
that
they
need
to
have
working
instructions
how
to
handle
covered
after
the
end
of
the
job,
oh
at
yeah
moment.
B
So
all
the
time
I
had
allocated
today
for
working
this
when
went
into
that
bit
bucket,
but
I
I
hope
to
have
an
updated,
zero
pack
implementation
within
10
days
or
so,
and
then,
of
course,
when
people
can
play
with
this,
maybe
we
we
get
more
more
substantial,
more
deep
feedback
and
on
what
we
like
about
this
and
what
we
don't
like
about.
A
Okay,
so
yeah
we
have
one
more
meeting
before
it
f110,
so
hopefully
we
can
do
a
status
update
for
for
next
next
scene,
great
okay
and
that's
it
about
working
group
documents.
C
I'll
just
jump
in
quite
briefly
here,
because
I've
seen
a
few
names,
please
make
sure
to
add
yourself
to
the
blue
to
the
blue
sheet
in
the
posted
minutes.
Yes,.
E
Is
that
displaying
yep
yep
all
right,
so
this
is
a
short
draft
on
a
way
of
tagging,
seabor
protocols
on
primarily
on
disk
and
I'll,
explain
why
first
part
of
it
came
from
the
conversation
about
a
month
ago
about
we
had
different
things
about
seaboard,
encoding
of
certificates,
private
keys,
and
this
lot,
this
kind
of
thing,
and
particularly
private
keys,
don't
tend
to
get
transmitted
and
there's
a
bit
of
a
disaster
in
the
peak
space
in
terms
of
the
file
extensions
where
everything
is
a
dot
pen
file,
and
unless
someone
has
a
convention,
you
know
blah
blah
blah
priv,
they
don't
actually
know
if
it's
a
private,
key
or
not
and
sun
knows
pem
is
always
a
private
key
and
crt
is
the
certificate
and
there's
a
bunch
of
other
things.
E
That
kind
of,
I
think
it's
been
a
bit
of
a
management
disaster,
particularly
for
less
technical
people.
File,
however,
will
mostly
tell
you
it's
a
unix
command.
It's
been
around
since
the
1970s.
E
That
would
mostly
tell
you
what
what
is
what
and
so
that
saves
you,
because
there's
actually,
you
know
obvious
encodings
in
that,
and
people
steeped
in
the
no
actually
can
usually
recognize
what
it
is
by
the
base64
after
a
while,
because
the
first
couple
bytes
tend
to
be
very
consistent,
and
so,
even
if
you
actually
say
I
don't
really,
I'm
gonna
compile
all
the
stuff
into
code
when
it's
source
code
and
on
a
constrained
system
when
it's
source
code
and
you're
putting
it
together,
you
may
actually
find
that
you,
you
can't
remember,
which
p
piece
went
where
and
which
part
was
for
which
part
which
thing,
and
so
I
think
it
would
be
useful
to
be
able
to
tag
things
usefully.
E
So
today,
right
now,
if
you
point
file
at
a
c-bor
file,
you
are
told
by
the
file
command
that
it's
a
c-bor.
If
it
starts
with
dot
cbor
550.
E
Then?
It
definitely
gets
identified
as
cbor
and
that's
great,
but
it's
not
as
wonderful
as
it
is.
So
I
basically
just
was
converting
a
bespoke
25
year
old
interface
that
basically
threw
structures
across
an
ipc,
and
I
did
this
because
I
didn't
want
to.
I
don't
want
to
have
to
link
the
c
code
into
rusters
or
or
python
to
be
able
to
do
ipc
properly.
I
said:
well,
hey
I
want
it
in
cbor,
then
I
don't
have
to
do
anything.
E
It's
just
a
cddl
between
the
two
and
I
was
just
dissatisfied
in
my
unit
tests,
because
I
have
actual
records
like
this.
This
ipc,
I
have
them
as
records
all
over
the
unit
tests,
because
that's
how
I
do
my
unit
testing
is
from
is
from
input
like
that.
E
So
my
proposal
is
that
we
start
with
the
regular
seabor
tag
and
that
we
put
another
tag
inside
of
it,
which
I
recommend
is
comes
from.
The
first
come
first
served
policy
and
that
you
ask
for
32
bits
at
least,
and
that
puts
it
squarely
in
the
four
byte
space
and
it's
relatively
easy
to
get.
I
got
one
this
one
up
here
on
the
screen
is
opsn
in
ascii.
E
If
you
were
to
look
at
it,
which
is
for
open
swan,
which
is
where
it's
going
into
the
mat
file
command
suggests
very
strongly,
is
the
first
point
is
that
the
initial
match
is,
is
32
bits
long
and
at
least
30
bits
long
and
it's
a
unique
match.
E
So
the
c
bar
tag,
followed
by
32
bits
of
per
protocol
tag,
makes
it
very
simple,
and
then
we
had
a
discussion
on
the
mailing
list
as
to
what
exactly
you
would
tag,
because
I
had
a
previous
proposal
that
did
some
other
things
and
someone
very
brilliantly
pointed
out
that
the
tag
b-o-r
is
a
byte
string.
It
spells
c-bor
in
the
disk.
E
So
if
you
were
to
look
on
the
disk,
you
would
see
four
bytes
of
this
55-799
tag,
followed
by
your
four
bytes
of
your
protocol,
followed
by
the
word
seabor
in
in
the
hex
dump
or
the
the
hask,
the
ascii,
which
would
be
a
very
good
clue
if
you
were
actually
looking
through.
You
know,
random
files
that
are
stuffed
and
you
didn't
know
what
they
were,
and
you
see
this
you're
like.
Oh,
I
could
probably
decode
this
with
cbor.
E
So
now
I
do
this
and
I
run
file
and
I
have
a
extension
that
says
it,
and
it
tells
me
exactly
what
it
is
and
that's
really
about
it,
and
it's
really
a
bcp
and
it
doesn't
require
any
allocations,
because
anyone
who
needs
code
points
will
just
go.
First,
come
first
serve,
and
this
simply
says
I
you
know
go
do
this
and
do
it
this
way.
The
other
point
is
we
had
a
conversation
as
to
whether
or
not
this
would
be
a
prefix
tag
to
the
original
item.
E
The
actual
thing
you
want
to
do
it
or
a
sieber
sequence,
and
I
went
for
recommending
a
seabor
sequence.
It's
not
really
a
lot
harder
to
parse
off
the
disk,
and
it
very
clearly
says
this
is
not
the
object.
You're
looking
for
the
next
object
is
the
one
that
you
want
and
then
that's
the
one
that
you
have
and
it's
free
of
tags
of
this
kind
of
thing,
if
you
don't
want
them
and
because
why
send
an
extra
12
bytes
over
the
wire?
If
you
already
know
what
it
is,
that's
it
any
questions.
C
B
E
A
E
Not
a
bad
idea.
Okay,
I
I'm
not
opposed
to
that.
You
know
five,
five,
seven,
nine
eight,
I
don't
know
whatever
will
make
something
up,
but
I
I
don't
have
a
problem
with
that
adopt
the
document
and
let's
make
that
change.
B
Yeah
there's
a
little
bit
of
arithmetic
behind
this
55799
yeah
and
I
think
the
most
important
thing
is
that
it
clearly
makes
this
not
a
utef
utf-8
file
or
not
a
unicode
file
in
any
of
the
other
unicode
transformation
formats,
and
as
long
as
the
number
you
pick
also
has
this
property.
I
think
very
good.
E
B
So
I
I
still
would
for
those
people
who
actually
have
to
do
a
dot
zebo,
not
a
dot,
zebo
sec
kind
of
file.
B
I
I
still
would
like
to
have
the
convention
spelled
out
how
to
do
this
in
one
sibo
data
item,
so
there
you
would
actually
properly
use
55799,
and
I
think
you
would
still
go
for
this
32
bit
tag,
and
I
probably
need
some
wording
about
there
that
it
really
needs
to
be
32
bits
in
the
preferred
encoding,
and
then
I
would
probably
just
put
the
thing
in
there.
So
you
would
not
remove
12
bytes,
you
would
remove
eight
bytes
and
otherwise
you,
you
are
fine,
so
you're
right
this.
B
E
Okay,
so
I
I
wrote
the
code
in
my
code
and
actually
one
of
the
things
is
that
my
ipc
actually
started
with
a
magic
number
which
was
rev,
which
was
changed
almost
every
single
revision
of
the
software,
because
the
the
structure
would
change
right
because
it
was
a
c
structure
being
thrown
across
and
so
what
the
thing
that
I
needed
was.
E
I
wanted
to
say:
okay,
I
need
to
be
able
to
see
that
this
is
actually
the
newer
one,
the
seabor
one,
and
so
that's
why
I
don't
actually
don't
care
what
what
happens,
because
I'm
like
mem,
compare
for
the
first
eight
bytes
and
if
it's
good,
I'm
I'm,
then
I
process
it
a
seabor.
And
if
it's
not,
then
I
you
know,
I
have
a
little
bit
of
legacy.
Support
that
I
will
remove
two
versions
from
now,
because
the
big
problem
is
when
the
software
restarts.
E
And
vice
versa,
even
all
right,
so
do
you
want
me
to
make
this
change
before
adoption,
or
does
the
working
group
wish
to
adopt
it
and
then
instruct
me
to
make
the
change.
B
A
A
Yes,
I
would
like
to
hear
more
people
explicitly
explicitly
say
they.
They
think
this
should
be
done
in
cyborg,
if
you
think
so
in
the
in
the
call,
and
then
we
can
confirm
this
in
the
we
can
ask
the
question
in
the
mailing
list
as.
A
A
Okay-
and
I
guess
we
need
it-
would
be
good
to
have
some
people
look
at
it
and
and
to
state
what
they.
A
A
Okay,
and
also
you
were
saying,
michael
in
the
mailing
list,
that
you
think
that
this
needs
to
be
marked
as
pcp.
E
Yeah,
I
think
that
bcp
is
the
right
right
thing
for
this.
I
guess
we
can
still
allocate
a
tag
at
in
bc
with
a
bcp,
because
we
don't
have
to
allocate
this
new
seaboard
sequence
tag
in
the
oneplus
2
space.
I
guess
it
is.
E
So
I
think
that
needs
a
document
right.
If
I,
if
I
remember.
B
E
Yeah
and
but
we
can
yeah
so,
but
we
can
also
right
yeah
exactly
I
wouldn't
say,
but
but
it's
it's
fine
that
when
we
write
when
we
do
the
first
come
first
serve
that
we
point
to
the
document
as
an
internet
draft
to
do
that.
Okay,
so
we
just
need
to
come
up
with
the
right.
The
right
number,
the
right
magic
in
that
number
and
okay
yeah,
but
I
think
it's
a
bcp.
B
Yeah,
I
think
there
will
be
standards
at
some
point
that
will
reference
this
and
say
our
tag.
Is
this,
and
but
this
document
can
be
a
recipe.
E
Mm-Hmm
that
one
yes,
okay,
so
while
encoding
this
bespoke
protocol,
I
had
ip
addresses
to
send
and
v4s
and
v6s,
and
I
looked
at
oh,
there
must
be
a
tag
for
that
and
I
found
tag
260
and
261
and
I
found
that
tag.
260
says
basically
it's
a
v4
if
it's
four
bytes,
it's
v6,
if
it's
16
bytes
and
it's
ethernet.
If
it's
six
bytes
and
I
then
said
well
what
about
eight
byte
ethernet,
which
is
definitely
a
thing,
and
I
also
realized
well,
you
know
you
might
be
forced.
E
In
my
case,
I
don't
care.
If
I
don't
I'm
not
that
sensitive
for
size,
but
on
the
wire,
you
might
be
forced
to
send
16
bytes
which,
when
there
may
be
many
zeros,
eight
bytes
of
zeros,
if
it's
a
prefix,
typical
prefix
and
then
I
looked
at
okay,
what
about
prefixes
and
in
both
cases,
these
point
to
essentially
point
to
a
github
wiki
page
as
the
documentation,
a
guy
named
rava?
I
don't
know
him.
E
I
didn't
recognize
his
name
to
do
this
and
I
looked
through
this
and
I
went
okay,
but
he's
telling
me
to
apply
it
to
arrays
and
maps
and
I
went
well.
I
don't
really
want
to
do
that
and
and
then
I
realized
that
that
the
way
that
it
does
prefixes
is
it
puts
the
prefix
as
the
key,
if
I
understood
it
correctly,
and
the
the
prefix
length
as
the
value
of
the
map
and
since
we're
not
allowed
to
have
duplicate
keys
in
a
map.
E
This
presents
a
problem
if
you
actually
need,
for
some
reason
to
send
two
prefixes
that
have
different
lengths
along
the
thing:
it's
not
crazy
in
some
configuration
situations
where
you
may
have
multiple
prefixes
that
have
that
differ
only
in
their
length
for
some
reason,
maybe
simply
because
the
way
they've
derived
or
because
that's
the
way
the
guy
typed
them
in.
E
E
My
subnet
and
I
really
want
to
put
the
prefix
length
first
so
that
I
don't
know
it
just
seems
it
just
seems
more
pleasing
to
me
to
do
that,
to
be
able
to
look
at
that
first
and
then
say:
okay,
this
makes
sense
and
I'm
gonna
gonna
put
it
in
the
right,
the
right
place
and
then,
finally,
that
I
would
only
send
as
many
bytes
of
the
prefix
as
were
not
zero.
So
I
could
omit
trailing
zeros.
E
If
I
wanted
to
do
that,
I
got
a
private
message
from
brian
carpenter
saying
that
the
document
needs
to
be
more
clear
about
when
the
prefix
length
is
not
a
multiple
of
eight
that
the
trailing
bits
are.
Zero
must
be
set
to
zero
by
the
sender
and
must
be
set
to
zero
by
the
receiver,
and
he
had
concerns
about
covert
channels
and
things
like
that
being
there
otherwise,
and
also
because
the
prefixes
tend
to
be
used
combined
later
on
with
an
or
operation.
E
So
you
really
got
to
make
sure
that
the
trailing
zero
bits
really
are
zero.
You
get
kind
of
a
disaster,
so
the
v4
tag
looks
identical.
E
It
uses
the
same
tag
for
v
for
prefixes
and
addresses,
which
is
the
understanding
that
an
array
means
that
it's
a
subnet
or
prefix
and
the
no
none
means
it's
not,
and
if
the
v6
address
shows
up
in
a
context,
we
expected
a
subnet.
Well
then
the
assumption
is
this:
is
it's
a
slash,
128
or
a
slash
32.,
so
this
would
be
a
typical
are.
E
These
are
our
example:
192
168,
1
2.0
again
the
last
byte
in
this
one
is
omitted,
it's
always
zeros
and
it
needs
a
one
plus
one
tag,
ideally
so
that
needs
a
document
and
sir
carson
suggested
it
could
go
into
notable
tags,
and
I'm
happy
with
that
as
long
as
the
working
group
is
willing
to
do
an
early
allocation
of
it.
Since
that
document
is
at
this
point,
expired
even
that's.
It.
B
E
So
the
only
reason
I
would
not
want
to
go
into
the
notable
tags
document
is
if
the
working
group
didn't
want
to
do
an
early
allocation,
in
which
case
I
would
prefer
to
simply
progress.
This
document
asap,
but
as
long
as
any
whatever
gets
the
allocation
happening
sooner.
That's
that's
yeah.
What
I
I
I
care
about
and
I'm
not
sure
about
the
schedule
for
notable
tags.
At
this
point,
I
I
didn't
look
at
it.
It
was
expired.
B
So
it
sounds
to
me:
it's
a
little
bit
making
this
more
complex
so
pursuing
this,
as
your
own
document
for
now
seems
to
be
the
the
least
complex
way
of
doing
this,
we
can
always
merge
them
later.
It.
A
Also
makes
it
easier
to
get
reviews.
You
know.
B
A
E
So
it
just
occurs
to
me.
He
said
it.
E
E
E
No
well,
I
actually
so
yeah.
So
it's
a
it's!
It's
olay's
document,
six-man
universal
rra
option.
It
has
been
discussed.
I
think
the
working
group
might
have
discussed
adopting
it,
but
I
don't
think
they
have
yet.
But
the
point
is
that
that
might
be
interesting
to
be
able
to
tag
rather
than
having
to
do
a
v4
and
a
v6.
Well,
what
make
no
sense
to
put
a
v4
tag
into
a
v6
ra?
Maybe
okay,
we'll
forget
that
idea.
E
All
I
was
going
to
say
is:
it
might
actually
be
useful
in
there
because
router,
the
bit
bytes
in
a
in
router
options
are
are,
are
scarce
but
certainly
having
a
way
of
encoding.
The
v6
prefix
makes
sense,
but
it
might
be
that
they
don't
need
a
tag
anyway.
A
Okay
sounds
good
and
also
this
seems
to
be
in
scope
of
our
working
group.
Anybody
objecting
to
this
document
being
adopted
and
moving
forward
in
the
working
group.
A
Silence,
okay,
so
we
will.
We
will
send
an
adoption
call
in
the
main
list.
B
So
once
this
document
is
adopted,
I
probably
come
up
with
a
question
whether
we
should
be
doing
something
we
have
been
doing
for
the
oid
case,
where
we
can
apply
a
tag
to
an
array
of
things
and
have
all
the
things
in
there
be
interpreted
as
ipv4
addresses
or
ipv6
addresses.
E
E
The
array
semantics
and
you
know
basically,
can
you
distinguish
between
these
two
sides
of
things?
Yes,
obviously
enough,
I
mean
if
the
fir,
if
the
first,
if
it's
an
array
and
the
first
item
is
an
array-
then
that's
obviously
a
subnet.
E
But
if,
if
it's
an
array-
and
the
first
item
is-
is
not
a
an
int,
then
it's
a
address.
I
don't
know
that
seems
pretty
loaded
to
me,
but
I
don't
know
right.
B
E
E
It's
not
such
a
big
deal.
You
actually
spend
more
time
more
lines
of
code.
You
know
getting
the
the
ad
the
content
into
your
own
structures.
Then
I
do
figuring
out
what
it
is,
but
I
don't
have.
I
don't
have
a
require
my
I
don't
have
a
requirement
to
have
sufficiently
long
lists
of
these
things
that
I
would
need
to
do
that,
but
I'm
sensitive
to
other
people's
needs
here.
A
Sounds
good,
I
just
want
to
note
that,
yes,
we
will
do
this
adoption
course
and
please,
I
know
our
main
list
is
a
bit
slow
responding.
Sometimes
so
you
know
if
you
think
this
is
a
good
idea,
please
go
ahead
and
reply
to
the
mail
thread,
even
if
you're
wearing
the
call-
and
you
know
you
have
already
stated
that
you
think
they
should
be
done
and
you
do
not
object
so
that
helps
us
have
a
trace
of
of
consensus
and
adoption.
A
And
that
was
it,
so
we're
done
with
the
agenda
items
and
the
last
item
was
any
other
business.
So.
B
B
It's
maybe
worth
looking
at
what
what
happened
over
in
the
gid
spec.
I
still
don't
don't
know
whether
this
did
spec
will
ever
be
used
for
anything,
but
I
think
if
it's
there,
we
should
make
sure
that
they're
using
sibo
in
the
right
way.
So
I
just
advise
them,
for
instance,
to
no
longer
reference
7049,
but
reference.