►
From YouTube: IETF102-I18NRP-20180716-1430
Description
I18NRP meeting session at IETF102
2018/07/16 1430
https://datatracker.ietf.org/meeting/102/proceedings/
D
All
right
note:
well,
since
it's
Monday
I
will
actually
note
the
note
well,
so
there
are
a
bunch
of
BCPs
floating
around
in
this
organization
that
tell
you
what
participation
requirements
are.
Some
of
them
include
things
regarding
intellectual
property.
If
you
are
someone
who
is
worried
about
that,
you
better
go,
read
them
before
you
sit
here
and
look
cute
and
there's
other
things
about
other
behaviors
that
you
should
engage
in
and
not
engage
in.
Please
read
up
on
these
things.
D
D
So
the
on
administrivia-
and
this
is
this
Boff-
we're
going
to
do
pretty
much
chair,
driven
and
then
discussion
in
each
point
simply
because
I
think
it's
going
to
be
a
lot
of
explaining
what
the
problem
and
goals
here
are
and
the
mic
lines
will
be
open
as
we
get
to
each
topic
that
there's
not
a
long
discussion
of
each
thing.
In
any
event,
we'll
do
goals
and
on
goals
for
the
buff
up
front
review.
D
What
our
current
state
of
affairs
is
look
at
the
problems
with
the
current
state
of
affairs
look
at
some
strongman
proposals
that
were
put
together,
two
of
them
to
be
precise
and
then
figure
out
some
next
steps.
Anybody
have
complaints
about
this
structure
or
things
to
add
or
subtract
or
multiply.
I
love
happiness
and
silence,
always
a
good
thing.
D
Will
do
yeah.
We've
got
very
good
sound
in
this
place,
but
it
did
yes,
its
robust
I,
unusually
robust
myself
alright.
So
we
have
several
goals
for
the
session
today.
We
want
to
figure
out
the
problem,
space
of
internationalization
reviews
and
what
we
need
to
do
with
them.
We
want
to
come
to
some
high
level
agreement
on
solutions
to
the
problems
that
we're
having
doing
these
reviews.
Nowadays,
we've
got
two
proposals
that
are
on
the
list
and
we
will
describe
those
two.
D
They
either
come
down
to
basically
an
internationalization
review
team
or
writing
a
internationalization
considerations,
RFC
and
and
yes,
and
that
is
a
non-exclusive
or
it's
a
proper
or
so
both
of
those
things
might
occur.
The
group
might
decide
that
both
of
those
things
are
worth
doing.
Specific
non
goal
is
to
figure
out
how
to
handle
the
larger
internationalization
initiative,
some
things.
D
B
B
So
issues
come
up
and
there's
a
number
of
things
that
over
the
years
we
have
defined
that
are
have
impinge
upon
internationalization
character
sets
if
you
go
and
look
at
the
the
background
materials
RRC
2277.
So
some
of
these
things
have
been
around
for
quite
some
time
a
utf-8
there
are
other,
also
encoding.
So,
for
instance,
you
are
eyes.
Sometimes
we
try
to
encode
unicode
fold,
unicode
characters
outside
the
ascii
range
into
your
eyes,
internationalized
domain
names.
We
have
had
two
versions
go
rounds
on
that
ID
and
a
a
1980
2008.
B
We
have
some
non
DNS
protocol
identifiers,
so
we
had
precis
work.
That
was
done
to
address
a
lot
of
that
and
there's
a
lot
of
security
implications
which
I
forgot
to
put
the
RFC
in
there.
But
it
was
an
IAB
document
about
authorization
and
authentication
with
regard
to
internationalize
strings
and
so
interest
very
interesting.
Things
can
happen
there
when
we're
doing
in
comparison
want
to
know
if
something
is
allowed
to
have
access
to
something.
So
these
are
some
and
Alexei
and
I
actually
have
when
we
had
to
review
the
art
reviewers
going
at
one
point.
B
B
We
have
Internet
drafts,
which
are
not
directly
about
internationalization,
but
in
have
things
in
them
that
needs
to
be
looked
at,
so
they
might
have
some
text
strings.
Even
things
like
error
messages
that
can
be
sent
around.
Do
we
want
to?
How
do
we
want
to
handle
those
positive,
so
there
might
need
to
be
flagged
as
even
if
it's
utf-8
well,
you
know
it
does
need
a
language
tag.
Things
like
that,
so
we
will
less
things
comfortable.
Look
at
those
there's,
some
I
would
say
kind
of
relatively
straightforward
applications
of
some
of
our
existing
technology.
B
B
How
do
we
have
folks
look
at
these
things?
Well,
sometimes
people
reach
out
early
on
because
they
realize
there's
some
areas
that
they
don't
know
as
much
about,
and
so
they
might
need
some
help
and
that's
my
understanding
of
what
happened
with
lamps.
Our
working
group
I
think
people
reached
out
to
maybe
you
and
Patrick
ahlstrom
to
say
hey,
you
know,
we
don't
have
expertise
here.
We
need
some
help
that
works
well.
B
Sometimes,
if
people
are
having
awareness
that
they
need
to
reach
out
oftentimes
things
get
flagged
in
during
iest
review,
sometimes
earlier
in
working
group
last
call
or
ITF
last
call
often
area
directors.
A
current
area
directors
can
correct
me
if
I'm
wrong,
but
I
know
this
is
where
things
often
came
up
in
the
past
and
then
whoever
gets
flagged
says.
B
Oh
well,
we
need
to
find
someone
who
knows
about
this,
so
maybe
it's
John
or
maybe
it's
me
sometimes
Reznik,
Patrick
Falls
from
other
folks,
depending
on
what
the
topic
comes
up,
we
might
take
a
look
at
things
worst
case.
Maybe
people
weren't
really
recognizing
that
there
was
an
issue
and
and
review
doesn't
happen
at
all.
Hopefully,
that's
not
really
what
happens,
but
we
don't
really
have
a
systematic
way
to
find
what
would
needs
review
and
who
does
these?
B
Well,
we
have
some
people,
they
know
some
things
and
people
find
them
and
see
if
they
can
apply
have
that
knowledge
apply
to
the
things
they're
working
on.
This
is
kind
of
okay,
in
the
sense
that
you
know,
we
typically
find
the
right
folks
sometimes,
but
it's
kind
of
ad
hoc
right,
so
someone
maybe
reaches
out
to
John
or
they
reach
out
to
me
or
someone
else
and
say:
hey
I
could
use
some
help
on
this.
B
If
they're
aware
we
don't
you
know
so
that's
kind
of
how
things
work
now,
it's
sort
of
ad-hoc
adhocracy,
if
you
would
and
then
we
sort
of
informally
provide
feedback.
It's
a
one-off
thing.
We
don't
necessarily
capture
what
gets
said,
and
so
we're
not
really
you
know,
figure
it
out
have
a
stable
of
advice
that
we
provide
Pete.
You
want
to
talk
about
some
problems,
I
like
to
talk
about
the
easy
stuff,
yeah.
E
D
They're
not
reliably
identifying
the
documents
needing
review
the
the
worst.
The
worst
case
scenario
was
not
identifying
them
at
all,
but
even
having
it
happened
during
iesg
review.
I.
Think
all
of
the
art
area
directors
for
the
past
few
years
have
gotten
very
used
to
finding
keywords
in
all
sorts
of
other
our
internet
drafts
and
going
that's
going
to
be
an
internationalization
thing,
because
you
use
the
word
string
and
probably
we
need
to
go
look
at
that,
but
that's
not
sustainable
and
it's
very
late
in
the
process.
D
We
also
have
a
pretty
small
number
of
people
who
we
can
hand
these
things
off
to
with
at
least
reasonable
chops
it.
None
of
us
I,
believe,
are
saying
that
we
have
to
have
the
expert
of
all
experts
do
this
every
time,
but
we
need
people
who
at
least
have
read
through
and
understand
the
issues
well
enough
to
be
able
to
identify
that
there
might
be
an
issue
there
and
the
problem
there
is
that
we're
not
training.
D
F
G
D
A
couple
of
proposals
have
been
put
out:
one
is
to
basically
form
something
that
we
do
in
lots
of
other
areas
as
well,
which
is
an
internationalization
Directorate,
so
not
including
documents
that
are
specifically
about
internationalization
per
se.
This
would
be
going
through
documents
on
some
sort
of
regular
basis
to
review
for
common
internationalization
issues
also
would
include,
if
requested
or,
if
discovered,
giving
guidance
to
particular
working
groups
on
the
applications
of
existing
internationalization
technology.
D
So
the
sort
of
thing
that
happened
with
lamps
trying
to
walk
through
and
help
solve
some
of
the
thorny
or
internationalization
issues
that
come
up,
although
again,
that's
the
kind
of
stuff
that
probably
requires
a
dedicated
working
group
to
address
for
a
particular
internationalization
kind
of
problem
space,
but
for
more
general
things,
certainly
that
Directorate
can
help-
and
you
know,
maybe
that
for
purposes
of
a
working
group
might
be
like
a
design
team
that
works
on
internationalization
internationalization
stuff.
In
the
first
place,
the
other
proposal-
I'm,
sorry
we're
still
one
Directorate.
D
The
the
idea
is
much
like
the
gen
art
or
sector.
It
would
not
review
every
internet
draft
that
went
through
simply
just
time
constraints
there,
as
I
was
talking
about
with
regard
to
a
bee's
doing
it.
There
there's
a
relatively
easy
way
to
triage
documents,
to
figure
out
whether
it's
worth
giving
them
a
more
thorough
review
to
look
for
particular
kinds
of
data
that
is
floating
around
in
a
particular
protocol
spec
where
necessary.
B
D
Other
thing
going
on
there
is
I
can
tell
you
that
I
have
the
I
can't
describe
to
you
what
the
possible
internationalization
problem
is,
but
I
know
it
when
I
see
it
so
those
of
us
who've
been
doing
it
a
while
and
know
the
issues
can
kind
of
go.
You
know,
that's
gonna
be
a
problem,
but
some
of
it
is
complicated
to
describe
what
exactly
you're
looking
for.
If
it
says
string,
that's
easy.
D
H
Dan
is
this
is
barely,
but
Dan
is
not
asking
about
getting
it
down
into
it,
he's
just
asking
for
triage
and
at
the
triage
level.
Yes,
looking
for
strings
looking
for
utf-8
somewhere
in
there
there's
a
pretty
easy
way
to
say
this
document
probably
doesn't
have
internationalization
issues.
This
document
needs
a
little
more
look
but
you're
right.
No,
not
much
of
it
is
written
right.
I
This
is
Andrew
Sullivan.
If,
if
we
could
write
down
reliably
how
to
do
that
triage,
we
wouldn't
have
to
have
this
Boff,
and
that's
that's
actually
part
of
the
problem
here
that,
because
the
the
sense
is
that
the
sense
develops
really
from
the
from
the
observation
that
gee
there's
a
payload
here,
we're
actually
somebody's
supposed
to
do
something
with
this
payload,
and
that
turns
out
to
to
mean
that
there's
something
in
there
that
you
that
somebody
has
to
care
about.
I
But
there's
something
in
there
that
somebody
has
to
care
about
is
not
really
a
triage
strategy.
That's
a
you
know,
that's
a
somebody
needs
to
have
a
look
at
all
the
things
and
and
while
they're
going
along
people
need
to
need
to
notice.
Oh
well,
there
could
be
an
issue
here
and
then
you
send
it
off.
Yeah.
D
The
the
security
ones
were
I
think
the
more
interesting
discovery
that
those
of
us
who
did
these
reviews
noticed
because
the
security
people
would
sometimes
say
no,
no
we're
just
treating
that
as
a
binary
string
to
do
whatever
manipulations
on.
But
some
of
us
realize
wait
a
minute,
but
you're
get
that
by
that
supposed
binary
string
from
user
input
and
that
user
input
is
usually
going
to
be
through
a
keyboard
with
certain
features
and
and
certain
ways
of
encoding
that
input
that
might
be
different
between
different
keyboards.
D
F
Cleansin
here
we
may
be
using
triage
in
slightly
different
ways
or
not
understanding
in
the
same
way
where
that
stops
identifying
a
document
that
needs
further
attention
is
pretty
easy
and
we
could
probably
write
down
a
bunch
of
heuristics,
if
not
rules,
for
identifying
those
documents
and
get
it
right
almost
all
the
time.
The
difficulty
is
comes
in
with
what
happens
next
and
that
tends
to
require
some
specialized
expertise
and
discussion.
B
F
There's
an
additional
complication
there,
which
is,
if
you've
got
a
small
pool
of
people
who
can
do
the
work,
then
those
people
are
not
available
to
do
the
reviews
of
the
work.
If
you
wanted
to
pendick
review
and
one
of
the
things
which
is
interesting
about
the
list
of
ID's
who
add
up
earlier,
is
that
if
you
discount
the
people
who
had
heavy
involvement
in
writing
of
those
IDs
from
the
list
of
available
reviewers,
the
pool
of
available
reviewers
gets
really
small.
I.
B
Totally
agree
the
spears
in
Andrzej,
by
the
way,
sorry
and
I-
think
that's
where
there's
two
things
there
I
think
John
one!
Is
that
that's
where
things
order
on
perhaps
a
design
team
in
the
sense
that
let's
say
we
want
to
address
these
fun
issues
that
we
had
with
you
know,
unicode,
seven
and
and
so
on.
We
did
have
a
small
group
of
people
who
had
awareness
of
what
those
issues
were
and
looked
into
them,
and
they
me
you
know.
B
D
H
The
slide
says
small
team.
To
start,
maybe
five
people
we
have
had
I,
don't
know
whether
we've
had
a
formal
ie
TN
Directorate
before,
but
we've
at
least
had
a
de-facto
one
before,
and
we
have
had
an
IAB
internationalization
program
before
and
both
of
them
have
collapsed.
What
makes
us
think
that
establishing
a
new
one
will
have
a
different
result:
Ted.
C
J
Tory
I
would
actually
say
that
it's
different
in
two
ways,
one
of
which
is
very
salient
to
this,
and
one
which
is
perhaps
a
bit
more
nebulous.
One
of
the
things
that
makes
it
a
bit
more
nebulous
is
that
in
many
ways
an
ITN
program
of
the
IAB
must
buy
the
Navy's
nature,
think
about
liaison
relationships
and
how
it
interacts
with
other
bodies.
J
J
The
other
thing
is:
people
are
driven
in
different
ways
to
participate
in
things,
and
people
who
are
driven
into
Directorate
review
are
often
simply
trying
to
make
sure
that
the
work
of
the
ITF
completes
successfully,
and
that
may
not
be
the
same
driver
for
people
who
are
trying
to
see
new
frameworks
or
new
resolution
mechanisms.
That
may
be
significantly
different
from
where
we
are
today.
It's
like
saying
the
IRT
F
and
the
ITF
may
have
different
participants
because
they
have
different
goals.
K
H
I
reviewed
a
document
relatively
recently
that
just
thought
it
was
sufficient
to
say
this
field
will
be
case,
insensitive
utf-8
and
you
know
they
thought
that
was
enough
and
and
there's
no
reason
to
think
they
shouldn't
think
that's
enough,
because
that's
you
know
that's
as
far
as
they
knew.
They
didn't
know
to
tag
that
from
more
information
or
they
would
have
come
to
us.
So
there's
that
issue,
but
the
w3c
has
recently
put
out
a
statement
about
increased
focus
on
internationalization.
Can
you
say
something
about
that?
Thank.
K
Managing
and
doing
some
of
the
internationalization
review
work,
and
we
recognized
the
w3c
community
recognized
that
that
was
not
enough
and
recently
launched
an
internationalization
sponsorship
effort
trying
to
bring
in
more
support
to
fund
more
of
the
involvement
of
experts
in
internationalization
work
across
a
variety
of
areas,
including
identifying
gaps
in
the
platform,
as
well
as
reviewing
and
improving
the
work.
That's
being
done
without
a
specific
internationalization
focus.
B
Hi,
this
is
Peter
again,
I
am
actually
involved
with
the
internationalization
working
group
at
the
w3c
I
think
it's.
It
works
quite
well,
especially
the
tagging
of
issues
and
things
like
that.
Don
and
I
are
on
those
calls
quite
often
and
I
think
that
work
has
done
well
to
Berry's
point.
You
know
why
do
we
think
we're
going
to
get
a
different
result?
This
time,
I
think
it
is
related
a
little
bit
to
what
you
said.
Ted
about
different
kinds
of
people
involved.
B
We
haven't
done
I
would
say
a
good
job
of
training
people
and
recruiting
people
to
work
on
these
efforts
at
the
ITF.
We've
had
some
tutorials
here
here
and
there
I
think
that
if
we
were
working
on
potentially
working
on
some
of
the
less
contentious
issues
where
we
do
need
to
do
sort
of
regular
reviews,
I
know
and
at
w3c
we
have
a
weekly
call
to
go
over
that
the
things
that
are
under
review.
B
If
there
were
a
regular
cadence
of
calls
that
we
had
or
meetings
of
some
kind
of
wherever
reviewing
the
documents
come
through
and
triage
and
things
together
and
that's
what
we're
I
think
we
get
back
to
working
as
a
team,
because
we've
got
individuals
who
know
about
things,
but
we
haven't
really
had
it's
more
of
a
team,
coordinated
effort.
I
think
that
would
help
people
feel
more
committed
to
that
and
be
working
on
things
ongoing.
That's
my
perception
anyway.
Yeah.
H
C
M
M
Theory,
I
think
the
purpose
for
directorates
is
to
and
they're
all
sort
of
chartered
by
some
portion
of
the
IHG
right.
So
typically
they
all
roll
up
to
some
ad
some
set
of
ATS
and
those
ADEs
have
some
responsibilities:
vzv
the
directorate.
Typically,
that
would
include
like
maintaining
the
membership
refreshing,
the
membership
being
involved.
In
some
of
these
discussions
about
recruiting
likely,
you
know
establishing
that
the
review
cadence
or
the
expectations
about
what
happens
to
these
reviews
and
whatnot
and
I
think
it's
important.
N
M
B
N
This
is
your
idea.
My
comment
about
the
Directorate
Licht
rate
is
how
we
can
obtain
diversity
of
the
expert
from
all
over
the
world,
because,
as
you
see
then
most
of
people
here
is
from
jacket,
states
or
European,
and
very
few
Asian,
sir
or
Arabic
such
languages.
But
the
population
of
such
language
is
very
wide
and
high
penetration
in
the
world.
So
the
story
says
five
is
small.
I
think
also
and
invites
much
more
language
expert
from
the
wallet
it's
very
important.
D
Yeah
I
think
that's
probably
right.
I
think
one
saving
grace
is
that
for
the
the
80/20
rule
is
that
80%
of
these
things
are
things
that
can
be
dealt
with
by
anyone
who
has
basic
expertise
and-
and
it's
probably
even
more
than
80%
I
would
guess,
but
the
the
there's
a
chunk
that
actually
somebody
with
appropriate
skills
would
be
required
for
and
so
yeah
I
think.
The
point
is
well-taken.
L
Chris
Newman
another
point
I
wanted
to
make
about
Directorate
is
there
may
be
people
who
don't
want
to
be
like
full
members
of
a
Directorate
and
commit
a
lot
of
time,
but
are
happy
to
help
out
and
yo
II
have
provided
a
way
to
make
sure
they
stay
up-to-date
on
on.
What's
the
you
know,
the
latest
things
that
should
be
watched
for
and
in
help
with
reviews,
so
I
encourage
you
to
consider
making
room
for
that
sort
of
participant.
F
First
of
all,
I
would
have
agreed
with
80/20
six
months
ago,
but
my
experience
and
a
number
of
things
have
happened
in
the
last
six
months-
is
such
that
lowering
that
number
quite
a
bit.
The
second
observation
is
that
to
do
well
on
this
beyond
a
certain
point
in
this
goes
to
yashida's
son's
comment.
F
You're
going
to
have
to
recruit
some
outside
expertise.
The
difficulty
with
rikuo
nura
difficulties
with
recruiting
outside
expertise
is
that,
if
we
inflict
upon
those
people,
in
addition
to
our
higher,
is
and
rich
compensation
packages
uncle
Dylan
IETF
bureaucracy,
they
won't
be
here.
So
one
of
the
issues
is
that
it
the
more
we
push
this
into
the
traditional
Directorate
box,
the
less
likely
it
will
be
to
succeed,
given
that
need
for
outside
expertise
and
the
time
issue
mention
a
number
of
other
things.
So
be
a
little
careful.
There.
M
Elisa
Cooper
I,
don't
really
know
what
bureaucracy
that
is.
I
haven't
really
seen
that
I
actually
got
a
piano,
responder
Chris
as
well,
just
to
note
that
in
general,
for
example,
we
have
like
a
diversity
of
review
Cadence's.
So
you
can
have
people
who
are
part
of
the
Directorate
who
review
once
a
quarter,
and
you
can
have
people
who
review
twice
a
month.
There's
probably
not
enough
people
in
the
world
to
populate
this
thing
that
that
would
be
a
useful
model.
But
if
there
are
then
it's
not
like
has
to
be
one
size
fits
all.
O
Adam
Roach
actually
got
up
here
to
ask
Chris
if
he
could
clarify
something
about
his
comment.
I
can
read
what
you
said:
two
different
ways
and
one
is
I,
think
what
ELISA
kind
of
alluded
to,
or
it's
like
someone
who
doesn't
necessarily
have
the
time
commitment
but
might
want
his
date
like
either
on
an
infrequent
but
periodic
basis,
or
you
know,
I
want
to
be
the
person
who
is
involved
for
internationalisation
as
it
pertains
to
say,
email.
O
The
other
way
that
I
can
read
it
is
you
know,
sort
of
experts
on
tap
that
you
can
reach
out
to
who
are
experts
on
you
know
like
Indic
languages,
you
can
say:
hey
I
need
some
input
on
this
and
I
am
not
sure
exactly
which
of
those
two
you
are
referring
to
the
second
one.
I've
actually
already
heard
the
first
one
is
kind
of
new
and
I'm
just
curious.
L
Chris
Newman
no
I
was
more.
It
was
more
time
constraint,
you
know.
If,
if
you
like
it,
you
know,
if
you
ask
me
to
participate
and
not
had
to
be
on
a
bi-weekly
meeting,
I'd
say
no.
If
you
asked
me
to
participate
and
I
was
reviewing
a
document
every
two
weeks
or
something
I
might
I
might
very
well
say
yes,.
J
J
B
B
You
know
you
know
in
a
in
the
particular
internet
context,
but
to
reach
out
to
them
with
regard
to
that,
instead
of
having
them
necessarily
involved
all
the
time
that
we
people
that
we
can
call
on
as
needed,
I
I
would
see
that
model,
probably
working
better,
because
those
are
those
deeper
issues.
The
day-night
might
might
not
want
to
be
triaging
and
reviewing
all
the
time.
I
think.
B
Triage
is
a
function
of
whatever
team
gets
put
together,
whether
it's
a
Directorate
or
something
else
they
need
to
perform
triage,
but
then
once
they
figure
out,
this
document
needs
review.
Oh,
this
is
a
straightforward
issue.
We've
got
a
RFC.
Perhaps
that
talks
about
these
things.
We
pretty
much
know
what
to
do.
We've
documented
what
the
usual
right
answer
is
here.
That's
one
thing,
but
they
know
now.
We've
got
a
deeper
issue.
P
Harilal
lustrum
know
just
just
want
to
remind
people
about
the
original
meaning
of
triage,
which
was
the
Roman
legions
after
the
battle
had
a
process
by
which
the
doctor
looked
at
you
and
divided
you
into
one
of
three
groups,
one
is,
will
get
healthy
anyway.
This
is
likely
to
be
the
ones
that
you
can
just
give
them,
give
them
a
piece
of
advice
or
even
say
you.
P
F
John
cleansing
again,
let
me
point
out:
we've
again.
It
goes
to
some
extent
to
Ted's
comment.
Let
me
point
out:
we've
got
a
much
easier
in
some
ways:
problem
than
w3c,
for
example,
although
in
some
ways
harder
w3c
has
had
to
deal
with
challenges
like
how
you
punctuate
Mongolians.
So
it
looks
right
on
the
page.
My
suspicion
is
the
amount
of
IETF
expertise
in
the
punctuation.
A
Mongolian
is
quite
limited
that
if
you
put
together
a
by
person,
Directorate
team,
you
are
unlikely
to
have
any
of
it.
F
They've
got
a
whole
page
layout
issues
and
a
whole
lot
of
issues
which
are
very
specific
to
specific
issues
with
specific
languages
we
have,
as
far
as
I've
been
able
to
tell
none
of
that.
Our
hardest
problems
are
knowing
that.
If
you
do
something
in
a
particular
way,
it
is
likely
to
screw
up
some
language
you
have
never
heard
of,
and
that
requires
enough
broad
expertise,
as
distinct
from
language
by
language,
specific
expertise
to
detect
those
kinds
of
issues
and
at
least
to
understand
the
world
enough
to
get
the
expertise
needed
to
address
them.
F
H
This
is
Mary,
I
almost
entirely
agree
with
what
John
just
said,
the
only
quirk
we
don't
need
to
know
the
details
of
Amharic
or
something
or
Mongolian,
but
we
do
need
to
know
that
in
some
languages
this
issue
comes
up
so
like
we
need
to
know.
Arabic
has
variants
and
other
languages
have
other
things,
and
we
don't
need
to
know
how
to
resolve
that.
We
need
to
know
that
that
issue
crops
up
and
we
need
to
know
how
to
deal
with
it.
H
D
Given
the
lull
in
the
conversation,
let's
quickly
also
look
at
the
other
proposal,
which
is-
and
this
may
impact,
how
we
design
proposal
one
or
how
they
will
interact.
So
the
other
proposal
on
the
table
is
writing
up
an
RFC
to
describe
internationalization
issues,
so
this
is
not
going
to
necessarily
be
an
introductory
tutorial
for
beginners
to
learn
internationalization.
This
is
aimed
at
developing
expertise
such
that
you
could
do
different
levels
of
this
sort
of
review,
so
it
would
describe
key
issues
and
principles
for
reviews
and
documents.
D
D
H
B
Would
help
because,
as
I
say,
we
do
have
this
oral
tradition,
a
lot
of
its
in
John's
head,
but
other
people
like
you,
know,
Patrick,
hallström
and
another's
in
mark
McKay
and
Pete,
and
so
on.
We
find
knowledge,
but
we
haven't
captured
it
anywhere.
So
I
think
it
would
be
helpful
to
capture
more
of
that
on
paper
so
that,
as
we
do
bring
folks
up
to
speed,
they
have
an
awareness
of
what
some
of
the
issues
are
that
need
to
get
addressed
and
even
beyond
that,
I
think
how
to
think
about
these
sorts
of
problems.
H
H
John
has
done
at
least
two
plenary
presentations
on
internationalization
and
other
people
have
done
others
in
various
contexts
and
the
result
is
usually
people
walking
away
with
a
glazed
look
in
staggering
that
a
step
one
so,
but
if,
in
addition
to
this
and
like
I
said
this
might
be
a
fine
thing
to
do,
but
I
think
we
need
a
different
level,
something
with
addressed
to
people
with
no
a
I,
18
and
knowledge.
That's
telling
you
things!
D
Q
Q
So
that's
why
we
started
BCP
56
bits
to
document
that
knowledge
and
part
of
that,
and
why
I
think
that
this
is
important.
This
proposal
is
that,
if,
if
you
don't
have
a
shared
truth
of
this
is
the
correct
thing
for
for
documents
to
do
in
this.
In
this
situation
or
all
rightie
of
specifications,
then
then
you're
gonna
have
contention.
You're
gonna
have
someone
coming
in
on
the
review
team
imposing
their
opinion
or
their
view
of
what
internationalization
is
which
may
not
align
with
the
rest
of
the
review
team
and/or
reality
so
I
think.
B
I'll
give
a
I
think
the
w3c
has
done
a
good
job
of
this.
The
there
was
a
character
model
for
the
internet
document
that
was
old
and
has
been
updated
and
a
lot
of
good,
better
things
have
gone
into
that,
and
so
yes,
having
something
along
the
lines
of
what
you're
saying
mark
I
think
would
be
very
helpful.
R
Dave
Taylor
so
I'm
going
to
go
back
to
one
of
John's
comments
and
then
give
two
analogies.
The
comment
of
Jones
I
want
to
focus
on.
Is
you
said
your
opinion
that
the
percentage
now
is
different
from
it
was
six
months
ago?
That
tells
me
that
there's
a
sizable
percentage
of
information
that
you
might
need
to
know
when
doing
a
review
or
even
a
triage.
That
is
not
necessarily
a
static
enough
that
as
soon
as
you
put
it
into
an
RFC,
it's
not
update.
R
The
two
analogies
that
I
want
to
give
is
one
is
the
way
that
mid
doctors
worked
now.
The
one
is
the
way
that
the
RFC
style
guide
worked.
Okay,
in
both
cases,
the
set
of
things
that
you're
reviewing
for
were
somewhat
dynamic
and
there
was
a
subset
that
became
relatively
fixed
and
so
both
of
them
kind
of
start
off.
R
They
didn't
have
to
wait
for
an
RFC
once
you
had
a
subset
of
that,
they
became
stable
enough
and
unchanging,
and
often
that
portion
of
it
moved
into
an
RFC
and
I
believe
that
happen
in
both
the
start
style
guide,
as
well
as
any
guidelines
for
moobs,
and
so
that's
the
model
that
I
think
one
would
actually
need
to
do
here
says
the
RFC.
Isn't
the
starting
point.
You
need
like
a
wiki
page
or
something
as
a
starting
point
now
might
be
a
subset
of
it.
That
goes
when
RFC,
so
that
may.
D
R
Think
the
first
step
is:
let's
charter,
a
working
group
ripoff
or
something
whose
task
it
is
to
define
an
RFC
I,
think
that's
the
only
way
to
frame
it.
I
completely
agree
with
a
showing
of
knowledge
right,
which
is
you
know,
the
style
guide,
admit
doctors
and
stuff.
The
original
sharing
was
not
on
IFC
right,
it
was
web
page
wiki
style,
so.
D
A
question
I'm
guessing
a
DS,
although
anybody
can
obviously
contribute
to
the
question
is:
is
there
a
willingness
to
charter
a
working
group
whose
output
is
not
an
internet
draft?
That's
going
to
go
for
RFC,
but
more
in
documentation
in
some
sort
of
space
that
that
will
force
people
to
concentrate
on
it.
I
I.
R
S
Alexei
Melnikov,
we
we
had
some
discussions
in
IHG
recently
about
documents
who
really
liked
not
to
be
published
because
they
have
no
temporal
usefulness.
You
know-
and
you
know
this-
this
is
not
necessarily
the
same
example,
but
I
personally,
don't
think,
don't
see
a
problem
with
a
group
which
is
chartered
to
maintain
certain
documentation
without
a
specific
outcome.
As
long
as
we
can
measure
success.
H
Dan
York
and
I
was
just
to
echo
days
comments.
That
was
my
comment.
Was
it's
not
an
RFC?
It's
it's
a
set
of
things.
This
is
when
I
think
about
the
work.
I
was
doing
earlier
about
deployment
of
various
protocols.
You
know
it
was.
It
was
a
suite
of
materials.
It
was
a
thing
like
that
to
Lexi's
point
yeah
this.
There
is
a
value
in
looking
at.
How
do
you
do
this
in
a
way?
That's
dynamic.
That
is
something
that
you
do.
The
other
piece
is
to
realize
that
we
don't
have
to.
H
You
know,
do
all
this
alone.
There
are
many
other
resources
out
there.
You
know
such
as
the
w3c
s,
but
others
other
entities
that
are.
This
is
not
a
new
topic
right,
so
part
of
the
thing
is:
can
you
gather
people
who
can
look
at
what
are
the
ones
that
are
useful
for
people
to
know
what
are
the
tutorials
that
have
been
published
that
are
useful
in
some
way,
but
it
just
it's
it's
more
than
one
document.
It's
a
suite
of
something.
D
F
F
F
But,
but
that
particular
guideline
applies
to
a
lot
of
the
stuff
we
do
here
in
terms
of
of
having
the
intuitions
for
quote
triage
end
quote.
The
thing
which
worries
me
about
the
wiki
kind
of
approach
is
you've
got
to
be
sure
that
whatever
that
stuff
is,
and,
however
you
presented,
which
makes
much
less
difference,
is
carefully
reviewed,
so
it's
actually
right.
Yep.
I
I
There
are
lots
of
people
who
are
working
on
these
things,
which
are
true
and
now
there's
the
suggestion
for
a
document
here
and
I.
Think
it's
worth
noting
that
one
of
the
problems
that
we've
had
is
that
we
are
not
alone
in
this
space,
but
there
are
a
number
of
people
who
seem
to
want
to
own
it,
and
so
one
question
that
I
guess
I
have,
if
we're
going
to
tackle
it
with
either
of
these
techniques
or
both
is
what
we're
gonna
do.
I
If
somebody
comes
along
and
says
none
of
you,
people
are
qualified
to
do
this
and
none
of
you
people
should
be
doing
it
and
please
sit
down
and
shut
up
and
go
away,
and
let
us
take
back
our
our
proper
area
of
expertise
and
I'm.
I,
don't
want
to
get
into
the
question
of
whether
there
is
such
a
proper
area
of
expertise
or
whether
any
of
the
experts
in
question
actually
have
that
expertise.
But
what
do
we
do
about
that?
I
O
This
is
Adam
Roach,
just
I,
think.
If
a
group
came
to
us
and
said
you
know
we
phrasing
what
you
said:
they're
effectively
stopped
trying
to
get
this
right.
We
would
say:
okay,
but
it's
a
none
company's
incumbent
on
you
now
to
take
on
the
load
of
making
sure
that
we
do
and
I
think
that
would
actually
not
be
a
terrible
outcome,
assuming
they
actually
were.
Experts.
J
J
Team
and
also
building
up
the
connections
between
particular
expertise
outside
the
ITF
and
bringing
it
in
time
can
can
take
place
over
time.
We
don't
have
to
run
up
against
this
problem
and
say:
I
hit
my
head
on
the
bricks
and
now
I'm
going
to
sit
down.
We
we
can
say:
okay,
there's
some
grout
I
can
start
pulling
it
at
this
particular
brick.
Get
it
a
little
bit
better
and
and
move
on
from
there.
So
I
think
that
there
are
some
some
action
items
we
can
take
from
this.
J
That
really
do
result
in
things
getting
better
without
necessarily
us
fooling
ourselves
that
they're
going
to
resolve
the
entire
problem.
I
mean
I
had
a
conversation
at
one
of
the
ICANN
meetings
with
a
language
expert
in
Ethiopian
languages
which
cost
multiple
different
language
families
using
a
single
script,
a
different
problem
than
the
one
John
was
just
alluding
to,
and
it
just
reminds
you
that
it's
always
a
little
bit
worse
than
you
thought
in
the
internationalization
world.
J
M
It
is
to
get
right
all
of
the
things
which
are
all
true,
but
if
the
notion
is
to
eventually
try
to
involve
more
people
in
this
work
and
inspire
people
to
look
at
this
area
and
say,
hey,
there's
some
hard
problems.
Maybe
those
would
be
fun
to
work
on.
Maybe
I
would
find
them
compelling,
as
somebody
who
is
new
to
networking
or
natural
languages
or
both
I
think.
B
I
think
this
is
Peters
and
Andre
I
think
that
gets
back
to
what
ted
is
talking
about,
which
is
incremental
improvement.
We
are
where
our
base
is
very
low
right
now
we
were
very
ad
hoc.
We
don't
bring
people
along
into
the
into
learning
about
these
things,
I
think
if
we
had
more
structure
a
little
bit
more
structure
around
it
and
we
could
try
to
bring
lift
up
the
level
of
knowledge
about
some
of
the
folks
that
participate.
That
would
be
great.
It
is
a
very
difficult
area.
B
F
Actually
I
was
going
to
say
something
quite
contradictory:
I,
don't
believe
that
this
work
is
fundamentally
harder
than
a
deep
understanding
of,
for
example,
internet
routing,
and
what
all
the
issues
are
there?
It's
just
different
and
our
difficulty
is
that
we
have
not
a
lot
but
a
reasonable
number
of
people
here
who
have
that
level
of
understanding
of
of
internet
routing,
and
we
have
a
much
smaller
number
of
people
here
who
have
that
level
understanding
of
this
problem
in
part
because
it
isn't
taught
in
computer
science
network
engineering
school.
F
The
other
observation
is
that
I
am
all
in
favor
from
making
incremental
improvements,
but
what
stimulated
Patrick's-
and
my
request
for
this
off
was
the
fact
that,
ultimately,
a
binary
decision
has
to
be
made
by
the
area.
Directors
in
the
is
G,
and
that
binary
decision
is
about
whether
or
not
we
are
ready
and
competent
to
review
any
particular
document,
and
that
decision
may
have
some
fuzzy
edges,
but
it's
ultimately
not
a
making
things
a
little
bit
better
kind
of
decision.
It's
say
when
are
things
good
enough
decision.
D
So
we
are
right
at
time
and
I
think
we
can
just
creep
over
a
minute
or
two
to
at
least
for
us
to
express
what
we're
getting
a
sense
of
from
the
room.
I've
first
heard
that
people
seem
to
think
that
a
director
at
a
group
of
some
sort
is
an
appropriate
thing
to
undertake.
It's
gonna
have
a
range
of
tasks,
maybe
with
a
range
of
participant
types
which
include
triage
include.
D
Some
more
expert
review
include
the
ability
to
maybe
give
a
document
or
a
particular
issue,
to
a
super
expert
to
help
with
a
particular
gnarly
issue,
I'm
also
hearing
that
an
internationalization
RFC
is
not
going
to
be
the
target
as
much
as
some
sort
of
internationalization
documentation,
which
might
involve
an
RFC
might
involve
some
more
dynamic
text
and
there
getting
a
group
together
to
work
on.
That
would
be
a
good
thing.
T
D
Okay,
so
I'm
seeing
five
hands
in
eight
yeah,
okay,
so
a
smallish
group
are
there
folks
in
the
room
who
feel
like
they
have
the
time
and
inclination
to
learn
more
deep
expertise
and
contribute
to
such
reviews
and
and
yes
about
the
same
number
and
different
people.
Note
so
we're
talking
in
the
dozen
ish
category
and
for
what
that's
worth
to
forming
such
group.
F
D
N
U
D
I,
like
that
idea
very
much
so
right,
so
yes,
and
so
part
of
next
steps,
which
Harold
just
mentioned,
is
what
form
are
willing
to
use
to
continue
this
discussion
and
we
have
a
mailing
list.
Do
we
need
to
do
something
in
addition
to
that
and
what
are
our
timeframes
for
next
steps?
He
turns
to
the
ADEs
who
might
be
implicated.
B
O
A
mailing-
this
is
probably
just
trying
for
this
next
steps.
I
think
I
want
to
confer
with
my
co
80s
to
sort
of
figure
out
what
a
reasonably
forward
here
is.
I,
don't
think
it
can
take
a
lot
of
time
because
it
was
you
know,
I
think
the
messages
we
heard,
who
were
relatively
clear
but
I
just
want
to
make
sure
on
the
same
page.
First.