►
Description
Document editing and sheperding at IETF93
2015/07/19 1500
A
Edd
you
team,
stuck
in
there
I,
never
said
editing,
we're
not
going
to
talk
about
document
editing
if
you're
here
to
find
out
anything
more
about
XML
to
RFC
or
enrol
for
any
of
that
stuff
goes
somewhere
else.
But
if
you're
here
to
talk
about
document
shepherding,
that's
what
this
is.
So,
let's
get
started.
A
Rfc
4858
is
the
RFC
that
started
off
the
whole
document.
Shepherding
thing
the
title
of
it
is
document
shepherding
from
working
group
last
call
to
publication,
and
you
notice
it
starts
at
working
group
last
call
it
shifts
work.
The
purpose
of
it
is
to
shift
work
from
the
responsible
ad
who
had
to
do
all
this
checking
to
someone
who
has
been
following
the
working
room
and
has
some
idea
of
documents
history
and
can
take
some
of
this
work
off
of
the
responsible
ad
and
coordinate
with
the
working
group
on
it.
A
In
addition,
it
provides
another
set
of
eyes.
It's
always
good
to
have
yet
one
more
person,
who's
responsible
for
doing
a
thorough
review
of
the
document,
so
it
started
off
with
working
group
chairs
and
secretaries
I.
Think
the
document
says
that
it's
going
to
be
a
working
group
chair
or
a
secretary,
but
there's
an
IAS
g
statement
about
non
working
group
chair
shepherds
and
it
actually
encourages
the
use
of
that.
A
So
we've
moved
towards
sometimes
using
other
volunteers
as
shepherds
and
still
the
chairs
are
the
ones
who
appointed
specifically
I've
got
a
couple
of
quotes
here
from
the
document
shepherding
document
RFC
4858
I'm
not
going
to
read
them
to
you,
but
basically
it
tells
you
what
the
primary
meet
the
objective
of
this
is
and
set
up
sets
up.
The
document
Shepherd
right
up
most
document
Shepherds
are
most
familiar
with
the
document
shepherd
right
up.
A
You
have
some
questions
to
answer
or
in
the
new
version
that
I
like
because
I
composed
it
it's
it's
I
call
it
the
essay
version
where
you
just
fill
in
four
sections
and
and
just
tell
me,
what's
different
about
this
dog.
What's
notable
about
this
document,
that's
mostly
what
the
Shepherd's
think
they're
doing.
A
A
A
Here
are
some
tasks
that
this
is
actually
also
a
quote
from
4858.
The
bullets
here,
providing
the
document
shepherd
right
up
during
a
bee
evaluation.
Managing
the
discussion
during
ietf
last
call
following
up
on
feedback
during
iesg
evaluation
following
up
on
feedback,
and
the
discussing
comment
ballots
that
bia
SG
puts
it
in
general,
keep
following
up
and
keep
the
working
group
in
the
loop.
So
I
propose
some
time
ago
what
I
call
extended
documents
shepherding
and
what
I
mean
by
that
is
starting
earlier
in
the
process
and
really
focusing
on
being
an
active
shepherd.
A
That's
it
started
off
as
an
as
an
internet
draft
that
I
was
putting
through.
I
started
off
having
it.
This
is
my
opinion
and
people
didn't
like
that.
We
shouldn't
be
publishing
people's
opinions
in
the
IETF
stream,
so
I
said
I'll
send
it
through
the
independent
stream
and
now
that
I
etf
process
stuff
shouldn't
go
through
the
independent
stream.
So
then
I
went
and
put
it
through
as
a
proposal
for
IETF
consensus
and
it
kind
of
petered
out.
It's
now
in
the
in
the
working
group
shows
wiki.
So
it's
a
place
here.
You
are
all.
A
While
you
are
all.
If
your
chairs,
you
all
have
access
to
the
working
group,
chose
wiki
and
can
tweak
it.
Actually,
I'm
not
sure
I
don't
know
if
it's
restricted
to
chairs.
I
think
it
might
be
that
anybody
with
a
date
with
a
wiki
I
deacon
but
anyway,
so
if
anybody
here
is
allowed
to
edit
this
document,
that's
this
page
has
ideas
on
how
to
change
it.
You
should
do
you
should
do
that
you're
empowered,
but
what
this
document
does?
It
sets
out
a
timeline.
A
The
comments
don't
just
sit
there.
There
are
responses
for
them,
there's
an
iteration
of
the
draft
that
goes
through
when
there's
no
revision
of
the
draft
for
four
months,
the
ship's
that
that's
a
that's
a
failure
right.
The
Shepherd
should
be
keeping
track
of
this,
and
I
should
say
it's
not
always
a
failure.
The
dot
the
working
group
may
have
intentionally
set
this
one
on
a
back
burner
for
a
while.
The
point
is
that
the
shepherd
should
be
responsible
at
the
early
stages
for
making
sure
the
document
doesn't
languish.
A
If
the
working
group
means
to
progress
this
document
now,
the
shepherd
should
be
making
sure
the
working
group
is
working
on
it
and
progressing
the
document.
Now,
what's
there
in
in
the
wiki
now
are
my
suggestions
as
I
said,
feel
free
to
add
your
own
to
tweak
it,
but
the
the
reason
this
is
extended
is
first
of
all
it
emphasizes
the
depth
and
follow
up
and
making
sure
that
it's
not
just
about
doing
the
shepherd
right
up.
So
you
can
send
this
to
the
iesg
it's
actually
actively
following
the
document.
A
It
starts
off
by
encouraging
making
a
stronger
push
toward
using
non
chair
shepherds
not
just
to
offload
work
from
the
ad,
but
to
offload
work
from
busy
shepherds
to
give
other
people
an
opportunity
to
participate
at
a
higher
level
in
the
working
group
and
perhaps
sorry
busy
chairs.
Thank
you
to
give
up
give
people
a
chance
to
start
doing
some
chair,
related
jobs
and
prepare
themselves
to
be
chairs
that
sort
of
thing
and
it
starts
earlier.
It
gets
you
it
gets.
Somebody
involved.
A
The
idea
here
is,
as
soon
as
the
working
group
starts
discussing
a
document
if
this
suits
the
chairs
view
of
how
this
working
group
operates.
A
point
is
Shepard
then,
and
have
the
Shepherd
follow
the
document
through
all
the
stages
in
the
life
cycle
and,
of
course,
it's
suggestions.
This
is
all
flexibility.
You,
you
still
run
your
working
group,
the
way
you're
going
to
run
your
working
group.
A
We've
had
a
lot
of
suggestions
within
the
IAS
g4.
Well,
let's,
let's
tweak
something
here
in
our
process,
so
we
can
save
a
couple
of
weeks
in
iesg
processing
on
the
document
that
doesn't
shorten
the
documents
lifestyle
cycle
by
much
this
stage
here
working
group
document
usually
takes
up
most
of
the
documents
life
that
can
be
anywhere
from
six
months
to
12
months
to
we've
seen
six
years
eight
years
for
some
documents,
I
think
we
can
eliminate
the
six
years
in
eight
years.
A
A
So
here
are
some
of
the
examples
of
the
what
the
document
on
the
wiki
does
is
for
each
stage
in
the
life
cycle.
It
says
it
lays
out
some
tasks
that
I
think
the
Shepherd's
should
be
doing
at
that
stage,
and
so
here
are
some
examples
in
IETF
last
call
monitor
the
last
call
comments.
Make
sure
that
specifically
requested
reviews
arrived
so
there's
one
that
the
shepherd
right
up
says
what
what
reviews
do
you
think
are
needed
for
this
document?
If
you've
asked
for
some,
that's
the
some
some
reviewers
may
need
to
be
product.
A
For
those
make
sure
the
document
editors
respond
to
the
reviews.
We
often
see
Directorate
reviews
or
specifically
requested
reviews,
come
in
and
there's
no
response
to
them
and
no
changes
are
made
based
on
them.
That's
the
Shepherd's
supposed
to
be
watching
for
that.
Keep
the
dialogue
going,
making
sure
the
issues
are
dealt
with
and
seeing
to
it
that
all
of
this
comes
back
to
the
working
group
and
working
group
is
aware
of
what
changes
are
being
requested
and
agrees
with
them,
and
that's
a
really
important
thing
you're
going
to
see
at
every
stage
of
this.
A
This
is
iesg
evaluation,
and
you
see
the
last
bullet
is
pretty
much.
The
same.
We've
had
a
lot
of
questions
about
what
changes
happen
to
a
document
after
it
leaves
the
working
group
and
whether
the
working
group
still
agrees
with
what
the
document
says.
After
all,
those
changes
are
made
and
I
never
want
to
hear
anybody
say
the
ad
made
me
change
this.
That
may
or
may
not
happen,
but
regardless
of
whether
it
does
the
working
group
should
be
aware
of
that
and
should
still
agree
with
the
document.
A
Big
rest,
nudging,
authors,
nudging,
reviewers,
nudging,
chairs
nudging
gave
us
nudging,
everybody
all
the
way
from
the
initial
presentation
to
the
work
to
the
working
group
to
the
RFC
publication,
the
RFC
publication,
which
includes
after
we're
done
after
the
is
jeez
approved
it
you've,
still
got
off.
48
and
Steven.
Farrell
brought
up
a
question
today,
because
there
have
been
a
series
of
documents
that
had
quite
a
lot
of
discussion
during
off
48
to
the
tune
of
dozens
of
messages
between
the
authors
and
the
RFC
editor,
and
that
sort
of
stuff
is
often
missed.
A
A
And
I
give
an
example:
I
can
give
some
excel.
I
know:
we've
had
some
internationalization
issues
come
up
in
some
documents,
where
there's
been
a
lot
of
discussion
about
exactly
how
something's
not
quite
right
exactly
how
do
we
resolve
it
and
for
those
I
know
we
took
it
back
to
the
working
group
and
we
said
here's
what
came
up
during
all
48.
Does
anybody
have
any
questions
or
comments
about
this,
and
in
that
case
it
all
went
smoothly,
but
I
can't
quote
documents
I,
don't
have
do
you
remember
anything
in
particular?
Well,
in.
A
That
was
during
all
48.
Okay,
that's
I
thought
that
was
I
thought
I
had
to
discuss
on
that,
and
that
was
during
the
discussion,
but
that
was
still
during
iesg
evaluations.
Oh
right!
Yes,
yes,
yes,
yes!
So
in
weirds,
which
was
do
you
know
what
what
weird
is
about?
Other
people
might
not.
It
set
up
a
protocol
for
replacing
the
whois
protocol
and
it
involved
having
a
registry,
a
new
registry
that
I
Anna
had
to
keep
up
in
a
strange
way.
It's
not
a
typical
iono
registry,
where
a
document
says
put
this
in
registry.
A
If
the
registry
was
built
from
stuff
that
registrar's
sent
in,
and
so
we
had
a
lot
of
tweaking
of
that
during
isg
evaluation,
but
then,
when
I
Anna
actually
went
to
create
the
registry
after
approval,
there
were
some
more
issues
that
came
up.
We
had
to
go
do
more
tweaking
and
again
we
let
the
working
group
know
what
was
happening
and
nobody
nobody
commented
so
and
for
those
stages
we
usually
accept
that
no
comment
means
you
agree
at
other
stages.
In
the
documents
lifecycle,
you
may
want
explicit
a
few
people
explicitly
saying.
A
A
Shifting
too
much
to
the
Shepherd's
and
thank
you
in
thirty,
seven
languages,
I
believe
or
maybe
it's
34
anyway,
alright,
so
I'll
leave
it
on
this
slide
and
I'd
like
to
open
it
up
to
questions.
How
many
of
you
have
how
many
of
your
chairs,
so
we
got
about
half
the
room,
maybe
the
room.
How
many
of
you
who
are
not
chairs
have
shepherded
documents?
A
few?
A
A
A
C
A
So
I'll
tell
you
that
the
two
things
that
I
most
want
a
shepherd
to
do
that
often
don't
get
done
or
don't
get
them
properly.
One
is
a
really
useful
shepherd
right
up
and
it's
why
I
try
to
shift
it
to
the
essay
style,
shepherd
right
up
and
that
seems
to
have
worked.
Although
a
lot
of
people
don't
really
like
the
ship
that
that
format,
they're
used
to
the
question
and
answer
here-
are
12
questions
or
18
questions
or,
however
many
there
are
answer
each
one
and
you
see
a
lot
of
just
yes.
Yes,
no!
A
No!
Yes,
that's
not
terribly
useful.
What
I
want
to
see
when
I'm
reading
the
write
up
is
tell
me
about
what
happened
to
this
document
in
the
working
group
and
the
questions
that
just
have
yes/no
answers.
If
yes
or
no
is
the
norm,
answer
for
that
question,
I
don't
need
to
see,
and
so
the
Shepherd
right
up
the
essay
version
says
tell
me
a
little
bit
about
the
technical
aspect
of
the
doctor.
A
Tell
me
a
little
bit
about
what
happened
to
this
document
in
the
working
group
and
what
kinds
of
what
kinds
of
discussions
were
notable
and
were
there
any
controversies
here?
Is
there
anything
we
ought
to
be
aware
of
so
we
don't
ask
the
same
questions
over
again,
and
so
we
get
some
idea
of
how
the
consensus,
developed
and
pete
has
something.
B
B
It
was
proposed
standard,
but
they
really
don't
care,
so,
let's
just
make
it
informational
or
vice
versa.
If
the
answer
was,
we
absolutely
want
this
to
be
a
bcp,
because
we
wanted
it
to
have
musts
and
shoulds,
and
we
can't
do
that
in
an
informational
document.
You
might
get
a
different
answer
from
the
iesg.
That
assume
that
the
reason
the
questions
being
asked
is
because
it's
going
to
be
argued
about
and
so
give
an
explanation
for.
B
Okay,
if
these
bozos
on
the
isg
argue
about
this,
here's
why
we
did
it
that
way
or
here's
what
the
answer
is
same
thing
with
the
IPR
question.
Have
all
the
authors
acknowledge
that
the
IPR
assume
that
someone's
going
to
say
wait
a
minute
I
thought
that
there
was
an
IPR
disclosure
about
this
or
there
should
have
been?
Did
you
actually
check?
A
Yeah
I'll
drop
a
knee
here
and
there
and
the
risk
of
embarrassing
a
good
shepherd,
but
princess
karen
o
Donoghue
did
a
write-up
recently.
That
said
in
answer
to
that
question,
peach
is
brought
up
proposed
standard,
but
the
working
group
is
ambivalent
about
that.
We
considered
informational.
Here's
what
we
discussed.
We
settled
on
proposed
standard
now.
If
the
IAS
g
starts
to
discuss
that
we
can
say,
go
look
at
the
right
up
and
Karen's
told
us
some
little
bit
about
what
the
conversation
was
and
we
have
some
place
to
start
with.
A
I've
had
some
Shepherd
write-ups
that
answer
some
of
the
questions
that
I
started
having
as
soon
as
I
started.
Looking
at
the
document
as
a
shepherd
right
up
tells
me
the
answer
to
that
question:
I
don't
have
to
go,
make
a
comment
in
my
ballot
or
put
it
discuss
on.
It
is
I
already
have
the
answer.
That's
cool
and
then
the
the
essay
version
has
this
little
checklist
at
the
end
where
all
the
yes/no
questions
are
accumulated
and
you
can
just
go
through
them
and
say
yeah,
nothing,
nothing,
nothing
to
say!
Wait!
A
This
one
I
should
say
something:
the
idea
of
the
new
version
is
only
tell
me
what's
different.
Only
tell
me
what's
notable
about
this
document,
so
I
certainly
encourage
the
Shepherd's
to
use
that,
but
check
with
your
ad
make
sure
your
ad
is
happy
with
that.
Some
of
the
80s
aren't
crazy
about
it,
which
is
why
we
did
not
change
the
shepherd
right
up.
We
just
gave
the
the
two
versions
as
options:
hi
hi,
my
name
is
Debbie
coolio.
A
So
individual
solutions
and
what
we
generally
have
been
doing
is
trying
to
find
somebody
to
be
a
shepherd
for
the
individual
mission
it
or
it
also
offloads
the
the
job
of
watching
everything
happen
and
making
sure
he
gets
followed
up
from
the
ad.
It
also
gives
another
check
another
set
of
eyes
on
a
document
that
probably
hasn't
gotten
as
much
review
as
it
should
have
I
know.
Pete
is
really
against
doing
doing.
Protocol
work
as
may
be
sponsored
documents
because
they
don't
get
very
much
review
as
working
today.
C
C
A
A
The
author
approaches
an
ad
and
says
I
have
this
document.
It
seems
to
fit
in
your
area
with
your
Shepherd
informant
and
the
ad
can
say
sure
the
ad
can
say
no
but
go
check
with
this
ad
or
the
ad
can
say
it's
unlikely
that
any
ad
is
going
to
pick
this
one
up
and
I'm
certainly
going
to
recommend
to
them.
They
don't
that
has
happened
anyway,
so
that
was
there
were
two
things
that
I
had
that
could
make
that
that
I
would
like
to
see.
Shepherds
do
better,
that's
one
of
them
and
I.
A
C
A
That
idea,
I
can
go
look
through
my
comments
and
find
some
and
approach
the
Shepherd's
to
see
if
they
wouldn't
mind
my
calling
them
out
as
simpler.
The
second
thing
I
would
like
to
see
done.
Better
is
following
up
at
every
step,
I'd
like
to
see
the
Shepherd's
collecting
issues
from
the
reviews
making
sure
the
issues
are
tracked,
some
way,
whether
it's
through
an
official
tracker
or
just
the
Shepherd's
ticking
things
off
in
a
spreadsheet
and
making
and
having
the
shepherd,
make
sure
that
all
the
comments
have
been
dealt
with.
A
I
want
to
see
a
shepherd
come
to
me
at
the
appropriate
time
like
when
last
call
closes
I'd
like
to
see
the
Shepherd
come
to
me
and
say
we
have
two
issues
to
be
addressed
from
last
call
before
the
document
is
revised
for
IHG
evaluation.
After
a
is
g
evaluation,
ok
I've
got
three
changes
that
need
to
be
made
before
you
approve
the
document.
I
don't
want
to
have
to
guess
that
and
I
often
I
do
I
have
to
watch
what's
going
on
watch
all
of
it
myself.
That's
the
part.
A
I
really
want
the
Shepherd
to
take
care
of,
and
sometimes
I
you
guys
have
to
ask
ok.
I
haven't
really
been
following
this
very
closely.
I
haven't
really
had
time
to
follow.
It
very
closely.
Tell
me
when
it's
ready
for
me
to
click
the
approve
button
and
then,
of
course,
because
that's
my
responsibility,
I'm
doing
the
same
thing.
I
told
you
before
that
the
working
group
chairs
have
to
do
with
the
shepherd.
A
C
It
creates
a
dialogue
between
my
ad
and
myself
of
this,
where
I
should
ask,
or
is
this
apply
or
is
this
who-
and
it
would
be
great
if
there
was
a
point
of
contact
associated
with
that
question,
which
said,
if
you're
wanting
to
ask
about
this
particular
you
know
xml,
review
syntax
review
whatever
start
here,
and
then
you
can
work
your
way
to
an
expert
over
here
or
something
like
that.
Otherwise,
it
tends
to
be
especially
if
it's
areas
that
are
completely
unrelated
in
general,
with
the
working
groups
efforts.
A
It
when
I
get
action
I,
and
so
this
this
one
is
for
me
to
make
sure
that
this
stuff
is
collected
in
one
place.
There
is
a
page
if
you
go
to
the
iesg
web
page,
which
you
can
get
to
off
of
the
main
IETF
page.
There
is
a
page
for
directorates
and
that
lists
all
the
directorate's
I
need
to
make
sure
that
that
list
tells
you
whom
to
contact
for
each
of
these.
There
is
also
a
mailing
list
now
I'm,
a
mailing,
alias
whatever
that.
A
C
A
B
B
C
C
So
that's
one
suggestion
one.
My
request
is
that
what
happens
when
you
send
off
a
request
to
the
right
people
and
you
get
nothing
back
and
you
get
crickets
or
so
one
of
the
questions
that's
come
up.
Is
how
long
do
you
wait
before
you
say
we
attempted,
and
then
you
know
we
dealt
with
this
scenario
recently
and
we
just
basically
it
became
this
dialogue
between
my
ad
and
I
you're
gonna
be
my
new
ID.
So,
but
how
do
we
wait?
How
long
should
we
wait?
What
should
we
do?
Should
we
go
to
someone
else?
C
A
I'll
start
at
the
end,
is
this
good
enough
for
the
iesg
review?
The
answer
to
that
is
what
what
your
judgment
is
on.
What
you
think
is
good
enough
and
that's
going
to
depend
on
how
confident
you
are
if
you
have
three
lines
of
simple
a
dnf
and
you
can't
get
anybody
to
review
it
for
you,
but
you've
copied
it
from
something
else,
and
it
looks
like
it's
written
all
right.
You
know
ask
for
somebody
to
take
a
look
at
it
if
nobody
does
you're
confident
enough
with
any
of
this.
I
want
that.
A
So
even
for
the
have
all
the
authors
confirm
that
they
have
made
all
the
eye.
Dr
disclosures
and
they're
performing
with
bcp
78
and
79.
The
boilerplate
in
the
in
the
draft
says
that
this
conforms
with
bcp
78
79
number
of
people
have
complained.
Why
are
you
asking
me
this
stupid
question?
I've
just
submitted
20
drafts
that
all
have
the
same
boilerplate?
Yes,
I'm,
conforming
with
it.
A
If
you've
got
people
like
that
in
your
working
group-
and
you
know
them,
then
you
know
that
they're
reliable
with
this
I
don't
have
to
ask
the
question:
is:
can
you
in
good
faith,
confirm
that
this
is
that
everything's
in
control?
It's
your
judgment
and
if
you,
if
you
want
Amit,
if
you
want
to
confirm
that
by
asking
the
authors,
that's
great
and
most
chairs
do.
If
you
want
to
say
its
peak
I
nope
it's
on
top
of
it,
you
don't
have
to
ask
that's!
A
Perhaps,
and
that's
your
judgment
now,
if
somebody
comes
later
an
ad
who
knows,
XML
really
welcomes,
and
so
are
you
know
an
ad
who
knows
battery
charging
crap
goes
and
looks
at
the
battery
move
and
puts
of
this
mega
discuss
on
it
like
Pete
with
one
you
know
you
may
get
the
often
you
get
somebody
who
has
a
special
expertise.
You
never
knew
they
had
and
they
come
out
and
they
say
wait.
A
A
The
questions
other
comments,
please
so
yeah.
My
big
message
here
is
that
we
need
to
make
sure
that
the
resources
that
you
guys
need
to
do
the
shepherd
job
are
readily
available
to
you.
You
don't
have
to
go
running
all
over
the
place
to
do
it.
Oh
the
other
part
of
your
question
was:
how
long
do
you
wait
and
how
much
do
you
look
people
again
judgment
on
that?