►
From YouTube: Tutorial: What Makes a Good Internet-Draft?
Description
This tutorial provides some guidelines for how to create a good Internet-Draft. Presented by Adrian Farrel, Independent Stream Editor, and Spencer Dawkins, Transport Area Director.
B
C
C
So
why
us
who
are
we
Spencer
is:
is
the
Wonder
hamster
and
I'm
the
old
dog
for
legacy
reasons
of
our
own
Internet
domains
and
and
so
on?
Spencer
is
currently
serving
as
a
transport
area
director
and
has
done
so
for
the
last
five
years,
and
that
means
he
reviews
all
of
the
internet
drafts
that
are
produced
by
the
IETF.
So
he's
got
quite
a
wide
view
of
that
and
he's
been
around
even
longer
than
me,
which
is
why
his
beard
is
gray
and
mine.
C
Isn't
he's
done
lots
of
stuff
in
the
ITF
on
the
ice
tree,
he's
in
the
internet,
research
task
force
and
here
on
the
IAB
as
well?
Previously
yeah
yeah
I
have
just
been
appointed
as
independent
streams.
Editor,
sorry,
independent
submissions
editor
and
that
means
I'm
responsible
for
RFC's
come
outside
the
IETF
process.
I
was
six
years
on
the
IES
tree
as
the
the
routing
area
director
and
I've
chaired
working
groups
across
sort
of
half
at
the
IETF
together,
I
looked
it
up.
C
C
So
what's
the
point,
what's
the
point
of
being
here?
What's
the
point
of
posting
an
internet
draft,
maybe
you've
got
an
idea.
You
want
the
ITF
to
adopt.
Maybe
it's
something
you
want
to
get
published
as
an
RFC
to
influence
the
standards
that
govern
the
internet.
Maybe
it's
just
some
thoughts
that
you
want
to
share.
It's
it's
like
their
discussion
document.
C
Whatever
the
point
of
it
is
the
implication,
is
you
want
people
to
be
able
to
read
it
and
understand
it?
You
want
the
process,
the
publication
process
to
run
through
successfully,
and
that
may
be
through
the
is
G
or
the
IRT
F,
or
the
IAB
or
the
ISE,
and
so
you
actually
need
to
get
the
the
people
responsible
for
that
publication
process
to
agree.
Yes,
this
has
to
be
published
and,
lastly,
you
want
the
RFC
editor
team
to
actually
be
able
to
publish
it.
C
C
So
let's
talk
through
the
basic
process.
Steps
take
the
microphone
out
of
here,
so
I
can
see
the
side
which
I
can
probably
read,
and
you
probably
count
from
the
back
of
the
room,
but
oh
yeah,
the
slides
are
online.
If
you
want
to
follow
along
that
way,
so
anyone
can
post
an
ID
right.
Anyone
there's
no
membership,
no
requirements.
All
you
have
to
do
is
format
it
and
give
it
to
the
submissions
tool.
C
There's
no
checks
on
the
quality,
all
the
content
of
what's
posted.
There
is
a
retrospective
check.
If
you
post,
something
that
is
defamatory
or
illegal,
then
your
internet
draft
will
be
removed,
but
otherwise
anything
having
posted
it.
It
just
sits
there.
So
if
you
do
not
ask
people
to
comment
on
it
to
read
it,
then
nothing
happens.
C
C
Usually
the
working
group
chairs
go
and
poll
their
working
group
and
say:
do
we
want
to
work
on
this?
Are
we
happy
and
some
working
groups
have
teams
of
reviewers
to
vet
the
documents
and
obviously
the
chairs
review
it
once
it's
adopted,
the
working
group
owns
the
draft
and
what
that
really
means
is
that
they
are
responsible
for
the
content.
So
beware,
you
have
your
favourite
idea,
the
best
idea
in
the
whole
world.
You
bring
it
to
a
working
group.
C
When
you
produce
a
draft
and
submit
it,
there's
an
automatic
requirement
that
you
include
a
piece
of
boilerplate
text.
The
boilerplate
text
is
something
you
need
to
read
and
you
may
need
to
show
it
to
your
sponsoring
company's
legal
team
because
it
says
stuff
about
IPR
and
copyright.
I
am
NOT.
A
lawyer
and
I've
never
played
one
in
a
television
series,
but
I
do
know
that
this
text
says
that
you
are
making
a
personal
declaration
and
statement
to
the
IETF
about
IPR.
So
you
are
personally
responsible
for
what
you
say.
C
A
Could
I
just
add
one
of
the
definitions
of
not
Luke
loans,
so
favorably
is
that
you're,
not
you're
kind
of
saying
this
is
not
going
to
be
considered
for
standard
as
a
standards
track,
so
this
is
not
it.
This
is.
This
will
not
be
a
standard.
If
it,
you
know
you
basically
that
the
the
bus
you
got
off
of
by
doing
something
different.
C
So
the
the
the
standard
approach
in
the
software
industry
and
hopefully
is
now
properly
taught
in
computer
science
courses-
is
that
the
later
you
find
about
the
more
expensive
it
is
to
fix
so
finding
above
it,
specification
is
half
the
price
of
finding
above
at
design
and
that's
half
the
price
of
finding
it
in
code
and
finding
it
in
the
field
is
really
expensive.
So
we
want
to
do
the
same
with
our
documents.
C
We
want
to
get
the
bugs
out
early
before
they've
caused
damage
and
before
people
have
become
emotionally
attached
to
what
they've
written
it's
remarkable.
How
often
something
comes
up
with
a
late-stage
problem
with
the
document
and
the
authors
really
really
don't
want
to
change
it,
because
it's
their
baby?
Well,
if
they
introduce
their
baby
to
the
world,
much
younger,
it
can
be
molded
into
the
right
shape.
C
C
C
C
The
style
guide
talks
about
two
things:
it
talks
about
the
sections
you
have
to
have
and
what
information
has
to
go
in
each
section
and
it
talks
about
some
basic
do's
and
don'ts
about
how
to
write
the
text,
how
yeah,
how
deep
to
nest
your
bullet
points,
whether
to
write
in
slang
or
not
so
the
mandatory
stuff
for
sections
that
are
in
there
all
of
the
the
boilerplate.
That's
about.
C
Ipr
about
copyright
about
the
the
fact
that
internet
drafts
expire
every
six
months,
that's
got
to
be
present.
To
be
honest,
the
way
I
do
it
is
I,
go
and
look
at
the
most
recent
draft.
That
was
posted
and
passed
all
the
checks
and
copy
it,
but
you
can
get
a
lot
of
it
free
with
the
tools
you
have
to
have
a
security
consideration
section.
You
have
to
have
an
eye
on
a
consideration
section
and
we'll
talk
about
those
in
a
minute.
C
That
is
on
another
slide
as
well.
There
are
other
sections
that
may
be
helpful
implementation
status,
telling
people
you've
already
written
code,
and
it's
already
in
the
field
helps
people
know
how
to
take
what
you've
written
other
sections
that
people
might
say
are
needed
are
not
really
needed.
So
if
you
read
RSC
4041,
it
suggests
you
should
have
morality,
consideration
section
you
don't
have
to
do
it.
C
You
should
do
it
you'll
be
struck
by
lightning.
If
you
don't
so
security,
it's
got
to
be
there
and
it
has
to
be
meaningful.
It
doesn't
have
to
be
there
from
day
one
or
it
can
be
empty
from
day
one,
but
by
the
time
you're
asking
for
working
group
last
call
it's
got
to
be
real,
it's
got
to
be
what
you
think
is
right,
so
don't
just
write
what
you
think
you
can
get
away
with
a
security
police
will
read
it
and
will
tell
you
you're
wrong.
C
C
C
So
there's
lots
of
relevant
existing
security
documents
that
may
be
related
to
your
protocol.
You
can
go
out
and
you
can
look
at
if
you
go
to
the
MPLS
working
group
and
search
for
the
word.
Security
you'll
find
a
number
of
RFC's
that
are
relevant
for
MPLS.
So
that's
a
good
approach.
Talk
to
working
group
chairs.
They
should
know
existing
security,
work
related
to
your
area
of
work
and
then
look
at
some
period
of.
C
Rfc's
that
give
guidance
on
how
to
write
security
sections,
how
to
think
about
security
and
privacy
go
to
the
security
Directorate.
A
security
Directorate
is
a
group
of
people
who
get
security
and
they
have
volunteered
their
time
to
review
stuff.
So,
when
you've
written
your
draft
and
it's
yeah
the
basic
protocol,
it's
there,
you
can
go
to
the
Directorate
and
say
help.
Can
you
can
you
point
me
at
what
sort
of
security
things
I
should
be
doing
with
this
document,
and
you
can
always
ask
for
help
pass
for
help
on
mailing
lists.
C
C
Ayana
iona
is
Ayana.
Considerations
is
another
mandatory
section
and
you
cannot
be
clear
enough.
Okay,
the
text
in
this
you
should
treat
as
if
you're,
giving
formal
protocol
instructions
to
a
protocol
engine,
even
though
the
person
who's
consuming.
It
is
a
human
because
if
this
goes
wrong,
your
protocols
broken.
C
So
you
have
to
say
which
registries
be
precise,
which
registries
are
you
talking
about?
Use
the
name
exact
name
of
the
registry.
Don't
just
say:
there's
a
registry
for
protocol
bits
that
you
know
nail
it
down,
write
down
exactly
what
you
want
the
registry
to
say
when
the
changes
have
been
made,
don't
say:
oh
allocate
me
a
value
say
allocate
a
value
saying
this
is
its
name.
This
is
its
purpose.
Is
the
reference
basically
cut
and
paste
from
the
registry
that
exists
into
the
draft
and
then
fill
in
what
you
want?
C
C
Big
bugbear
do
not
use
explicit
values
when
you
write
an
eye
on
a
section
for
an
existing
registry.
Don't
do
it
unless
you've
done
early
allocation
or
you're
actually
creating
the
registry.
It's
not
your
job
to
put
values
in
there.
It's
Ayanna's
job
and
if
you
put
values
in
there,
you've
run
all
sorts
of
problems,
because
somebody
else
might
have
written
a
draft
also
putting
values
in
and
now
you've
got
contention.
The
whole
point
of
Ayane
is
to
sort
that
out
so
use
tags
conventionally
use,
TBD
and
identify
which
one.
C
Manageability
considerations:
it's
not
a
requirement
that
you
do
this,
but
you
know
honestly,
you
should,
because
you
want
your
protocol
to
get
used,
you
want
it
to
get
managed
properly
and
be
usable.
You
need
to
think
about
these
things
at
day.
One
an
absolute
classic
is
om.
Om
tends
to
get
left
out
of
protocols
right
the
way
through
until
they're
in
the
field,
and
then
it
has
to
be
reverse
engineered.
So
a
good
protocol
spec
starts
talking
about
manageability
early.
C
C
C
It's
a
it's
a
proof
that
the
document
describes
the
protocol
that
works.
It's
a
proof
that
there
is
readability
and
comprehension
of
your
document
and
it's
okay
to
say
at
the
end
of
this
nobody's
implemented
this
yet.
But
two
people
have
told
me
that
they
are
they're
really
planning
to
it's
just
giving
extra
information.
C
Interestingly,
this
status
is
stripped
from
an
RFC
before
publication,
because
it's
reporting
on
on
a
snapshot.
It's
saying
this
is
where
we
are
in
the
process.
The
draft
is
here.
This
is
what's
been
done
when
the
RFC
is
published,
its
archived
forever,
and
so
this
would
get
out
of
date
immediately.
So
it's
stripped
out.
C
C
So
when
there's
a
document
it
needs
to
get
proper
review
in
the
right
working
group
and
that
maybe
multiple
working
groups
increasingly
we're
finding
drafts
Span
working
groups,
it's
easy
to
get
attention.
You
send
a
mail
to
multiple
mailing
lists,
done
if
you're
doing
writing
the
document
about
a
special
subject
that
crosses
areas
get
the
reviews
and
the
examples
that
I've
given
here-
young
security
management,
they're
all
special
review
teams
out
there
to
help
get
this
right,
so
you
may
be
living
somewhere
in
the
internet
area.
Writing
a
young
model.
C
A
classic
is
the
the
must
may
should
etc,
etc,
language,
it
looks
kind
of
trivial,
but
it's
actually
quite
important
to
get
it
tight,
because
this
is.
This
is
almost
legal
language
for
the
interpretation
of
the
implementer
and
that
the
classic
is,
if
you
write
should
when
you
receive
this
packet,
you
should
do
this
you're,
also
implying
that
there
is
a
may
do
something
else.
You've
got
to
write
it
down,
otherwise
people
will
just
say:
arts
assured
I,
don't
have
to
do
it.
C
If
it's
an
experiment,
if
you've
gone
for
an
experimental
RFC,
then
you
have
to
actually
describe
the
experiment.
How
will
I
know
if
this
experiment
is
a
success?
What
are
the
parameters
if
it
is
it
in
a
closed
garden?
Is
it
in
the
Internet
at
large
and
then,
lastly,
in
my
pet
nits
and
gripes
is
backwards
and
forwards
compatibility
if
you're
extending
the
protocol
or
you're
writing
a
new
protocol
to
replace
something
talk
about
how
that
interacts
with
what's
already
there
if
you're
writing
it,
any
protocol
talked
about
how
it
can
be
extended.
C
What
are
the
future
going
to
look
both
directions
and
you
can
explicitly
rule
us
out?
You
can
say
this
just
does
not
work
with
bgp
three,
but
if
you
don't
say
it,
people
don't
know
it.
So
there
there's
some
resources
here.
There's
that
there's
a
checklist
things
to
think
about
make
sure
you've
done.
C
C
English,
it's
really
easy
for
me
to
say
right,
good
English,
it's
not
too
difficult
for
Spencer
as
an
American
to
say
right,
good
English,
but
it's
I
I
understand
that
getting
English
right
is
pretty
hard.
If,
if
it's
not
your
native
language
and
frankly,
don't
ask
me
to
write
an
ID
in
Swedish
or
Chinese,
because
I
can't
so
you
know
I
know
I'm
lucky
here,
but
you
want
people
to
be
able
to
read
your
draft.
You
want
it
to
be
in
clear
English
and
you
want
it
to
be
comprehensible.
C
C
You,
then,
is
that
what
you
meant,
but
you
might
not
understand
what
they're
saying
so,
there's
a
there's,
a
real
danger
that
you
introduce
technical
errors
right
at
this
very
last
stage
of
edit,
and
some
of
that
is
because,
when
you're
copy
editor
says
this
word
should
be.
That
word.
You
say:
oh
I
didn't
understand
that
because
I'm,
not
a
native
speaker,
so
let's
change
it
and
then
you
actually
change
the
meaning
so
use
resources.
C
C
You
can
talk
to
other
people
in
the
working
group
and
there
are
actually
usually
in
working
groups,
quite
a
lot
of
people
who
are
keen
to
get
involved.
They
want
to
help.
So
if
you
say
help
I've
got
this
draft
I'd
like
somebody
to
review
it
for
me
for
English,
then
you'll
get
answers
and
then
look
at
the
RFC
editor
pages
for
various
working
group
workshop,
outputs
and
advice
on
writing
ideas
and
there's
some
links
there
and
lastly,
there's
an
online
spell
checker
for
internet
drafts.
It's
not
got
a
perfect
dictionary,
but
it's
pretty
good.
C
It's
good
I
do
spell
and
I
think
it's
another
of
Henricks
wonderful
tools.
Excellent,
so
I've
got
just
a
few
more
slides
before
handing
over
to
Spencer
diagrams.
Okay,
it's
your
writing
text,
but
do
diagrams
diagrams
really
help
make
them
simple
diagrams.
It's
nice
and
easy
to
understand.
A
clear
diagram,
modern
art
is,
is
harder
to
understand.
Ask
the
art!
Ok!
Drawing
pictures
in
ascii
is
not
a
natural
thing.
Some
people
can
just
do
it.
Some
people
can't
don't
worry.
If
you
find
it
hard
get
somebody
to
help.
C
B
C
I
went
and
looked
at
a
random
picture.
I
might
know
how
this
looks
from
the
back
of
the
room
yeah.
It
looks
as
bad
as
it
looks
on
my
laptop,
so
this
is.
This
is
a
really
good
example
of
you
know
what
not
to
do.
There
are
circles.
There
are
kind
of
what
all
the
lines
made
up
of
zeros
or
o's
like
why.
Why
is
that?
That's
just
not
a
simple
line
and
then
there's
some
random
floating
characters
in
there
as
well.
C
A
C
Yeah
mine,
the
real
unicorn,
since
it's
just
a
two-dimensional
one
right
did
you
say
what
you
meant
to
say?
Okay,
not
just
is
it
clear,
but
is
it
clearly
what
you
meant
to
say
so,
a
knowledgeable
reviewer
and
a
friendly
one
will
parse
confusing
texts
sympathetically,
and
if
you
wrote
it
yourself,
you
maybe
don't
even
notice
it's
confusing,
because
it
says
what
you
mean.
You
wrote
it
so
you
know
what
it
means.
C
Try
to
think
about
it.
From
the
perspective
of
a
newcomer,
somebody
who
doesn't
know
the
topic
very
well
and
has
picked
up
your
draft
for
the
first
time.
Will
it
be
clear?
So
sometimes
issues
are
our
English.
We've
talked
about.
Sometimes
it's
writing
style
short
sentences
are
good,
but
writing.
A
complex
short
sentence
is
bad,
so
my
advice
is
normally
use
more
words,
but
keep
the
sentences
short
use,
bullet
points.
Bullet
points
really
help
and
don't
use
the
passive
voice,
or
rather
the
passive
voice
should
not
be
used.
C
We've
also
talked
about
our
sc2
and
1:9
language,
so
ask
for
reviews,
asked
for
more
reviews,
ask
for
more
reviews
and
then
run
it
through
the
blast
calls
and
get
everybody
to
look
at
it.
Probably
that
list
should
really
highlight
ask
for
reviews
from
friends.
They
don't
even
have
to
be
technically
knowledgeable
friends.
C
So
I
picked
one
from
some
who
works
at
the
same
company
as
me,
and
you
can
try
to
read
this
and
parse
it
in
your
own
time,
but
basically
about
halfway
through
this
abstract
I
when
I
first
read
it
I
felt
I'm
drowning
I
have
no
idea
what
this
is
talking
about,
it's
confusing
and
and
it's
garbled,
and
why
would
you
put
an
abstract
in
a
drug?
The
abstract
is
the
bit
that
people
read:
okay,
it's
the
hook.
Why
would
you
put
that
out
there
unless
it's
clear.
C
Yeah,
why
have
you
written
this
draft?
Okay,
do
try
to
tell
your
readers
what
it's
there
for,
don't
just
say:
oh
here's,
an
idea.
Sometimes
it
can
be
really
hard
to
write
this
down.
But
but
if
you
don't
know
it,
then
nobody
else
will
know
it
so
state
the
problem
you're
solving.
Clearly,
in
a
few
words
with
a
picture.
That's
like
key.
If
you
don't
know
what
problem
you're
solving,
then
what
are
you
doing
if
you're
not
using
and
pre-existing
mechanism
say
so
say
why
you're
not
doing
it?
C
Not
just
I've
got
something
new
and
it
works,
but
I've
got
something
new
and
it
works,
and
it
solves
this
deficiency
be
aware
that
a
problem
statement
is
not
the
same
as
a
use
case.
They're
often
sometimes,
and
use
cases
with
I
think
we
see
a
lot
at
the
moment
of
drafts
that
try
to
list
they've
got
a
wonderful
new
idea
for
a
protocol
and
then
at
least
ten
use
cases,
and
maybe
nine
of
them
are
just
throwing
stuff
at
the
wall
to
see
what
sticks
and
one
of
them
is
the
golden.
C
C
C
C
C
C
All
right,
when
you
post
your
first
version
of
a
draft,
it
does
not
need
to
be
complete,
it
should
have
all
the
boilerplate.
It
should
have
placed
markers
for
staff,
but
if
you
write
a
full
spec,
the
first
time
you
post
it,
it
is
perfect,
then
probably
nobody's
going
to
review
it
and
discuss
it.
It's
kind
of
interesting
psychology,
but
you
want
to
hook
the
working
group
into
talking
about
your
your
document,
but
by
the
time
you
go
to
working
group
last
call,
and
certainly
by
the
time,
the
document.
C
But
by
the
time
you
get
to
working
group
last
call
the
the
document
really
needs
to
be
right.
You
must
think
it's
complete
as
independent
submissions.
Editor
I've
just
picked
up
a
document
that
which
has
got
a
publication
request.
It's
got
three
sections
in
it
that
say:
TBD;
okay,
that's
not
ready
for
publication.
A
Time
so
this
is
at
the
part
of
Adrian
I
asked
me
to
help
with
and
so
I'm
talking
about,
the
stream
specific
stuff.
I
stream
is
what
a
lot
of
us
do,
but
it's
not
everything
that
we
do
in
that
stream.
We
publish-
and
this
is
the
only
omission
that
I
know
in
the
material
that
I
have
four
kinds
of
RCS
of
standards
track.
Dcp,
experimental
and
informational
RFC's
there's
some
guidance
about
where
you
kind
of
want
to
target
that.
A
But
this
is,
you
know,
figuring
out
what
the
right
answer
is
is
still
a
minor
art.
Take
take
your
best
shot
and
you
know
the
area
director.
Even
the
is
GE
may
ask
ask
questions
about
that,
but
clay
I
say
start
out
what
looks
like
the
right
place
to
you.
We
usually
publish
dress,
produce
my
working
groups,
a
DS
can
sponsor
drafts,
but
why
not
from
working
group
and
working
groups
have
a
lot
of
flexibility,
but
publication
requested
sets
off
a
list
of
machinery
next
slide.
Please
the
machinery
looks
like
this.
A
There's
a
formal
state
machine
for
publication.
I've
got
that
there.
The
reason
you
care
is
because
state
transitions
are
going
to
generate
email,
that
you
will
see
as
an
author
or
as
a
shepherd
or
as
working
group.
Chair
and
I
have
those
listed
there.
So
the
ad
looks
at
it
and
says:
here's
what
you
know:
here's
what
I
see
the
expert
reviews
from
the
Directorate
get
set
off
at
at
last
call
time
and
get
set
off
foretell
a
chance.
A
The
last
call
that
you
get
comments
that
you
get
from
the
community
as
a
whole
become
estat.
You
get
from
other
area
directors
during
isg
evaluation,
the
state
going
to
approve
this.
Is
you
know
you?
You
may
think
that
you're
finished
take
a
deep
breath,
you're
not,
and
then
things
that
are
in
the
RFC
editor
queue
watch
for
questions
from
the
editors
resolving
comments,
and
this
this
is
I,
think
I,
hope
this
is
valuable
for
people.
A
B
A
My
next
one
area
review
teams
during
last
call
and
even
but
just
before,
a
tell
a
chat
and
from
other
IEDs
during
IHG
evaluation.
Your
best
strategy
is
to
reply
to
emails
with
comments
to
you
know
to
basically
say
yes
I.
This
is
what
I
think
you're
telling
me,
and
this
is
what
I
think
the
response
for
that
should
be.
You
don't
have
to
make
changes
to
accommodate
every
comment,
even
from
an
area
director.
So
please
do
the
right
thing
and
talk
to
your
document.
Shepherd
about
winda,
submit
updated
drafts.
A
That's
hard
to
give
general
guidance
for
one
of
the
things
is,
you
know,
should
you
know,
I
got
I
got
review
comments,
should
I
submit
an
updated
draft
and
the
area
director.
Can
you
know
your
area
director
can
look
and
say?
Well,
actually
no
one
else
has
ballade
on
it.
This
would
be
a
great
time
to
do
that,
because
none
of
the
area
directors
have
have.
You
know,
I
mean
they're
gonna.
They
need
to
see
the
new
stuff
that
you've
made
change
us
to.
A
A
That's
a
good
thing,
because
that
means
that
you're
not
getting
discusses
on
things
that
the
area
director
think
matters
but
doesn't
trust
the
authors
or
the
working
group
or
the
other
area
director
to
address
the
comment.
Okay,
we
probably
about
a
year
in
after
I,
joined
the
isg.
We
you
know
we
really
kind
of
we've
really
kind
of
changed
on
that,
and
it's
been
it's
been
that
way
for
several
years.
A
Where
you
know
we
say
we
don't
you
know
we're
not
going
to
I
know
you
can
put
it
in
a
comment
for
a
no
objection
position.
That
says
you
know
this
is
not
a
discuss,
but
it's
pretty.
You
know
it's
it's
pretty
worrying
to
me.
You
please
take
it
seriously.
So
you
make
see
comments
like
that.
Those
are
not
blocking
positions,
but
somebody
who
is
like
you
and
is
trying
to
help
is
trying
to
tell
you
something
they
may
be.
A
A
A
The
current
describes
criteria
and
non
criteria,
or,
at
this
link,
feel
free
to
take
a
look
at
them.
The
common
things
that
I
see
in
discuss,
ballots
on
you
know
the
last
three
or
four
is
geez.
I
would
say
the
number
one
that
I
do
is
I
can't
understand
what's
going
on
here.
So
please
keep
non-experts
in
mind
when
you
write
the
people
who
implement
this,
the
stuff
that
you're
doing
may
not
be
in
the
working
group
so
making
sure
that
people
who
haven't
been
involved
in
this
protocol,
since
somebody
thought
it
up
can
implement.
A
It
actually
turns
out
to
be
a
really
useful
thing
to
for
somebody
to
wonder
about
this
one
a
lot.
This
reveals
personal
identifiable
information,
so
keep
BCP.
Who,
or
does
that
138
in
mind
when
you
write
the
most
recent
surprise
that
people
have
had
it
was
the
realization,
slowly
and
painfully
at
disgust
palette.
Time
basically
saying
you
know
you
don't
think
of
IP
addresses
as
personal
identifiable
person
identifying
information,
but
a
lot
of
IP
addresses
and
ipv6
prefixes
are
only
used
by
one
person.
A
So
when
somebody
does
something
from
that
IP
address
or
even
from
that
prefix
they're
kind
of
telling
you
something
you
know,
and
you
might
you
know,
maybe
you
want.
Maybe
you
want
to
tell
people
something
you
know
in
your
protocol
about
that.
But
don't
let
that
be
a
surprise
to
you.
When
somebody
asks
about
it,
you
know
we
can.
We
can
do
a
lot.
A
A
This
can't
this
is
a
safe
on
the
open
Internet
because
of
congestion
or
because
of
security
or
because
of
scaling,
or
you
know,
whatever.
Whatever
reason
something
that
has
happened,
and
it
happened
about
a
year
in
after
I
joined,
the
is
G.
So
2014
was
that
we
started
I'm,
a
transport
guy
right
so
I'm
here
from
the
transport
area
and
I'm
here
to
help
you.
You
guys
really
have
to
laugh
at
some
of
these
statements
because
they're
intended
to
be
curious.
A
But
to
say
you
know
we
were
we
were
we
were
doing.
We
were
doing
a
lot
of
stuff,
we
think
encapsulations
on
UDP,
so
no
congestion
mechanism
there
and
it
was
not.
You
know
some
of
the
some
of
the
things
that
were
being
encapsulated.
It
had
better
congestion
mechanism,
you
know
control
mechanisms
than
others,
okay,
so
Adrian
and
some
other
folk
David
black
I
want
to
call
out
David
black,
and
there
were
some
other
folks.
A
I
spend
a
lot
of
time,
trying
figure
out
how
to
do
this,
and
what
we've
ended
up
with
is
saying
that
if
you
are
doing
something
that
is
useful
on
the
open
Internet
but
may
not
be
safe
on
the
open
Internet,
please
explain
to
us
what
the
considerations
are
to
run
it
on
the
open,
Internet
and
then
explain
to
us
which
of
those
considerations.
You
can
not
worry
about
so
much
if
you're
on
a
managed,
closed
network,
and
we
know
and
I
think
basically
all
of
the
recent
encapsulation
drafts
have
have
done
something
like
that.
A
So,
like
I
say
you
know,
it's
like
tell
us
what
you
know
tell
us
the
truth.
You
know,
but
you
know
think
about
this
stuff
up
front,
and
you
know
you'll
get
a
lot
more,
no
objections,
and
yet
you
know
I
even
bet
yes
on
some
of
this
stuff.
Why
is
this
should
not
a
must
may
be?
This
is
very
much
in
the
this
is
very
much
in
the
vein
of
what
Adrienne
was
saying
about.
A
If
it's,
if
it's
a
shirt,
then
that
means
somebody
may
not
do
it,
and
if
somebody
doesn't
do
it,
what
what
was
supposed
to
happen
and
we
we,
we
asked
that
question
fairly
frequently
and
there's
a
real
range
of
answers
that
we
get
back.
One
is
that
there
are
certain
oh
there.
There
is
a
reason
that
you
would
not
do
this
awesome.
A
A
Please,
the
IRT
F
stream
I'm,
also
in
the
IRS
G,
so
I
see
these
also
they
publish
RFC's
in
their
own
stream
and
the
process
for
doing
that
is
given
in
that
link,
RTF
specific
requirements
for
content
in
these
drafts.
Please
note
that
you
know
if
there's
things
like
saying
this
is
you
know
this
is
the
this?
Is
the
output
of
a
research
group
and
research?
If
you
don't
know
much
about
the
RTF
research
groups,
don't
have
to
have
consensus
that
this
is
the
right
thing
to
do.
A
B
A
A
You
know
seven
years
and
I've
written
I
wrote
the
only
one
that
didn't
come
through
a
research
group
that
I've
seen
go
through
the
IR
s
G,
so
you
don't
have
to
go
through
a
research
group
depending
on
what
you're
doing
I
was
writing
a
draft
saying
here's
how
to
here's?
Here's
what
I
etf
people
should
know
about
the
IRT
F,
so
didn't
come
through
a
research
group
and
that
you
know
lawyers
published
that.
But
this
stuff
will
usually
come
through
a
research
group.
They
are
not
standards
track,
they
are
experimental
or
informational.
A
There
is
no
last
call
outside
the
research
group
before
publication.
This
is
very
much
in
the
you
know:
let
it
let
a
hundred
flowers,
bloom
view
of
what
research
groups
do.
The
recent
the
RS
G
provides
review
and
guidance
for
these
through
these
drafts,
and
this
explained
in
the
publication
process
of
the
link.
The
is
G
does
what
is
called
a
RFC
57:42
review
for
conflicts
with
IETF
standards
process,
but
does
not
approve
IRT
F
stream
drafts.
A
If
you
crack
up
an
RC
57:42,
it
basically
says
these
are
the
five
things
that
the
isg
can
say
back
to
the
IRT
F
Stream,
you
know,
I
are
to
have
stream
controller,
and
you
know
one
of
them
is
we
don't
have
any
objection
to
this?
One
of
them
is,
you
know,
one
of
them
is.
This
is
related
to
work
in
this
working
group,
but
but
that
shouldn't
block
publication
and
then
you
know
then
you're
like
read
down
to
number
five
or
something
like
that
which
is
basically.
A
This
is
extending
a
IETF
protocol
in
ways
that
should
go
through
the
IETF
process.
So
that's
all.
That's
literally
all
that
you're
supposed
to
get
back
from
the
IES
is
gee.
If
you're
publishing
in
the
IOT
of
string,
the
next
slide,
the
IAB
also
publishes
in
its
own
strain,
and
this
is
described
in
RFC
4845.
A
These
are
often
produced
by
entities
that
the
IAB
overseas,
like
IB
programs,
I've,
actually
I've
got
a
draft
in
the
in
the
IAB
stack
evolution
program.
That's
supposed
to
go
out
for
community
comment
like
next
week
or
something
published
by
the
RFC
editor,
which
the
IR
IB
over
has
oversight
responsibility
for
published
by
the
Ayana,
which
is
the
same
thing
workshop
reports,
that's
for
workshops
that
the
IAB
has
done.
I've
actually
got
yeah.
That's
that
that's
one
of
those
so
and
these
are
not
standards
track
either
they
are
usually
informational
and
you'll.
A
You
can
publish
in
the
experimental
drafts
there,
that's
not
common,
but
it,
but
it's
possible.
The
process
is
that
you
will
be
working
closely
with
at
least
one
IB
member.
The
whole
IB
will
review
your
work.
The
IB
will
issue
a
ITF
wide
call
for
comments,
but
this
is
not
an
ITF
last
call.
They
are
not
bound
to
seek
consensus
or
resolve
comments.
The
IEP,
you
know
this
is
one
of
the
beauties
of
the
IEP.
Is
the
IEP
can
do
the
right
thing,
even
if
they're
the
only
people
in
the
room
that
think
it's?
A
That
is
the
right
thing.
That's
why
I
think
that's
why
they
don't
publish
on
the
standards
track,
the
IAB,
the
iesg.
You
can
say
those
guys
are
right
and
they
can
write
a
BCP.
That
says
those
guys
are
right
and
you
should
read
the
and
we're
going
to
act
like
those
guys
are
right,
but
you
know
so
that's
the
linkage
they're.
Basically,
the
iesg
does
not
perform
RFC
57:42
conflict
review
for
the
IEP
string
in
the
independent
stream.
C
Yeah,
so
the
independent
streams
publishes
RFC's,
but
they
are
not
IETF
RFC
s.
Okay,
be
very
clear.
You
can
come
to
the
independent
stream
and
get
an
RFC
and
you
can
pretend
to
be
an
intern.
I
atf
RFC,
but
it's
not.
The
process
is
in
48
46
and
what's
going
on
we're
publishing
documents
that
are
rooted
in
the
IETF
standards
process,
so
we're
following
that
paradigm,
but
we're
only
producing
information
or
experimental
RFC.
Is
it's
not
intended
to
publish
Lots,
and
my
first
pushback
is
always
why
can't
this
be
an
IETF
RFC?
C
But
if
there's
some
Internet
related
technologies
that
are
not
on
the
IETF
agenda
and
not
going
to
be
on
the
ITF
agenda,
if
there's
a
discussion
of
options
or
experiences,
if
there's
a
vendor
specific
protocol,
that
needs
to
be
documented
so
that
other
people
can
produce
implementations
that
interoperate.
If
there's
a
discussion
of
IETF
processes
or
protocols,
then
those
are
all
in
scope,
and
this
last
one
is
quite
important
and
I
think
it's
why
the
independent
stream
exists.
C
C
The
editor
sent
stuff
out
to
volunteers
for
review
and
has
an
editorial
board
to
help.
Those
reviews
are
not
mandatory,
they're,
not
part
of
the
consensus
process
but
they're.
They
form
a
discussion,
and
you
might
look
at
this
in
the
same
way
as
a
journal
paper
would
get
reviews
and-
and
you
would
discuss
with
the
journal
editor.
C
C
C
C
Get
to
summarize,
okay
make
your
draft
clear
and
meaningful.
That's
the
most
important
thing
so
write
carefully.
You
want
it
to
move
quickly
and
smoothly
forward.
So
do
all
the
right
things
early
for
the
process
and
and
for
the
cookbook
you
personally
you're.
Writing
the
draft.
You
are
responsible
for
making
that
high
quality,
so
that
means
fix
the
problems.
C
Early
ask
for
help
early,
don't
leave
it
right
up
to
the
editorial
stage,
saying
yeah,
they'll
sort
it
out
do
not
rely
on
the
iesg
in
the
RFC
editor
to
catch
and
fix
your
problems,
because
that
way
is
a
good
way
to
meet.
Those
problems
persist
to
introduce
new
problems,
to
delay
your
work
and
make
everybody
unhappy
with
you.
D
C
D
A
Anybody
want
to
question
that,
but
what
we've
been
doing
for
recent
I
well,
recent,
probably
seven
or
eight
years,
is
we've
started
doing
technical
presentations,
these
tutorials
on
Sunday
afternoon,
including
technical
material
they're,
like
you
know,
here's
here's
this
here's
the
deal
about
yang,
so
we've
had
you
know
so.
We've
had
one
of
those
that
right
now
that
is
probably
the
most
common
way
that
we
do
something
like
training
for
our
material
number.
One
number
two:
it
is
fine
for
people
to
write
users
guides,
and
things
like
that.
A
So
you
know
if
it's
useful
to
produce
that
in
a
working
group.
That
would
be
fine
if
it's
not
useful
to
produce
that
in
a
working
group,
but
you
understand
how
something
works
and
I'm
not
talking
you,
the
person
who's,
trying
to
understand
it.
I'm
talking
about
you,
the
person
who
already
understands
it.
You
know
you
know
this
is
this
is
I'm
writing
a
check
for
you
here.
This
is
a
fine
thing
for
the
independent
stream
editor
to
publish
and
say
this.
A
Is
this
person's
best
understanding
of
how
this
works
and
what
yo
the
ways
you
should
use
it,
and
what
the
benefits
are
things
like
that?
That
should
be
fine,
and
you
know
that
would
be.
That
would
be
a
pretty.
There
would
be
a
pretty
easy,
57:42
conflict
review,
which
is
you
know
we
would
have
no
objection
and
we
would
actually
be
thrilled
if
somebody
made
the
protocols
easier
to
implement
and
easier
to
deploy
and
easier
to
operate.
I
think.
C
C
There
they're
still
dry
SP
documents
but
they're
helping
to
explain
the
use
case.
The
other
thing
is
there
is
nothing
to
prevent
an
informational
pointer
reference
to
educational
material,
about
a
protocol
or
a
use
case
that
lives
outside
the
ATF.
The
only
thing
I
would
say
about
that
is
try
to
make
it
a
stable,
persistent
URL,
so
that,
when
the
things
in
RFC
in
in
five
years
time,
people
will
still
be
able
to
look
at
that
informational
material.
C
A
D
D
C
You
can
let
you
can
simply
let
it
expire
and
forget
about
it
every
six
months,
an
internet
draft
is
technically
dead.
My
advice,
however,
is
to
repost
the
draft
replacing
the
text
with
a
statement
we've
given
up
on
this.
For
these
reasons,
so
that
the
last
thing
that
was
ever
posted
as
a
draft
with
that
name
is
a
tombstone.
A
There's
actually
a
dead
state
in
the
data
tracker
for
a
draft
that
is,
you
know
we
almost
made
it
over
the
wall,
but
it
was
not.
It
was
not
ready
in
time
or
it
was
it
not
as
an
idea
as
we
thought
it
was,
and
so,
if
you
tell,
if
you
take,
you
know
to
the
working
group
chair
tells
the
area
director
that
the
air
director
can
mark
it.
That
way,
and-
and
you
know
everybody
everyone
will
know
and
I
think
that's
one
of
your
part
of
your
question
is
how
would
anybody
know.
E
C
A
A
You
know
in
my
I
got
that
as
the
area
director
and
my
guidance
back
was
everybody
stops
reading
when
they
get
to
the
I
Ana
considerations
they
think
they're
through.
If
you
know,
if
you
want
anybody
to
read
this
I
would
move
it
above
about
that.
There's
there's
no
rule
about
that.
But
don't
you
know
that's
kind
of
the
way
the
community
thinks
about
things
and
and
you're
kind
of
fitting
into
the
way
the
community
thinks
about
things
so
that
people
will
understand.
You
know
well
know
what
they're
looking
at.