►
From YouTube: IETF100-ThursdayLunchSpeakerSeries-20171116-1230
Description
THURSDAYLUNCHSPEAKERSERIES meeting session at IETF100
2017/11/16 1230
https://datatracker.ietf.org/meeting/100/proceedings/
A
Have
to
say,
maybe
you
don't
doesn't
matter
it
doesn't
matter,
but
let's
just
have
a
conversation
about
this.
A
brief
narcissistic
moment.
Many
of
you
recognize
me
from
working
here
at
the
ITF,
where
I
was
routing
ad
and
did
a
bunch
of
chairs
and
drafts
and
stuff
like
that,
but
I
haven't
it
around
the
last
couple
years
and
the
last
couple
years
I.
A
So,
just
like
at
the
ITF
open
source
and
cisco
as
a
company,
we
can't
go
any
faster
than
our
customers
and
our
customers
were
trying
to
understand
what
Sdn
controllers
and,
if
the
orchestration
etc
etcetera
data,
big
data
and
what
it
means
for
networking
they're
trying
to
understand
what
it
means
for
them
and
if
we
as
an
industry,
can't
go
faster
than
our
customers
or
the
operators
can
absorb
that
technology.
We
need
to
do
we
needed
to
take
a
turn
or
I
thought.
A
A
We've
seen
that
over
the
last
eight
years,
and
so
what
I
wanted
to
do
now
going
on
about
seven
six
years
ago,
was
us
or
reduce
the
fracturing
in
the
industry
and
at
that
time,
around
Sdn
and
NFV,
because
the
industry
was
just
exploding
with
controllers
and
there
was
a
predominance
of
one
particular
protocol.
No
subjective
statements,
but
that
protocol,
but
nonetheless
that
protocol
was
not
standardized
and
therefore
it
was
evolving
at
a
different
pace
for
different
purposes
and
there's
wide
variety.
A
Interest
in
maybe
doing
the
stuff
for
my
field,
I'll
do
the
talking
here
so
back
in
91
I
made
a
number
of
outrageous
claims,
prove
only
through
emphatic
assertion
and
let's
see
how
some
of
those
came
out,
what
I
want
to
mention
just
upfront
what
in
sum,
what
happened?
There's
still
a
massive
potential
for
the
ITF
to
engage
with
open
source
communities.
Really
it
hasn't
happened
to
the
extent
that
the
industry
needs
it
to
happen
and
I'll
discuss
some
of
those
issues.
Many
open
source
communities
have
formed.
There
are
now
a
plethora
of
different
communities.
A
Many
of
them
are
home
debt,
the
Linux
Foundation
and
there's
a
vibrant,
vibrant,
very
large
community,
around
developing
networking
technologies,
which
is
awesome
that
went
earlier
quickly.
Needless
to
say,
different
types
of
Stax
controllers-
virtual
flew
bars,
it's
it's
fractured
out
there
and
there
and
there's
no
clear,
skated
trajectory
necessarily
and
certainly
no
standardized
trajectory
that's
coming
there,
and
then
you
can
also
read
the
rest
where
that
I
have
on
the
slide.
A
Many
of
those
open
source
communities
are
also
now
standardizing
or
believe
that
they're
standardizing
packet
formats
etc
within
their
community
and
not
necessarily
bringing
those
back
to
the
ITF
we
run
for
another
standard
body,
etc.
So
the
last
thing
I
want
to
call
up
on
the
slide,
and
it
was
from
a
comment
made
at
the
plenary
last
night
and
I'm
likely
to
think
about
this.
We
many
of
us
know
who
are
35
your
friends
and
working
here
at
the
IETF.
A
We
built
our
careers
here
we
were
able
to
advance
our
careers
by
working
deeply
at
the
ietf.
Needless
to
say,
things
have
changed.
Many
folks
are
advancing
their
careers
and
deeply
engaging
in
their
career
path
by
engaging
in
open
source,
and
maybe
some
of
the
shifts
that
we're
seeing
of
where
people
are
going
R
because
the
advancement
of
their
careers
and
where
they
need
to
go
and
have
on
their
resume
that
they're,
interacting
with
open-source
communities
and
building
open
technologies,
and
many
of
that
is
there
and
so
something
to
contemplate
at
least
so.
A
Let's
just
talk
about
some
of
the
claims
I
made
in
91
will
find.
We
will
see
now,
three
years
later,
they
weren't.
In
fact
our
greatest,
so
the
ITF
can
take
a
leadership
role.
Couple
things
that
have
emerged
is
that
the
api's
platforms
and
frameworks
will
be
the
future
standards
front
for
software
driven
networking.
Second,
the
same
standardization
application
for
those
higher-level
concepts
we
need
to
think
through
if
and
which
standards
body
may
take
on
some
of
this
pieces.
A
Can
we
make
the
IETF
more
agile
and
perhaps
changes
in
the
liaison
process
and
how
the
ITF
was
going
to
work
with
other
ST,
O's
or
or
open-source
was
critical
called
to
create
more
code
in
ideas,
claims
around
rough
consensus
versus
parliamentary
procedure
as
many
open-source
communities
derive
consensus
in
a
different
way,
and
we'll
talk
about
that
and
you
know
what
running
code
now.
So
what
is
still
true
is
that
standard
geeks
are
don't
necessarily
make
the
best
open-source
Peaks,
and
it's
just
you
know.
It's
just
a
claim
for
our
argument
here
and
really.
A
A
So
next,
when
thinking
through
ninety
one
and
the
culture
and
crazy
embrace
good
open-source,
there
are
foundations
that
have
good,
neutral
third-party
management
and
how
to
be
able
to
interact
with
those
foundations
was
a
request.
But
what
I
want
to
call
it
to
the
orange
box
on
here,
and
much
of
that
is
taken
out
of
what
it
means
to
be
an
RFC
with
a
reference
to
open
source
by
design.
A
Open
source
is
constantly
iterative
and
moving,
and
so
therefore,
as
we
think
about
trying
to
base
a
reference
or
base
standardize
technology
on
a
fundamentally
a
code
base
in
technology.
That's
moving.
These
are
not
necessarily
diametrically
opposed,
but
they're
different,
so
just
quickly
quick,
taxonomy,
again
I'm,
expecting
people
in
this
presentation
to
be
listening
and
or
reading,
and
doing
a
combination
of
both
there's
a
difference
between
components
or
modules
platforms
and
opens
reference
platforms
and
there's
some
examples
that
embedded
in
there
of
what
I
mean
a
virtual
switch
is
a
component.
A
So
so
what's
happened
is
the
Linux
Foundation,
which
is
taking
a
large
role
in
creating
that
that
developer
community.
Our
networking
recently
last
May
released
a
document
on
how
to
harmonize
stos
and
open
source,
and
the
driver
is
already
the
drivers.
A
Obviously,
where
there's
a
big
rise
in
open
source
and
operators
are
using
it
and
they're,
defining
their
use
cases
and
solving
some
of
the
problems
there
and
some
of
the
recommendations
again
are
fairly
high-level
and
throughout
this
talk,
we'll
get
down
a
little
bit
into
some
of
the
nitty-gritty
of
what
we
can
actually
do
as
the
IETF
to
actually
inner
inner
work.
So
next
thing
I
wanted
to
mention
in
the
last
three
years
open
source
foundations
in
the
history
of
the
internet.
Over
the
last
10
years,
every
foundation
seemed
to
beget
another
foundation.
A
I
mean
it
was
biblical
to
how
fast
those
foundations
did
reproduce.
Everybody
with
the
technology
or
a
project
wanted
to
create
a
new
foundation
around
themselves
and
have
a
community
and
just
to
keep
this
simple.
It
meant
that
you
had
to
invest
in
it,
and
so,
in
the
last
several
years,
the
Linux
Foundation
has
taken
a
number
of
these
different
projects
and
now
created
an
umbrella
project.
Now
the
point
here
isn't
just
the
overall
governance
goals.
A
It's
that
even
the
organizational
structure
of
communities
and
how
they
interplay
and
interact
is
fundamentally
more
dynamic
than
we
may
be
seeing
in
standards
bodies.
And
so
let's
talk
about
specifically
what's
happened
since
91
in
the
idea
draft
was
recently
put
out
on
how
to
normatively
reference
an
open-source
community
from
that
draft.
I've
worked,
I've
talked
to
those
authors
at
least,
and
there's
a
number
of
pitfalls.
To
avoid
one.
Is
that
you're
chasing
behind
the
code
trying
to
document
what
it
actually
did?
A
Second,
the
sari
they
claim
from
the
ITF
that
there
must
be
multiple
operable
interoperable
instances,
isn't
necessarily
the
goal
of
that
community
they're
working
quickly,
iteratively
on
one
particular
architecture.
For
this
and
in
general,
to
make
my
point.
If
there's
major
architectural
flaws
or
if
there's
issues,
then
another
community
will
form
to
flood
those
those
flaws,
and
so
there's
there's
fundamentally
different
goals
of
interoperation,
because
one
is
a
subway,
a
problem
in
a
use
case
that
we
have
and
the
goals
potentially
of
that
normative
reference
may
not
be
met.
A
So
a
couple
things
that
I've
talked
about
with
the
author
here,
all
the
way
again
I
think
this
is
critical,
work
it's
necessary,
but
not
sufficient
to
be
able
to
have
cross
SDO
and
open
source
communication
as
well
as
progression
together.
How
do
we
pick
the
projects
that
are
within
the
on
the
the
list
of
open
source
projects
that
can
be
chosen
for
a
normative
reference
working
group
gonna?
Do
this
we're
gonna?
Do
this
by
consensus?
Is
a
draft
off
we're
gonna
pick
a
winner
interesting?
A
A
That's
my
point
and
then
there's
other
other
nuances
of
working
at
platforms
and
api's
and
in
particular,
that
many
projects
or
communities
have
sub
projects
within
their
architectures
and
if
you
only
want
to
refer
to
or
have
a
reference
normative
reference
to,
one
particular
module
or
one
particular
piece
that
needs
to
be
worked
through.
I
think
it's
critical
to
be
able
to
normally
leave
reference,
open
source
and
code
etc.
But
there
are
some
more
work
needs
to
be
done
here
to
make
this
work.
Second
and
we're
gonna
go
into
a
couple
examples
right
now.
A
The
aim
catalog
work,
which
many
of
you
know
something
about-
is
an
experimentation
from
the
IETF
side
to
create
a
live
set
of
models
for
networking
devices,
keeping
it
simple
that
actually
is
not
trying
to
produce
documents.
It's
trying
to
produce
live
models
that
they
can
work
together,
and
this
becomes
critical
as
an
experiment,
because,
frankly,
we
can
show
it
as
success
and
the
reason
why
I
was
a
success.
A
Frankly,
as
you
can
see
listed
in
numerically
one
through
six,
a
number
of
tools
were
built
around
this:
for
education,
for
referencing,
for
compilation
for
dependency.
Checking
this
is
what
about
searching,
etc.
The
application
of
metadata
that
we'll
talk
about
to
be
able
to,
describe
and
and
put
measure
metrics
and
variables
around
the
different
models
that
utility
of
tooling-
and
many
of
you
have
worked
on
on
the
hackathon-
became
critical,
so
that
metadata
from
a
code
point
of
view
becomes
one
definition
of
a
health
metric
as
depending
on
the
data
that's
put
in.
A
But
what
I
personally
find
most
interesting,
because
I
think
this
is
the
path
forward
for
sto
relationships
in
multi
SEO
relationships
as
well
as
to
open-source
is.
There
is
a
dependency
diagram
instead
of
having
the
ITF
state
in
the
industry.
A
yang
models
done
here.
There
is
a
relationship
that
the
broadband
form,
the
MAF
and
a
wide
variety
of
other
stos,
as
well
as
open
source
projects
open
configures.
A
The
big
conversation
as
well
are
producing
these
models
for
their
specific
technologies
and
areas
of
the
industry
that
they're
trying
to
work
on,
and
this
dependency
graph
actually
pulls
them
together
into
one
unified
place
and
therefore,
with
the
metadata,
the
compiler
can
show
that
the
entire
system
can
work
so
that,
but
this
wouldn't
be
a
recurring
theme
from
him.
Couple
of
other
pieces
that
came
out
was
looking
at
the
communities
at
the
Linux
Foundation,
the
overlap
between
IETF
technology
and
open-source.
A
So
this
is
taking
that
dependency
graph
up
one
level
and
what
becomes
interesting
about
this
in
conversations
with,
hopefully
within
the
IETF,
we're
catalyzing
conversations
towards
this
end.
This
shows
where
IETF
technology
is
being
used
and
therefore
influence
is
already
occurring
and
can
occur
to
greater
degree,
and
it
shows
where
there's
an
on
overlap
and
the
ITF
can
say
we
better
pay
attention
here.
A
We
better
go
see
what's
going
on,
because
we
have
a
long
history
here
at
the
ITF
of
understanding
how
to
do
things
securely,
etc,
and
perhaps
some
of
these
communities
can
take
advantage
of
that
as
well
as
perhaps
we
can
make
sure
that
the
implementation
is
done
well
and
not
with
other
causing
other
issues.
Couple
lessons
from
this
immediately
once
I
said
that
the
data
models
were
alive.
The
goal
of
this
effort,
even
though
sponsored
by
the
ITF,
was
not
the
RFC
ID.
A
It
was
being
able
to
point
to
a
live
tree
of
dependent
models
that,
together
being
worked
on
independent.
Oh,
we
worked
on
independently,
create
the
new
way
to
operate
a
software
software
driven
network
or
a
model
driven
network,
and
so
that
that's
been
a
fascinating
thing
to
learn
about
this,
and
you
can
see
that
over
300
modules
are
are
now
using
this.
The
next
one
was
when
we
went
through
the
education
program
or
when
the
leaders
of
this
project
did.
It
was
developed.
A
The
tools
for
your
customers
and
the
customers
are
not
necessarily
solely
the
ones
that
the
ITF
it's
for
these
other
SDS
and
open
source
communities
and
through
continued
tooling,
simplify
the
development
process.
So
this
is.
This
has
been
a
a
very
good
experiment,
but,
as
we've
potentially
talked
to
this
week
and
on
email
lists
about
is
a
2.0.
A
All
of
this
work
has
been
privately
funded
and
worked
on
with
this
tools
have
been
privately
funded
and
with
the
potential
process,
change
towards
live
models
referring
across
SD
O's
and
across
open-source
communities,
not
saying
the
private
funding
is
going
to
dry
up
I'm,
just
saying
for
longevity
of
the
process
and/or
of
the
tools
and
to
build
a
community
around
this,
it
needs
to
it
needs
to
move
through
a
different
level,
and
that
may
be
part
of
the
role
of
our
sock
or
the
ITF
in
Casa.
2.0.
A
We'll
give
you
another
quick
example:
it's
one
m2m
very
similar
if
you're
familiar
with
this,
it's
related
to
IOT,
just
to
keep
it
simple
and
if
you're
do
work
in
IOT,
you
know
that
it
is
one
of
the
most
fractured
industries
ever
created
by
mankind
and
here's
a
bunch
of
logos
of
the
groups
that
that
claim
to
have
the
sole
answer
for
IOT.
That
was
a
joke
by
the
way.
A
Rolls
out
like
this,
so
I've
gotten
completely
drunk
on
my
kool-aid
that
the
way
to
understand
the
impact
of
the
technology
is
not
on
the
RFC
ID,
but
on
the
product
of
the
RFC
and
the
product
of
that
RFC
is
represented
by
a
multi
sto
multi,
open
source
community
dependency
graph,
which
in
fact
is
what
the
industry
ends
up
using
to
drive
what
they
need
to
solve.
In
the
industry,
so
a
couple
of
other
things
that
we've
tried
battering
91-yard
o
was
the
ITF
chair.
At
the
time
we
decided
hey,
let's
bring.
A
Let's
just
try
something
simple:
it's
being
done
all
over
the
industry:
let's
just
bring
hackathons
to
the
IETF,
and
thankfully
we
we
got
a
chance
to
in
continued
work
with
Charles
EKKO
who
really
brought
this
together.
Here's
a
picture
of
all
of
you
hacking
just
our
last
weekend
and
it
and
I'll
show
you
in
a
second.
It
has
become
one
of
the
most
relevant
ways
to
grow,
not
only
IETF
participation
but,
of
course,
take
IETF
technology
and
create
open-source
communities
around
it.
A
A
So,
at
the
plenary
last
night
there
was
conversations
about
a
T
yep
and
what
you
see
in
blue
on
the
left
hand,
side
is
total
attendance
in
what
you
see
in
orange
is
remote
attendance
and
you
can
pretty
much
AIBO
that
there's
a
best-fit
line,
and
so
it's
declining
attendance
and,
on
the
right
hand,
side
it's
hackathon
attendance,
and
you
can
pretty
much
I,
though,
that
that's
a
steeper
line
increasing
and
in
fact,
the
claim
that
it's,
the
fastest
growing
part
of
the
ITF
community
I.
Think
it's
true.
A
What's
interesting,
though,
at
the
yellow
line
are
the
individuals
who
are
only
coming
to
the
hackathon
and
not
staying
for
the
meeting
interesting
to
consider
whether
or
not
we
can
do
court
better
outreach
to
the
coders,
the
developers
of
the
open-source
community
for
the
technology
by
simply
taking
advantage
of
the
folks
who
were
coming
to
the
beginning,
part
of
the
hackathon,
just
a
just
a
quick
bit.
This
is
showing
and
blue
and
green
between
ninety
nine
nine
hundred.
A
When
we
first
started
asking
the
question,
is
this
your
first
ITF
and
is
this
your
first
hackathon
as
to
the
same
people
and
sure
enough,
there's
a
high
correlation
to
folks
showing
up
to
the
I
to
the
hackathon,
who
aren't
necessarily
feeling
potentially
any
any
membership
in
working
groups
at
the
IETF?
And
so
there's
a
couple
of
really
big
hackathon
communities
that
are
constantly
being
used
throughout
the
history
of
these
hackathons?
A
The
platform
for
network
data
analytics
opendaylight
and
fright
Oh
for
the
for
ting
frame,
but
in
reality,
I
want
to
make
sure
that
this
isn't
a
complete
list
nor
exhaustive.
But
it's
a
really
really
long
list
of
IETF
technologies
that
are
actually
being
developed
in
communities
created
around
that
technology
here
at
the
ITF,
and
that
is
the
best
news
so
therefore
to
get
the
modules
of
technology
that
are
being
created
into
open
source
and
to
get
those
architectures
that
I
showed
you
built
out
of
this
technology.
A
This
is
again
another
path
forward
that
we
should
think
about.
So
it's
like
a
pause
for
a
second.
Why
again
do
we
do
we
care
about
this?
We
care
about
this
because
the
end
operators
and
deployers
of
networks
end
of
this
of
this
technology
are
in
fact
investing,
and
this
is
seen
by
membership
in
those
there's
open
source
communities
and
their
investments
in
those
open
source
communities
that
one
they
want
to.
Learn
it
two.
They
want
to
understand
it,
of
course,
and
three
they'll
even
want
to
be
able
to
modify
it
themselves.
A
A
Us
represents
a
lot
of
things
that
I
worked
on
my
career,
so
it
here
is
shed
because
the
list
in
the
middle
is
in
fact
the
new
modern
technology
project
have
technologies
that
they
want
to
go
to,
but
the
way
they're
consuming
those
technologies
is
via
those
open
source
projects
that
I
mentioned
in
the
bubble.
In-Between
it's
a
it's,
a
fascinating
just
transition
of
network
operation
to
the
inclusion
of
open
source
and
were
they
looking
for,
make
it
simpler.
We
want
to
be
involved,
we
want
to
help
code
ourselves.
A
We
need
to
understand
how
that
stuff
works.
So
a
couple
quotes
again
from
one
from
Centrelink.
First,
we
don't
we
want
to
do.
We
use
open
source
because
it's
fast
and
again,
we
want
to
be
involved
and,
second
from
Bell
Canada.
We
want
to
be
able
to
not
only
potentially
progress
the
scanners
but
progress
the
functionality
to
solve
our
problems,
and
we
can
do
that
in
code.
So,
let's
just
do
a
quick,
I
promise,
a
quick
talk
about
what
other
SD
O's
have
done.
Besides
IETF
in
case
you
haven't
been
fallin
knows.
A
First,
the
MAF
has
fundamentally
changed
the
trajectory
of
what
they
work
on
based
upon
their
membership
and
we'll
just
talk
about
this
for
a
second
there's
about
two
hundred
and
ten
members
and
those
are
large
enterprises
and
telcos.
First,
they
needed
to
reinvent
themselves.
Many
folks
will
remember
and
have
worked
on
this
where
there
was
an
MAF
certification
for
ether,
metro,
ethernet
equipment
in
particular,
and
they
were
on
this
cycle
and
they
weren't
able
to
move
quickly
enough
to,
above
to
the
operational
needs
of
Metro
Ethernet.
A
That
I
meant
that
I
showed
just
a
second
ago
that
had
a
notion
of
orchestration
and
then
model
driven
api's
based
on
yang
to
those
devices
and
the
services
that
come
from
that
we
also
brought
or
at
radical
brought
said.
Hey
man.
Hackathons
are
working
great
and
other
standards
bodies,
and
you
have
this
architecture.
You
want
to
build.
Why
don't
you
build
that
architecture
in
code,
as
your
us
are
trying
to
specify
this
at
the
same
time?
A
So
Charles
is
basically
a
one-man
army
out
there
trying
to
get
standard
bodies
to
realize
the
value
of
producing
the
code.
At
the
same
time.
Second-
and
this
is
also
fascinating
for
the
is
a
2.0
conversation-
the
MAF
invested
in
infrastructure,
compute
storage
and
networking
to
build,
deploy
and
certify
their
new
architecture.
On
to
play
to
discuss
again
about
the
future,
the
IDF
idea,
the
MAF
different
than
the
IDF,
is
a
very
bounded
architecture
and
therefore
a
bounded
target.
A
And
once
again
you
take
a
look
at
the
tools
that
are
most
popular
in
that
standards
body.
Perhaps
a
sharp,
perhaps
not
a
fact,
very
similar
to
the
tools
that
represent
IETF
standards
and
standardized
technology
as
well.
But
there
is
always
the
case
that
you
know-
and
it's
definitely
been
the
case-
that
there's
still
a
fight
out
there
for
sto,
relevancy
or
standards
body
relevancy
and
in
some
cases,
a
defensive
territory.
And
so
there
are
a
ton
of
s,
views
and
they're
cropping
up
all
across
the
industry,
related
to
different
vertical
industries
or
different
universe.
A
Parts
of
the
infrastructure
like
IOT,
etc
and
don't
be
confused
by
people
staking
a
flag
and
saying
this
is
our
turf
and
work
is
going
to
happen.
That's
obviously
not
the
case
and
also
don't
try
and
mop
up
behind
on
the
stay
to
clearly
mop
up
behind
the
coders
and
write
specifications
over
what
was
done
in
the
code.
These
oftentimes
that
code
is
is
experimental.
A
So
when
choosing
the
partnership
with
the
open
source
community,
a
few
things
to
take
a
look
out
there,
there
are
a
ton,
if
not
more
open
source
communities.
You
can
see
this
in
the
rise
of
the
use
of
github
et
Cetera's
mentioned
earlier
open
source
similar
to
standards.
Bodies,
of
course,
is
driven
by
potentially
large
organizations
trying
to
create
strategic
market
and
technology
development
and
set
a
trajectory
happens
and
standards.
It
happens
in
open
source-
that's
obvious,
but
let's
just
call
it
out.
A
A
So
in
a
little
bit
of
a
provocative
statement,
I've
put
up
here
where
green
is
good
and
weird
pink
is
bad
of
different
standards,
organizations
that
have
tried
to
work
with
the
open
source
model
and,
as
you
read
some
of
the
details
you
can
see
potentially
or
at
least
what
I'm
calling
out
to
be
the
success
or
failure
criteria
of
those
different
organizations,
and
it's
not
that
CableLabs
is
doing
swimmingly.
It's
just
potentially
comparatively
to
the
others
they're
finding
some
success
because
cable
wraps
is
not
doing
swimmingly
is
what
are,
though,
so.
A
Let's
just
talk
about
a
couple
of
things.
So
it's
frequently
been
acclaim.
They
look
other
other
bodies
are
doing
certification.
If
we
have
an
IETF
certification
associated
with
our
technology
will
expand
and
will
show
our
influence
over.
This
there's
been
failure
situations
as
well
as
well
as
successes.
Let's
talk
about
those
one
again,
poor
specification
in
this
case
in
the
java
community,
the
the
specification
and
reputation
for
that
specification
became
so
bad
that
certification
became
meaningless.
A
Second,
the
etsy
NFV
architecture
and
OPN
fve.
What's
interesting,
is
for
peein
of
these
trying
to
run
a
certification
play
and
then
states
that
they
have
this
alliance
with
etsy,
but
doesn't
test
against
that
etsy
architecture,
kind
of
interesting.
How
that
all
works
out
so
again
be
careful.
The
one
that's
primarily
claimed
to
be
a
great
success
is
3gpp.
Where
that
certification
allowed
multiple
interoperable
products
to
be
built,
we
will
see
if
this
works
for
5g
whatever
5g.
Is
it
apparent?
It's
apparently
everything,
but
nonetheless
telephone
now.
A
A
Those
communities
are
working
on
directly
on
use
cases
to
solving
problems,
and
so
in
the
ietf
process,
which
is
give
me
a
framework
draft.
Give
me
a
bunch
of
use
case
drafts,
give
me
an
architecture
draft
and
when
you
get
all
that
done
well,
you
can
go
work
on
the
actual
technology.
This
is
a
great
place
to
go
mining
for
those
concrete's
use
cases
and
communities
that
are
working
towards
them.
So
the
last
one
I
want
to
mention
back
on
m2m
is
the
open
connectivity
foundation
which
very
new
organization,
but
here's
my
point.
A
They
have
pulled
out
all
the
stops:
certification,
testing,
open
source
standards,
device
models,
etc.
So
new
membership
funding
model
they
they're
now
bringing
together
all
the
techniques
that
have
been
claimed
to
be
successful
into
one
group,
but
just
note
that
there
are
bounded
on
a
specific
problem
and
it's
yet
to
be
determined
whether
or
not
an
industry
can
support
a
foundation
on
potentially
such
a
limited
case.
A
I
want
to
then
put
this
into
some
kind
of
map,
for
reference
going
in
the
upwards
direction
is
commercialization
versus
individual
representation,
and
going
from
left
to
right
is
the
notion
of
free,
complete,
open
use
to
a
closed
environment,
taking
a
look
at
50,
some
of
that
open
source
work.
This
is
just
one
person's
way
of
putting
things
into
a
diagram
and
onto
a
map
of
the
commercialization
and
the
influence,
our
specific
enterprises
and
vendors
towards
the
contributions
and
directions
of
these
groups.
A
You
can
see
where
I
Jeff
was
placed,
and
this
particular
case,
and
now,
when
looking
at
the
partnership
goals,
are
the
partnership.
Sorry
are
the
goals
aligned
between
a
standards,
body
and
open
source
community
and
the
map
changes
a
little
bit
and
a
few
more
things
are
at
it,
and
so
this
will
end
up
being
one
of
the
cultural,
as
always
the
cultural
variable,
and
whether
or
not
goals
are
aligned
will
end
up
being
one
of
the
primary
ways
to
to
see
if
the
SCO
and
open-source
community
have
working
together.
A
But
I
want
to
caution
because,
just
like
there's
been
an
aggregation
of
open
source
communities
allowed
Linux
Foundation
networking
project,
you
know
I've
got
a
call
out,
especially
boot,
especially
you
know
cake
having
Andy
right.
There
call
out
the
evolution
from
the
frame
relay
an
ATM
forum,
all
the
way
to
the
broadband
forum
and
the
abrogation
of
stos.
That
also
occurs
over
time
as
well,
and
that
easily
can
happen,
and
it's
important
for
that
to
happen
as
potentially
smaller
ideas
accumulate
into
larger
ones
as
well.
A
So
with
the
reference
back
to
the
social
there's
potential
for
some
s,
Tio's
or
some
technologies,
as
well
as
open
source
communities
to
be
very
small
fish
in
a
very
large
aquarium
depending
upon
how
quickly
the
relationship
in
inter
dependency
between
open
source
and
standards
bodies
actually
occurs,
and
it
it
kind
of
depends
and
I,
don't
want
to
say
that
things
become
irrelevant.
If
you
don't
adopt
open
source,
you
just
in
to
make
my
argument.
Things
become
nice
and
very,
very
specific,
and
that's
still.
A
Okay,
but
again
it's
the
scope
of
influence
in
the
industry
and
trajectory
is
different
in
this
case.
So,
just
the
here's,
the
partner,
here's
the
narrative
associated
with
this
section
really
is
that
success
is
when
there's
going
to
be
friendly,
common
goals,
sorry
friendly
communities,
open
cultures,
common
goals
and
IP
alignment.
We'll
talk
about
north
for
property
in
a
second
because
the
to
failure
pieces
one
is
around
intellectual
property
and
the
inaugural
property,
as
as
defined
and
discussed
here
at
the
ITF.
A
A
Bodies
is
fundamentally
different
than
that
and,
as
you
can
see
see
in
some
of
these
other
pieces,
this
is
a
real
conversation
to
have,
and
it
it's
at
one
scene
of
the
movie
not
at
the
first
scene,
but
at
a
couple
scenes
into
the
movie,
where
part
of
the
way
that
we're
thinking
about
the
preservation
of
intellectual
property
rights,
aligned
across
standards
and
open
source.
It
always
comes
down
to
lawyers
I'm,
not
a
lawyer,
but
nonetheless,
these
issues
are
real
and
we've
got
the
force
work
through
them.
A
So
what
I
wanted
to
move
to
the
next
section
and
and
kind
of
get
to
get
to
some
recommendations?
Maybe
I'll
talk
about
in
three
years
is
standards
just
creating
documents
above
into
someone,
who's,
building
functional
code
trying
to
solve
problems.
But,
in
my
opinion,
the
industry
leader.
The
future
is
one
that
is
understanding
that
there
are
multiple
communities
that
need
to
be
influenced.
A
So
there's
some
choices
for
the
IETF
along
these
lines
from
do
absolutely
nothing
and
potentially
be
focused
on
very
specific
technology
niches
work
in
the
middle
and
have
some
communication
going
back
and
forth
and
and
continue
the
open-door
policy
of
the
IETF,
which
is
hey.
If
you
want
to
standardize
something
come
to
us,
realize
that
one
of
the
most
important
rules
in
an
open-source
community
is
in
fact,
community
outreach,
which
is
hey
we're
working
on
this.
Do
you
want
to
work
on
this
together,
I'm
summarizing?
A
What
the
role
is,
but
that's
effectively
what
the
description
is
and
then
for
vigorous
relevance.
Consider
changing
the
process
again.
Live
repositories
like
in
the
yang
catalog,
as
I
mentioned
earlier,
and
work
realizing
that
it's
a
multi,
SEO
and
multi
Karpin
source
world.
So
really
the
meta
point
of
that
is
can't
look
at
the
world
through
the
keyhole
of
one
particular
piece
of
technology
we
have
to.
A
A
But
that
does
appear
to
me
to
be
a
fantastic
way
to
understand
that
relationship
and
that
tooling
is
really
setting
the
industry
trajectory
for
how
many
standards,
bodies,
communities
and
organizations
will
work
together
and
understand
the
relationship
so
taking
a
look
at
some
open
source
project.
Genesis
I
want
to
harken
back
to
three
years
that
in
the
early
days
of
Sdn
and
NFV,
there
were
many
boss.
There's
many
conversations.
A
I
even
gave
a
couple
of
talks
on
this
that
there
are
things
we'd
like
this
to
to
standardize
here
and
need
to
standardize
here
whether
it's
protocols,
state
machines,
events
packet
formats
and
then
all
the
way
up
to
a
PRS
and
platforms
potentially
and
this
organization
and
frankly,
no
other
standards
or
organization
really
took
up
that
charge.
That's
okay!
A
That's
the
way
the
world
unfolded,
but
that
did
give
rise
to
these
major
open
source
projects
that
end
up
creating
that
de
facto
standard
for
those
pieces
that
these
are
the
events
are
going
to
get
and
how
we're
gonna
build
our
state
machines.
So
the
cloud
here
really
is
you
do
need
to
seek
out
projects
them.
A
A
The
the
point
here
is
is
the
the
middle
sub-bullet,
which
is
that
that
creation
of
the
platformer
platforms
of
these
dependencies,
the
dependency
maps,
also
allow
you
to
understand
where
the
open-source
communities
are
going
in
into
it,
also
which
technologies
could
help
a
particular
open-source
community,
etc,
and
so,
as
has
been
done,
we
don't
necessarily
at
the
ITF
at
the
notion
of
relevancy
metrics
health
metrics
associated
with
a
particular
standard.
We've
talked
about
what
makes
a
successful
standard,
without
necessarily
applying
those
metrics
across
the
board
to
us
standards.
A
Efforts
here,
but
I
want
to
remind
you
again,
there's
no
academic
or
theoretical
basis
of
what
is
the
definition
or
what
are
the
variables
of
a
healthy
open-source
technology
and
a
healthy
open-source
community,
and
if
there's
gonna
be
a
relationship
between
standards
bodies
and
open
source.
Those
two
things
need
to
be
understood
potentially
independently,
so
as
as
I'm.
A
How
to
institutionalize
hackathons
is
really
just
a
starting
point,
but
it
allows
the
conversation
which
can
also
occur
in
parallel,
that
the
RFC
ID
should
not
be
the
only
metric
for
the
IETF,
and
that's
really
the
claim
here
that
the
product
of
the
RFP
are
RFC's
is
the
return
on
investment
and
the
value
associated
with
the
modular
technology.
Game
are
created
here,
and
so,
as
you
can
also
see
the
notion
of
iterating
standards
and
certainly
live
revision
control
systems
containing
in
this
case,
the
yang
models
is.
A
That's
done
of
a
lot
of
the
private
funding
for
some
of
these
tools,
and
so
these
other
pieces
I'm
really
looking
for
a
way
to
partner
with
the
IETF
or
partner
with
I,
saw
and
us
or
partner,
with
open
source
communities
to
bring
this
tooling
and
create
a
community
around
this.
Not
saying
that
necessarily
the
money
is
going
to
dry
up,
but
it's
not
positive
for
the
longevity
of
this
particular
idea
towards
the
trajectory
if
the
ITF
wants
to
take
it
on
to
simply
have
it
in
private
hands.
A
That's
my
only
point
there,
but
this
is
a
way
that
the
ICAP
can
directly
work
in
open
source
because
their
processes
are
also
driven.
This
left
so
I've
come
to
the
end
of
my
slides
and
I.
Wouldn't
take
any
questions.
I
made
again
a
number
of
outrageous
and
contentious
statements
here
prove
merely
through
emphatic
assertion
and
so
by
all
means
arm
any
questions
or
comments
that
you
may
have
I'd
be
willing
to
discuss
with
you.
B
Michael
Abramson,
you
had
IPR
there
on
one
slide.
Where
do
you
think
we
are
in
three
to
five
years
when
it
comes
IPR
and
open
source
and
SD
o--'s,
because
everything
is
getting
patented
I
think
people
are
paying
for
the
patent
and
it's
just
patent
writing
machine
out
there
forever
everybody?
Where
do
you
think
we
are
in
three
to
five
years?
So.
A
I'm,
not
a
lawyer,
god
only
knows
what
lawyers
are
gonna
figure
out.
Alright
I
know
they'll
figure
out
how
to
kick
I'm
running
but
other
than
that.
I
actually
probably
see
something
towards
more
the
open
licensing
that
that
we
see
out
of
Apache,
Eclipse
or
BSD,
or
something
along
those
lines
versus
Rand
because
random
we
do
it
standards.
Just
to
summarize
my
point
is
you
know
you
can
use
this.
If
you
don't
sue
me
and
I
won't
sue
you
and
anytime
I
feel
like
taking
that
away.
A
C
Rehema,
well,
you
mentioned
the
hardware
open-source
and
obviously
it's
not
evolving
as
fast
as
open
source
software.
Obviously
vendors
hardware
and
silicon
vendors
are
not
as
interested
to
go
into
that
very
quickly.
So
I
wonder.
What's
your
perspective
about
what
we're
going
to
see
three
years
from
now?
A
From
that
hard
work
and
anytime,
you
see
an
industry
de
facto
standard
mechanism
for
actually
creating
products
that
can
be
then
created
by
no
DM
supplier,
etc
and
brought
into
the
industry.
We
better
have
a
relationship
with
those
or
otherwise
the
standards
that
we
write
and
what
they
implement
and,
what's
the
put,
are
gonna
be
dramatically
different.
So
that's
my
result.
D
A
It's
a
feature
and
it's
a
feature
because
a
lot
of
the
implementations
have
different
design
patterns,
as
well
as
different
architectural
goals.
Different
functional
goals
can
scale
and
and/or
perform
in
different
mechanisms
and
that's
the
purpose
of
having
different
design
patterns
and
architectures
in
software.
None
of
it
may
have
to
be
standard
and
in
fact,
what
it
lets
say.
The
eventing
api's
platforms
as
well
as
protocols
are
fully
standardized,
but
what's
in
between
is
dramatically
different
and
they
appear
to
be
working
towards
the
same
problem.
I
encourage
it
I
think
it's
a
feature.
Another
bug.
E
Cambron
gondwana
and
more
from
the
applications
world,
rather
than
than
the
low-level
things
and
in
our
world.
Certainly
ATF
standards
come
through
as
a
bunch
of
pros
and
possibly
some
a
B
and
F
and
implementation
quality
is
very
dependent
on
how
well
somebody
read
the
standards
in
the
IMAP
world
in
particular.
There's
a
lot
of
very
bad
implementations
out
there,
because
it's
a
long,
complex
standard
and
people
aren't
reading
it
very
well.
I
would
say
that
testing
is
actually
more
important
than
the
frozen
some
ways.
E
A
I
agree
with
you
and
and
what's
interesting,
is
that
there's
at
least
been
one
open
source
community
fti?
Oh,
that
actually
was
one
of
the
first
to
create
a
functional
performance
and
scaling
test
framework
associated
with
the
open
source
they
were
creating,
which
was
matched
up
against
the
proper
protocol.
Behavior,
so
I
agree
dude,
it's
just
it's
a
it's
not
necessarily
a
norm,
now
quantities
back
from
just
talking
about
one
community
and
what
they
do.
A
But
as
I
as
I
showed
in
a
couple
of
my
examples
of
doing
testing
and
certification
based
upon
complex
challenging
standards,
the
standards
weren't
good
enough.
The
code
wasn't
good
enough.
This
certification
didn't
matter
because
none
of
the
awesome
secrets
either
previous
to
problems,
and
so
it's
not
a
guarantee
that
even
high
quality
testing
is
going
to
evolve
properly.
Comparison.
E
A
The
MAF
actually
on
the
infrastructure
side,
the
MAF
actually
has
that
with
certification
and
presenting
some
caveats,
but
I
appreciate
that
a
scoring
mechanism,
but
that
is
beyond
necessary.
That's
an
industry
scoring
mechanism
that
it
is
necessarily
a
standards
body
of
scoring
mechanism.
In
that
case,
thank
you.
The
man
himself
Charles
buckle.
F
So,
first
of
all,
how
many
people
might
wanted
to
point
out
that
the
hackathon,
really
in
large
part,
came
out
of
some
of
the
challenges
you
raised
three
years
ago,
I'm
the
some
of
the
recommendations
of
potential
paths
we
put
out.
So
so.
Thank
you
for
that,
and
thanks
for
the
support
to
get
the
hackathon
started,
I
think
it's
making
some
good
progress
in
helping
standards
move
forward
more
quickly,
testing
things
out
before
we
get
too
far
under
the
standards
producing
running
code,
proof
of
concept
that
type
of
thing.
F
But
what
you're
talking
about
here
made
me
think
that
we
need
to
go
probably
beyond
that
and
for
things
to
really
get
deployed
to
produce
code
that
isn't
just
a
proof
of
concept,
but
is
actually
a
useable
library,
something
that's
that's
solid
and
that
people
can
build
on
top
of
some.
That
has
good
api's.
That
type
of
thing
is
that
something
that
we
should
tackle
within
the
IETF,
or
do
you
think,
there's
other?
We
should
go
outside
to
make
that
happen
and
produce
those
types
of
libraries.
So.
A
When
you're
talking
about
trying
to
create
something
that
solves
a
functional
problem
and
can
be
potentially
used
in
production
or
product,
that
is
actually
defining
the
creation
of
an
open-source
community,
whether
that
can
be
done
inside
the
IETF
and
can
be
governed
funded,
supported
as
where,
as
a
community,
a
developer
community
inside
the
ITF
or
related
to
the
ITF
can
be
created.
That's
yet
to
be
determined.
But
if
that's
a
strategic
directory
that
the
ITF
wants
to
go
on,
they
potentially
could
meet
that.
A
G
Okay,
my
name
is
rich
Souls
and
I've
been
involved
in
a
lot
of
the
similar
kind
of
junctions
on
different
scale.
Smaller
scale,
I
think
one
thing:
there's
a
challenge
in
this
stuff
is
that
the
foundations
or
the
level
of
infrastructure
support
you
need
differs
greatly
by
projects.
On
the
one
hand
you
have
OpenStack
and
on
the
other
hand,
you
know
you
have
the
tiny
guy
in
his
garage
with
the
NTP
Foundation
I
think
there's
an
opportunity
for
the
IETF
to
help
the
stuff
and
also
tap
into
some
kind
of
revenue
stream.
A
That's
a
great
conversation
for
continuing
at
is
at
2.0
and
and
we're
and
getting
to
a
point
where
we
can
not
just
discuss
the
new
operational
model
of
the
IETF
and
our
software
lychees,
because
that
may
be,
but
the
actual
goals
of
the
organization
that's
directly
one
of
the
other.
So
thank
you,
okay.
So
thanks
everybody
I
appreciate
the
time
today,
hopefully
is
interesting
for
you
and
catalyzes
some
from
conversation.
Thank
you
very
much.