►
From YouTube: IETF114 DINRG 20220728 2000
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
A
A
So
I'm
the
culture-
and
I
have
two
co-chairs
sitting
in
front
today-
lisha
zhang
and
mark
fedon-
and
maybe
briefly
explain
this.
So
we
realized
that
mark
and
folks
had
organized
a
site
meeting
on
a
quite
similar
topic,
and
so
we
just
got
in
touch
and
in
touch
and
decided
would
be
good
idea
to
join
forces
and
like
use
this
venue
for
discussing
centralization
today.
So
thanks
for
making
it
to
the
meeting.
A
Just
a
few
housekeeping
items
I
mean
thursday,
you
probably
have
seen
this
a
few
times
so
we're
also
using
the
the
meat
echo
queue
management
for
managing
questions
and
so
on
so
make
sure
you're
signed
in
into
the
session
and
okay,
it's
a
bit
awkward
for
me
to
say,
but
please
make
sure
you're
wearing
your
n95
or
fap
fap2rp3,
masks
and
yeah
keep
your
audio
and
video
under
control.
A
Okay,
so
we
are
following
the
so
we
are
the
ietf
irtf
here
and
we
are
following
the
itf
ipr
rules,
so
in
short,
that
means,
if
you
hear
or
say
anything
that
is
ipr
related,
you
are
expected
to
inform
us
in
a
short
time,
there's
also
the
privacy
and
code
of
conduct
that
we
are
running
by.
If
you
are
not
familiar
with
these
rules,
please
check
out
these
pointers
here
and
then
finally,
yeah
just
to
remind
everybody.
A
So
this
is
now
a
meeting
that
happens
in
the
internet
research
task
force,
so
the
research
system,
organization
of
the
ietf.
We
are
not
doing
standards
here.
A
A
Okay,
and
so
if
you
haven't
been
around
in
dinergy
yeah,
just
think
of
I
think
about
signing
up
the
mailing
list,
and
so
we
really
need
a
note
taker
for
today.
So
we
have
one
hour
so
not
a
lot
of
time,
but
we
are
hoping
to
get
many
good
ideas
out
of
this
meeting
and
we
really
really
need
to
capture
those.
A
And
yeah,
as
usual
notetaking
doesn't
require.
You
know
a
complete
record
of
what
is
being
presented,
but
just
the
gist
of
the
discussions
or
questions
and
the
the
core
of
the
answers.
B
Dirk
is
being
very
polite
here,
but
I'll
just
pick.
Somebody.
A
Better
for
you
to
do
it
in
the
room.
B
A
Thank
you
very
much
so
there's
the
the
cody
nd
as
usual
with
the
agenda.
If
you
want
to
use
that,
that
would
be
great.
A
And
yeah,
so
we
have
a
fairly
light
agenda
today.
So
if
you
have
been
around
in
dinergy,
you
may
remember
that
we
had
a
really
interactive,
engaging
workshop,
say
a
bit
more
than
one
year
ago
and
with
panel
discussions,
presentations
and
so
on,
and
that
yeah
really
brought
many
many
good
ideas
to
light
and
so
lisa
and
I
finally
managed
to
to
capture
those
and
are
currently
in
the
process
of
finalizing
a
written
report
with
the
help
of
jeff,
houston
and
christian
hudema.
A
And
so
what's
what
we
planned
for
today
was
that
malicia
will
provide
a
summary
of
this
report
of
this
meeting
one
year
ago
that
has
yeah.
I
think
many
good
points
that
were
discussing
and
then
we
we
hope
to
get
a
do.
A
round
of
discussion,
get
feedback
or
new
ideas
and
kind
of
use.
A
This
meeting
for
like
also
starting
like
the
next
steps
in
in
this
direction,
so
like
figuring
out
topics
but
maybe
also
way
of
working
where
they're
working
and
yeah
without
further
ado,
I
would
say,
alicia.
Let's
start
with
the
summary.
E
E
E
This
is
a
very
incomplete
list
showing
on
the
slides
about
the
existing
internet
draft
on
the
topic
in
particular,
I
want
to
point
out
that
you
know
the
first
field
identifies
the
observed
centralization,
phenomenons
and
the
second
half
of
the
list.
You
can
see
that
there
is
a
great
attention
paid
to
the
security
issues
ongoing
on
the
today's
internet.
E
It
may
not
be
the
best
thing
to
say,
but
let
me
say
I
think
we
are
yet
to
gain
upper
hand
to
put
the
internet
security
really
under
complete
control.
I
think
that's
a
big
factor
that
I
think
accelerated
the
centralization,
as
we
gonna
elaborate
in
the
following
report.
E
In
short,
I
think
those
existing
work
actually
reflected
the
many
shared
understandings
and
the
insight,
but
so
why
this
not
very
early
workshop
on
the
topic.
E
We
set
the
goal
to
say
that,
instead
of
them
into
designing
new
solutions,
let's
first
understand
what
are
the
root
causes.
What
has
happened
in
the
last
30
some
years
that
turned
the
internet
from
originally
somewhat
pretty
decentralized
operations
into
where
we
are
today
exactly
what
happened,
then?
What
we
can
learn
from
those
observations
we
want
to
put
a
caveat
up
front.
E
The
workshop
discussions
are
very
limited
in
scope.
We
largely
focused
on
the
application
providers,
consolidation
centralization
and
didn't
get
into
other
aspects
of
the
consolidation.
For
example.
Internet
access
is
obviously
today
are
in
the
hand
of
very
few
providers
and
there's
a
number
of
other
aspects
of
the
centralization,
but
in
the
following
report,
summary
we're
going
to
mostly
focus
on
what
happened
to
the
applications.
E
So
this
is
only
one
effort
into
understanding
the
problem
space.
I
think
more
effort
will
be
needed
to
better
understand
the
overall
picture
in
the
workshop
jeff
houston.
I
hope
he's
here.
Thank
you
jeff.
If
I
say
anything
wrong,
please
correct
me.
So
jeff
houston
gave
a
great
opening
talk
and
told
us
that
the
centralization
is
nothing
new.
It's
actually
started
with
industrialization.
E
By
over
a
hundred
years,
you
can
observe
from
a
history
that
economic
economy
of
scale
does
drive
consolidation
so
that
the
industry
always
evolves
towards
centralized
control
over
specific
industry
sectors,
and
you
know
it's
kind
of
a
nature
when
the
corporates
wanted
to
maximize
their
profit.
E
That's
not
exactly
the
same
action
they're
going
to
result
in
maximizing
the
user
interest
or
society's
interest
as
a
whole
and
therefore
in
the
history.
You
can
think
back
that
time
in
1890
the
sherman
anti-trust
act
to
curtail
the
control
power
of
the
industrial
giants
and
many
years
later,
that
the
same
antitrust
act
also
dissolved
the
att
monopoly
in
the
telecom
industry.
E
E
Of
course,
time
changed
technology
changed
and
therefore
the
specifics
are
different.
The
past
the
industry
giants
they
exploited
the
labor
forces
to
accumulate
the
profit,
and
today
the
industry
of
the
internet
does
something
different.
They
gain
this
innovative
personal
data
and
then
profit
from
the
advertisement
gains
so
coming
to
the
next
slide.
Some
of
our
specifics.
E
This
is
given
by
christian
fuedema.
I
hope
it's
online
to
explain
what
exactly
happened
as
like
the
samples
through
the
past
30
years.
You
can
see
that
the
internet
centralization
before
we
paid
attention.
You
know
this
pioneer
companies.
They
actually
started
by
investing
in
to
provide
very
much
needed
service
when
the
internet
started
like
the
search
engine
like
the
email
like
social
networking
things
that
the
user
in
general
very
much
wanted.
E
Of
course,
over
the
time
those
services
get
consolidated,
we
started
with
different
search
engines
and
then
got
into
yahoo,
and
eventually
google
take
off
for
email,
I'll,
have
more
to
say
later
and
social
networking
facebook
was
not
the
first
one,
but
today
it
is
a
dominant
one
and
more
services
get
ideas
over
time.
So
you
can
observe
this
positive
feedback
loop.
You
get
more
users,
you
have
more
incentive
and
the
more
input
to
provide
those
users
better
services,
that's
gonna,
attract
the
more
users.
Then
you
get
higher
revenues
and
you
attract
more
users.
E
The
service
providers
wants
to
know.
You're
well
so
therefore
say:
when
you
do
search,
they
can
provide
you
exactly
which
one
but
the
more
they
know
about
you.
I
think
the
more
influence
they
can
have
over
you
through
the
information
selectively
provided
to
you.
There
is
an
accident
book.
That's
in
the
workshop
reference,
titled
surveillance,
capitalism.
E
So
the
the
networking
started
from
the
dominance
of
carriers
in
the
early
days.
You
know
people
mostly
paid
attention
to
say,
hey,
which
isps
actually
provided
service
for
most
users,
and
that
is
no
longer
much
of
a
topic
now,
because
the
attention,
the
money
and
the
profit
have
moved
up.
The
protocol
stack
so
later,
microsoft
became
the
dominance
in
terms
of
providing
the
operating
system
platform,
and
then
you
can
see
today
it's
not
so
much
operating
system
that
dominates
the
market
or
make
the
most
profit
inside.
E
E
So
if
you
move
on
to
the
next
slide,
the
applications
on
that
specifics
number
three:
the
applications
actually
is
much
easier
to
become
dominant
because
they
they
don't
have
to
do
anything
lower
layer
to
standardize
this
standard.
That's
that
inside
they
just
install
the
application
on
your
browsers
and
then
they
are
ready
to
go.
E
So
if
we
want
to
decentralize
the
applications
that
christian
warned
us,
that's
really
uphill
battle
against
the
centralized
monopoly,
because
decentralized
systems
or
applications,
you
need
standards,
and
we
all
know
how
long
and
how
much
effort
that
it
takes
to
get
into
the
standard
ready
and
go.
It's
really
far
easier,
cheaper,
faster
for
the
application
monopoly
providers
to
add
the
new
things
and
the
new
features.
They
don't
need,
standardization,
you
don't
need
to
coordinate
with
their
competitors;
they
just
do
it
and
deploy
the
code.
E
So
in
putting
the
workshop
reported
together,
we
actually
speculated
this
question.
That
is
we
commonly
hear
people
seeing
that
the
distributed
protocols
somehow
get
decentralized.
So
we
looked
into
answering
that
question.
It's
really
that
the
protocols
get
decentralized
or
something
else
actually
happened.
You
look
at
the
protocols
merely
shapes
the
bits
around.
E
Even
the
higher
level
protocols.
Dns
it'll
facilitate
that
shaping
by
translating
the
names
into
addresses
the
protocols
are
carriers.
They
don't
make
a
decision
on
where
the
pack
is
the
most
it's
the
application
deployment
that
actually
make
that
decision,
and
the
next
slide
that's
going
to
say
that
in
the
early
days
I
was
very
very
early
in
the
game
right
first,
at
mit
later
I
went
to
xerox
park
and
then
I
went
to
ucla.
E
Even
after
I
got
to
ucla
ucla,
we
provided
dns
services,
email
services,
our
own
web
services,
that's
the
norm
of
the
day.
The
in
the
applications
were
decentralized
provided
by
largely
individual
institutions,
but
the
ucl.
We
did
that
as
a
service
to
ucla
user
community.
They
are
not
revenue
generating
business,
but
once
the
service
providers
grabbed
those
applications
making
them
to
be
revenue
generating
services,
that
is
really
economical.
Forces
drives
consolidation,
the
bigger
the
better,
the
more
profit,
more
control
and
hence
bigger.
E
E
E
E
There's
a
big
percentage,
so
that
really
illustrates
that
centralized
service
does
have
their
edges,
because
the
lack
of
effective
security
mitigation
make
providing
individual
services
become
the
last
analyze
variable.
So
I
put
it
there
to
say
that
there
is
an
economy
of
skill
that
drives
the
centralized
service
to
become
a
bigger,
but
security
really
plays
a
big
role
in
accelerating
this
whole
process.
E
So
I
hope
those
two
slides
last
one.
This
one
helped
answer
the
question:
how
come
there
used
to
be
distributed?
Applications
become
centralized,
that's
because
they
become
a
revenue
generating
services
and
the
economy
drives
them
into
centralization,
and
thanks
to
mark
he
offered
this
reference.
You
can
see
the
the
pointer
to
the
real
paper.
This
is
just
the
you
know
somehow
evidence
to
show
that
for
web
hosting
there's
yet
another
service
that
got
centralized.
This
is
only
showed
a
five
year
span
of
web
hosting
for
websites
under
different
tlts.
E
As
you
can
see
in
the
five
years
the
centralized
services
roughly
roughly
doubled,
I
think
over
the
last
two
years
is
probably
getting
a
lot
worse
as
well.
So
if
we
move
to
the
next
slide,
we
can
say
that
make
it
short.
There
are
multiple
factors,
but
the
top
two
factors
driving
the
centralization,
its
economy
of
skill
and
security
threats.
That's
I
think
the
outcome
conclusion
out
of
the
workshop
just
make
it
simple.
So
what
to
take
it
away
going
to
the
next
slide?
It's
just
these
two
things,
I'm
repeating
myself.
E
The
centralization
becomes
so
rapidly
advancing
because
there's
very
few
regulations
to
counter
measure
the
big
market
force,
the
economy
economy
of
skill
really
motivates
corporations
to
go
big,
and
we
cannot
deny
that
the
fact.
As
a
matter
of
fact,
jeff
pointed
out
very
clearly.
This
is
not
really
the
pure
size.
E
The
real
concern
about
centralization
is
their
control
power,
their
influence,
they
decentralize
the
influence
and
control
over
all
the
users
and
over
the
security
of
the
society
as
a
whole.
That's
our
big
big
concern.
So
security
challenges.
You
know
why
security
accelerated
the
whole
centralization,
because
we
don't
really
have
this
magical
power
to
mitigate
the
security
threats
very
effectively
for
ddos
mitigation.
That's
the
fact
that
the
major
mitigation
means
is
absorbing
it,
and
the
only
big
systems
has
a
capacity
to
do
that.
E
There
are
limitations
on
the
existing
web
security
framework
that
doesn't
allow
your
phone
and
my
phone
communicate
directly.
How
we
communicate
go
up
to
the
clouds
go
through
the
facebook,
then
we
communicate.
Your
phone
doesn't
really
know
my
phone
home
that
it
belongs
to
there's
no
security
in
today.
No
security,
no
communication,
and
we
don't
really
have
security
solutions.
E
As
of
now
deployed
for
direct
user-to-user
communication
to
enable
truly
decentralized
applications-
that's
just
maybe
stopping
it
and
then
the
workshop
actually
spent
some
time
elaborate
on
the
difficulties
on
developing
and
deploying
security
solutions.
It's
a
hard.
It's
really
hard
and
people
pointed
out.
I
forgot
who
mentioned
that
security
itself
is
really
hard
and
the
interoperable
security
is
even
harder.
If
you
want
to
decentralize
applications.
E
So
the
society
strives
when
you
keep
a
balance
between
economy
between
the
regulation
and
the
technologies,
and
today
we're
in
this
kind
of
limbo
stage,
because
economy
drives
drive
these
things
forward
by
the
large
the
regulation.
I
think
they
try.
They
really
try,
but
the
a
big
challenge,
what
exactly
to
regulate
and
how
and
then
the
technologies.
E
As
I
mentioned,
we
are
finding
the
hand
behind
the
security
threats.
We
don't
have
effective
solutions
to
mitigate
them
as
of
now
so
as
a
result,
I
think,
at
least
for
us
drafted
the
report.
We
believe
that
the
effective
regulations
and
the
legislation
we're
actually
playing
a
deciding
factor
in
curtailing
the
unconstrained
market.
As
much
as
this
itf
community
has
a
brilliant
engineers,
researchers,
designers,
the
technology
solutions
alone.
E
We
don't
think
that
it
can
mitigate
the
centralization
on
its
own,
so
it
doesn't
mean
that
ways
just
step
aside
and
watching
how
the
regulation
goes.
As
I
mentioned
already,
regulators
need
our
help
to
figure
out
what
to
regulate.
I
think
it's
really
our
responsibility
to
help
them
and
we
work
with
the
regulations
effectively
by
providing
new
solutions
for
the
already
identified
security
problems.
E
E
So
this
is
my
last
slide,
essentially
all
those
one
more
so
further
discussions
are
needed
and
that
we
thought
that
thing
rg
could
be
the
platform
for
collecting
all
the
inputs
and
following
a
platform
to
carry
on
the
future
discussions.
I
think
they're
really
hard
questions
that
need
answers,
so
say
that
I
always
tell
the
students,
if
you're,
facing
a
big
problem.
Where
do
you
get
started?
You
ask
yourself
this
question:
if
the
god
grants
all
your
wishes,
what
do
you
wish
to
happen?
E
If
we
have
a
clear
answer
to
that
question,
then
we
can,
from
there
figure
out
how
we
get
there.
So
that
very
first
question
is:
what
do
we
wish
to
see
as
an
ideal
outcome
out
of
all
this
effort,
mitigating
decentralization
and
that
need
to
take
into
account
all
these
major
factors
in
the
play.
Big
data
has
advantages.
E
A
Yeah
we
have
a
few
people
lined
up
in
the
queue
and,
let's
start
with
jose.
G
Yeah,
I
I
really
I
really
like
this
workshop,
but
I
come
from
a
completely
different
world.
G
I
come
from
the
iot
world,
where
we
have
the,
I
would
say
the
internet
and
the
small
and
where
we're
at
this
point,
doing
distributed
and
very
decentralized
applications
inside
our
own
networks,
and
so
of
course,
there's
blockchain
all
kind
of
other
stuff,
but
I'm
not
even
going
there
because
we're
a
lot
of
time
we're
within
the
same
domain.
So
we
don't
really
have
to
deal
about
deal
with
this,
but
when
the
challenge
that
we
have
is
because
of
the
centralization
a
lot
of
the
systems
we're
using
or
are
vertically
integrated.
G
So
although
we
have
two
sensors
that
are
co-located
in
the
same
place,
the
first
time
they
talk
to
one
another
is
somewhere
in
aws
or
azure
or
google
or
whatever,
and
and
we're
working
really
hard
to
make
sure
that
we
can
create
the
applications
that
are
necessary
for
next
generation,
industrial
systems,
for
example,
or
or
or
smart
cities
where
sensors
talk
to
one
another
because
they
are
located
in
the
same
place.
G
So
I
I
understand
that
your
workshop
looked
at.
I
would
say
the
big
picture,
but
there's
also
this
internet
in
the
small
this.
This
need
to
have
co-located
devices
on
the
same
network,
to
be
able
to
talk
to
one
another
and
not
being
completely
vertically
integrated,
and
I
think
the
centralization
actually
is
preventing
that
and
the
reason
we
need
decentralization
is.
G
We
need
to
have
the
system
to
talk
to
one
another,
and
you
know
you
were
there
maybe
early
this
morning
when
somebody
was
mentioning
the
use
of
ai
and
networking,
and
if
we
want
to
have
data
driven
systems
and
ai
driven
systems,
we
need
to
have
local
information
to
be
integrated.
G
E
A
Thank
you
by
the
way
are
we
taking
notes,
I'm
taking.
F
Hi
this
is
dino,
I'm
not
going
to
ask
a
question,
but
I
want
to
get
a
reaction
by
using
an
example,
I
think
all
the
monopoly
companies
that
are
offering
centralization
are
helping
for
the
examples
you
use
with
email
and
sometimes
maybe
we
give
up
the
decentralization
for
the
advantages
of
centralization.
F
Where
we
would
want
to
talk
directly
to
each
other
still
encrypted,
but
across
the
airway
here,
20
meters,
that
this
would
be
a
good
feature,
and
I
have
a
feeling
they
would
be
motivated
to
do
it
if
we'd
be
willing
to
pay
extra.
For
this
feature,
just
like
we
pay
extra
for
not
having
advertisements
in
various
apps
reactions,.
A
So
let's
collect
a
few
views
and
statements
and
then
maybe
try
to
come
back
and
structure
this
a
little
bit
so
so
far
we
we
have
heard
you
know
things
about
local
communications
or
iot
in
the
small,
so
dino
just
alluded
to
direct
local
communication
and
the
like
trust,
anchor
topic
that
licia
actually
talked
about.
H
All
right
with
terrible
torah
from
open
exchange.
So
I'm
sorry
I
didn't
this
is
all
I
know
about
the
workshop.
So
I
only
it's
a
an
immediate
reaction,
so
you
ask
what
is
missing
well,
I
was
surprised
because
in
the
end,
you
listed
these
two
factors
as
the
major
factors
for
centralization
and
uk
security.
H
I'm
also
surprised
that
by
not
seeing
the
other
factors
that
are
often
mentioned-
I
don't
know-
maybe
it's
just
a
matter
of
terminology,
but
when
we
have
these
discussions
in
other
environments
that
mostly
in
europe,
mostly
maybe
in
competition
discussions,
there
are
other
worlds
that
come
up
like
one
gardens
network
effect,
self-preferencing,
so
a
number
of
practices
which
are
both
technical
and
business
practices.
So
I'm
wondering
whether
this
came
up
in
any
way,
whether
it's
just
a
matter
of
aligning
the
language
or
it's
really
a
part
of
the
analysis
that
is
missing.
E
E
E
Like
you
know,
I
already
mentioned,
we
are
so
great
in
terms
of
the
low-level
communication
technologies.
The
bluetooth
can
work,
a
wi-fi
can
work
directly.
What
is.
F
A
I
Yeah
hi,
there's
sachin
here
I
had
this
point
that
you
made
about
the
email
thing,
the
point
that
the
university
email
was
not
removing
as
much
spam
as
gmail
has
been
doing
and
just
the
email
example
you
gave
about
spam
yeah.
So
this
comes
back
to
the
point
that
I
was
actually
discussing
with
the
fast
mail
guys-
and
this
is
important
because
technically
speaking,
email
is
not
something
that
cannot
be
run
easily
by
normal
people.
I
mean
reasonably
technical
people.
I
We
are
all
capable
out
here.
I
guess
of
running
our
own
email
servers
at
least
80
percent
would
be
the
standards
this
server
just
run.
The
problem
happens
is
when
I
run
my
own
server
somewhere.
Someone
would
have
added
that
particular
ip
address
to
some
long
list
which
will
prevent
my
email
from
going
out,
which
does
not
happen
if
I'm
taking
an
email
service
from
a
gmail
or
a
microsoft,
or
somebody
like
that
and
that's,
I
think,
one
of
the
issues
that
we
possibly
should.
I
I
don't
know
how
we
can
solve
it,
it's
more
of
a
social
issue
rather
technically.
I
think
it's
a
solved
sort
of
a
thing,
but
we
need
to
get
back
to
the
basics
and
try
to
possibly
understand
what
people
really
want
to
decentralize
how
they
want
to
decentralize
it.
What
they're
giving
up
I
mean.
That's
just
my
thinking.
A
E
E
C
All
right,
hi,
andrew
campbell,
speaking
in
terms
of
you,
post
the
question,
what's
missing,
I
would
have
expected
to
see
loss
of
resilience
on
here.
Something
there's
been
a
significant
loss
of
resilience
in
the
internet
infrastructure,
arguably
because
of
standards
and
because
of
consolidation
in
different
ways.
I
think
that's
an
issue.
C
I
agree
with
victorio
that
it's
not
economics
of
scale,
it's
much
more
about
network
effects
at
the
application
layer,
so
we
expect
reference
for
that
and
the
other
thing
I
don't
think
I
saw
is
loss
of
privacy
and
again
because
of
the
application
layer
and
the
network
effects
of
the
application
layer.
C
There's
research
out
there,
which
tells
us
that
people
don't
want
to
give
up
their
personal
data
but
feel
they
have
no
choice
so
they're
reluctantly
giving
it
up
and
that's
an
enormous
issue
again,
that's
because
of
the
network
effects
they
feel
they
have
no
choice
but
to
use
these
apps
and
therefore
the
sort
of
digital
surveillance
that's
undertaken
at
the
application
layer
is,
is
a
huge
problem
and
it's
facilitated
by
the
consolidation.
A
Thanks
andrew,
it's
just
as
a
note
is
just
closed
the
queue
because
I
think
we're
running
out
of
time.
Otherwise
so
but
let's,
let's
move
on
mallory,
please.
J
Hi
everybody
mallory
noodle
from
the
center
for
democracy
technology.
I
think
that
the
report
back
is
really
interesting
and
I've
been
thinking
about
it
a
lot
so
thanks
for
presenting
it,
I
wonder
if
it
doesn't
look
very,
very
different
at
different
layers
of
the
internet.
J
I
mean
I
was
just
thinking
about
the
case
of
hardware
and
then
you
follow
that,
naturally,
all
the
way
down
and
it's
like
do-
we
want
a
bunch
of
companies
mining
for
rare
minerals
in
the
congo
like
no,
so
you
know
there
are
going
to
be
some
natural
points
where
things
get
centralized,
but
potentially
mapping
out
very
specifically
in
each
of
these
layers
from
each
of
these
perspectives,
all
along
the
stack.
J
What
are
the
sort
of
benefits,
sometimes
of
centralization?
What
are
obviously
the
detriments?
What
are
the
mitigations?
Also?
What
does
it
look
like
from
providers
as
users
versus
actual
end
users?
I
think
a
lot
of
us
in
this
space
think
about
end
users,
but
we
actually
come
to
the
space
as
as
implementers
in
our
users
in
our
own
right
in
a
different
sort
of
view.
J
So
I
think
that
that
work
could
actually
be
really
useful,
because
we
do
have
a
lot
of
really
interesting
examples
that
we
like
to
talk
about
a
lot,
especially
when
it
comes
to
things
like
interoperability
or
the
you
know,
anti-patterns
like
this
idea
that
you
know
blockchain
solutions
are
fully
decentralized
when
we
know
very
well
from
a
code
perspective,
they
aren't,
but
you
know
so.
These
examples
abound
and
a
lot
of
people
have
brought
up
really
good
ones.
Today.
A
K
K
K
K
K
This
is
not
telephony
anymore.
There's,
no
industrial
widgets.
How
do
you
regulate?
What
should
we
do
and
the
real
answer
to
that
is
we
don't
know?
No
one
knows
it's
new
experience,
okay,
so
those
two
won't
help.
Will
users
get
us
out
of
this?
No,
because
this
entire
industry
is
focused
on
giving
users
precisely
what
they
want.
K
Tik
tok
is
what
they
want,
not
what
they're
forced
to
do.
Oh
god,
I've
got
to
do
tick-tock
today,
no
way
this
entire
industry
is
actually
based
around
understanding
and
predicting
your
needs
and
meeting
them.
So
anything
you
try
and
do
to
curb
that
you're
actually
going
against
what
users
want.
I
want
my
search.
I
want
my
mail,
I
want
my
facebook
whatever,
and
so
users
aren't
going
to
help.
K
K
At
some
point,
it
becomes
such
a
damning
wall
of
stasis
that
nothing
ever
changes
that
the
pressures
to
change
knock
it
all
down.
That's
the
way
we
destroyed
telephony.
It
was
unpleasant
lots
of
folk
locks
lost
jobs.
All
of
these
changes.
Ultimately,
once
you
get
to
this
position,
there's
no
clean
way
out
other
than
actually
revolutionary
ways
that
actually
end
up
destroying
the
incumbents
in
the
existing
sets
and
replacing
with
different
structures.
K
J
E
E
So
therefore,
today,
when
we
talk
about
centralization,
we're
talking
about
different
things
that
get
decentralized,
not
the
low
level
ip
level
connectivity,
not
yet
I
don't
know
whether
it
actually
will
be
consolidated
in
some
near
future,
but
but
so
today's
centralization
is
really
at
a
much
higher
level
than
what
the
ip
started
aim.
The
form
I
want
to
make
that
very
clear.
E
As
of
today,
the
bgp
regulated
the
connectivity,
that's
not
in
any
single
hand
yet
regarding
the
revolution,
jeff
talked
about.
That's.
I
think
I
very
much
agree
with
that.
I
hinted
in
the
earlier
comments
about
how
to
do
search
today.
If
you
do
the
search
google
does,
then
google
find
the
best
way,
but
if
you
really
open
up
the
mind,
you
think
in
the
fundamentally
different
ways
of
doing
search,
that's
really
a
revolution
that
can
bring
you
a
very
different
world
from
them.
D
D
The
the
one
thing
that
we
asked
me
that
was
a
bit
mean
thing
in
alicia's
summary.
Is
that
a
huge
part
of
her?
The
problem
is
the
accumulation
of
power
and
the
accumulation
of
power
is
not
just
about
insulin
strength,
but
it's
also
about
big
data
and
big
data
can
be
used
to
make
better
product,
but
it
can
also
be
used
to
as
a
barrier
to
entry.
It
can
be
used
as
a
mean
of
influence
and
things
like
that.
D
Big
data
itself
derive
from
a
very
specific
decision
to
fund
the
internet
services
using
advertisement
and
make
them
appear
free
and
behind
this
power
and
decentralization
power.
We
have
this
advertisement
business
and
personalized
advertisement,
which
is
largely
driven
by
big
data.
Now
it
has
also
driven
the
acquisition
of
data.
D
L
Thank
you
looks
like
I
get.
The
last
word
what's
been
interesting
about
this
discussion
is
that
it
shows
that
we
need
further
discussion
and
that
this
particular
group
is
probably
the
best
place
for
it.
There's
so
much
interest
and
impassioned
interest
that
I've
just
seen
in
the
last
30
minutes,
but
also
just
a
couple
of
other
thoughts
and
ideas
for
direction.
One
is
measurement.
We've
already
seen
a
couple
of
really
good
presentations
in
the
workshop
last
year
and
and
researchers
looking
at
that.
L
L
And
finally,
I
know
we've
just
been
discussing
the
economic
point
of
view,
which
I
do
think
is
really
a
main
driver,
but
also
looking
at
the
unintended
consequences
of
regulation
and
other
aspects
that
show
up
around
the
ecosystem
of
the
the
technical
aspects
would
be
a
really
good
approach.
So
thank
you
for
having
this
discussion
today.
A
Thanks
dominic
yeah,
that
was
actually
a
good
statement
in
a
sense
that
so
I
mean
we
have
heard
many
inputs,
and
so
even
if
you,
for
example,
buy
you
know,
jeff's
a
pessimistic
view,
you
have
the
different
things
you
could
do
right,
so
you
could
say
you
know,
think
about
disrupting
developing
developing
disruptive
technologies.
A
But
even
if
you
do
evolution,
somebody
probably
should
analyze
things
very,
very
well
like
say.
So
what
are
the
the
factors?
What
what
is
has
been
the
history
and
so
on.
So
we
have
been
circling
around
all
these
different
potential
activities.
So
in
the
workshop
last
year,
and
also
today
and
so
yeah,
we
felt
would
be
maybe
interesting
to
identify,
say,
activity
topics
that
people
would
be
most
interested
in
for
say,
follow-up
activities
and
yeah.
It's
it's.
A
The
energy
started
with
a
say,
maybe
over
ambitious
approach,
some
some
years
back.
Actually,
when
you
know
people
thought
okay,
blockchain
is
is
saving
us
from
centralization,
and
so
now
we
have
taken
this
step
back
and
so
also
at
the
workshop
last
year
I
mean
we.
We
we
heard
really
very
well
thought
through
presentations
of
the
historic
developments
and
yeah,
also
the
commercial
centralization
concentration
and
the
question
is
so
what's
next.
So
what
where?
Where
can
we
actually
make
a
difference?
A
And
so
that's,
I
think
something
where
we
have
gotten
new
new
inputs
today,
but
we
need
to
structure
this
somehow,
and
so
that's,
I
think,
next
on
the
agenda
for
us.
B
So
I
think
that
might
be
my
cue.
We
only
have
a
couple
minutes
left,
but
let
me
sort
of
look
to
the
future
rather
than
the
workshop
in
the
past.
What
I'd
like
to
do
is
if
you
look
at
the
second
bullet,
that's
on
the
screen
right
now.
One
of
the
things
that
we
observed
in
the
last
12
to
18
months
is
that
there
are
many
people
writing
independent
submissions,
individual
submissions
that
are
talking
about
various
various
features
of
consolidation
and
centralization.
B
B
B
And
that
would
be
the
future
direction
here
is
to
activate
ti
nrg
by
actually
making
it
an
active
home
for
these
kinds
of
conversations,
but
also
not
just
the
conversations
but
real
contributions
right
contributions
on
economic
aspects,
contributions
on
architectural
aspects,
contributions
on
protocol
aspects-
and
we
see
that
there
are
individuals
already
interested
in
doing
that,
and
so
that's
the
proposal
that
I'm
going
to
make
for
moving
us
forward
here
in
dinrg
when,
given
that
we
are
nearly
out
of
time
jerk,
I
I
once
again
encourage
all
of
you
to
be
on
the
mailing
list,
where
I'll
actually
put
that
proposal
to
the
mailing
list
as
well.
A
Yeah
thanks
mark
there's
still
two
people
in
the
queue.
I
don't
know
if
you
want
to
honor
their
request
yeah,
so
I
mean
we
still
have
a
bit
of
time.
So
christian,
do
you
want
to
add
something.
D
I
think
that
if
someone
wants
to
associate
with
tommy,
he
makes
a
very
good
point.
I
mean
if,
if
we
think
that
cutting
off
the
trafficking
of
data
will
cut
off
the
oxygen
that
decentralization,
then
the
atf
has
something.
What
to
do,
which
is
minimizing
data
in
everything
we
do
and
minimal
data
will
be
minimizing.
Data
across
protocols
will
be
very
nice.
It's
something
that
we
can
do.
D
The
the
other
thing
that
we
could
do
is
look
at
the
very
example
that
leash
has
given
about
the
distributed
dots
and
things
like
that.
There
is
work
going
on
in
the
itf
to
have
open
solutions
like
those
problems,
so
the
more
work
we
get
there,
the
better
we
make
the
system
as
well.
These
are
two
concrete
area
in
which
we
can
work
might
very
well.
Let's.
D
M
Thank
you,
so
I
I
think
what
can
we
as
engineers?
Do
we
as
people
contribute
to
the
ietf?
What
can
the
iecf
do?
M
I
think
it
comes
down
to
the
best
we
can
hope
for
is
to
build
solutions
that
lower
the
bar
to
entry
and
make
it
inexpensive
for
anyone
to
contribute.
Now
you
know
if,
if
the
giants
end
up
then
providing
a
level
of
value
that
the
smaller
players
decentralized
players
couldn't
each
meet,
then
I
don't
know
that
there's
a
solution
to
that.
You
know.
M
Ultimately
the
consumer
decides
and
if
you
know
the
consumer
finds
gmail
to
be
better
than
running
a
mail
server
in
their
house
they're
going
to
do
that
as
one
of
the
few
people
who
run
a
mail
server
in
my
house,
it
used
to
be
easy.
It's
not
easy!
Now
it's
kind
of
kind
of
a
pain
in
the
ass,
but
you
know
people
move
to
where
they
find
value
and
if
they're
going
to
find
value
from
a
giant
they're
going
to
use
the
giant,
I
don't
know
that
we
can.
M
I
don't
I
don't
know
we
can
fix
that
and
I'm
pretty
confident
regulation.
I
don't
have
a
lot
of
confidence
that
regulation
could
solve
that
problem.
I'd
be
more
fearful
that
whatever
regulation
comes,
would
probably
make
things
worse
and
would
probably
empower
giants
more
than
it
would
hurt
them.
M
So
I
think
ultimately,
the
best
thing
we
can
do
is
to
focus
on
designing
solutions
and
empowering
technologies
and
evangelizing
technologies
that
democratize,
you
know
the
contributors
of
content
and
the
users
of
the
technology,
and
you
know
whoever
builds
the
best
mousetrap
wins,
whether
it's
you
know
a
a
bunch
of
small
companies
or
a
few
large
ones,
but
I
think
that
this
working
group
does
provide
value
in
you
know,
maybe
creating
a
level
of
consciousness
on
those
issues.
A
Okay,
we
are
out
of
time
lisa.
Do
you
want
to
say
some
final
words.
E
I
think
it's
a
really
interesting
discussion.
I
appreciate
is
the
last
speaker
input.
We
can
chat
more
afterwards,
but
I
guess
this
is
a
very
good
starting
point
like
mark
mentioned
that
we
can
get
to
this
thing
rg
as
the
platform
to
at
the
focal
point
to
continue
this
conversation.
I
really
agree
with
that
suggestion.
That's
so
to
say
now.
A
All
right,
thank
you
very
much
for
coming
and
have
a
nice
evening
in
philadelphia
hope
to
see
you
soon.
Bye-Bye.