►
From YouTube: Messaging for Web3 - April 23, 2019
Description
Messaging for Web3
A
D
A
A
A
Maybe
we
can
have
a
brief
sink
up
round
and
then
there's
some
stuff
you
want
to
talk
about
when
it
comes
to
swarm
and
stuff
follow-up
from
from
from
last
call
and
then
I
know
valid
labs,
there
were
some
questions
on
interfacing
and
so
on,
and
then
NIM
had
some
questions
and
simulations
on
it's
just
something
else
that
people
would
like
to
talk
about.
Forget
start.
B
Maybe
Oscar
we
could.
This
would
also
be
a
good
time
to
maybe
get
into
kind
of
a
little
bit
more
exhaustive
list
of
projects
in
a
space,
because,
while
digging
about
a
bit
deeper
into
interest,
form
and
things
for
the
first
intro
last,
we
go
Victoire.
You
know
what
was
really
really
good.
Yeah
I
was
doing.
Some
reading
came
across
some
other
projects
that
I
just
wanted
to
bring
up
and
yeah.
That's
a
topic
as
well.
B
A
E
E
F
Started
doing
some
things,
but
I
haven't
been
I
haven't
made
some
virus
mostly
what
so,
two
weeks
ago,
at
the
call
we
had
some
by
the
way
I'm
Jeff
I'm,
also
a
researcher
at
web
3
I
also
sometimes
do
I.
Do
many
things,
though
not
enough.
Lately,
the
yeah.
There
was
a
request
from
validity
labs
to
have
sort
of
an
understanding
of
what
the
layering
of
did.
F
All
of
the
different
components
is
so
I
started
trying
to
make
some
notes
on
that
sort
of
during
the
call
last
week,
which
was
probably
not
the
ideal
time
for
that,
but
anyway,
so
I
will
at
some
point
very
soon.
Probably
today
finish
those
finish
up.
Those
notes,
that'll
just
give
a
sort
of
a
high
level
picture
of
what
things
go
where
and
yeah
that,
hopefully
will
simplify
well,
it
will
hopefully
help
structure
conversations.
D
Victor
trollin
from
the
stone
team-
and
he
me
joining-
is
a
source
of
blood
and
Lewis
Lewis
doesn't
have
a
mic
so
he's
renting.
He
is
the
author
of
PSS.
We
have
blood
I,
don't
know
he.
He
can
speaker
himself
that
his
is
doing
from
whisper
and
networks
play
absolutely
assesses,
we're
so
over
to
blood.
B
Yeah
hi
I'm
Sebastian
from
the
late
labs
city
or
validity
labs
here
and
hopper
hopper
is
basically
the
project
that
how
we
call
it's
what
we
develop
here
at
village,
labs
Robert,
is,
is
the
main
developer
behind
that
Robert
is
in
vacation
for
the
next
two
and
a
half
weeks,
so
not
with
us
here
yeah.
So
it's
it's
we're
aiming
basically
if
having
a
privacy-preserving
messaging
protocol-
and
you
have,
over
the
last
two
weeks
finishing
up
a
little
bit
of
Reece
restructuring
of
our
implementation.
B
F
There's
one
other
thing
I
can
mention,
so
it
looks
like
Harry's,
not
in
the
media.
I,
don't
think
anybody
from
panoramix
or
Nia
is,
but
though
there's
one
person
who's
commonly
in
the
meeting
is
Harry
helping
and
hopefully
we'll
have
more
people
from
the
NIMH
project
in
at
some
point
in
the
future.
That's
they
have
a
kind
of
a
mix
that
built
there's
some
things.
F
A
Some
experiences
on
a
fast
but
a
bit
a
bit
of
a
better
sort
of
requirements
list
and
so
on.
You
can
see
a
link
here.
We
were
posting
our
discuss
as
well,
and
then
Adam
has
been
doing
some
great
work
on
documenting
so
far
current
protocol
in
the
various
parts
of
it.
So
so
we're
currently
in
the
process
of
sort
of
breaking
that
up
into
smaller
pieces,
and
what
that
allows
us
to
do
is
that
we
couldn't
be
writing
or
whisper.
A
I
also
want
to
mention
briefly,
because
I
think
I've
been
a
bit
sloppy
in
some
communication
to
post
that
the
data
Sinclair
does
not
put
any
sort
of
strong
requirements
on
transport
layer
so
that
shouldn't
be
Israel
interfaces.
Concern
I
started
a
solid,
an
issue
with
sort
of
the
rough
interface
in
in
riot,
but
essentially
second,
here's
an
issue,
but
essentially
all
the
requirements
and
some
come
from
that
from
the
image
layer.
A
Above
so
that's
that's
most
of
the
conversation,
security
layer
and
that's
what
things
like
MLS
comes
in
where
there's
some
salt
ease
when
it
comes
to
these
these
requirements
for
sequencing
service
or
whatever.
But
it's
not
something
we
looked
into
too
much,
but
what
you're
gonna
try
to
have
some
person
and
at
all
representatives
that
or
end
this
group
as
well
at
MLS
hackathon,
it's
probably
gonna
be
Igor
from
the
core
team.
A
What
else?
Oh
I
guess
also:
we
did
this,
this
basic
proof
of
concept
to
play
on
with
swamp,
ESS
and
feeds,
which
is
essentially
a
log
based
form
of
communication
where
we
used
essentially
using
possesses
at
a
whisper
and
then
using
feeds
to
sort
of
up
to
also
push
your
updates
in
them.
When,
when
you
you
come
online
again,
you
can
look
in
the
other
person's
feed,
and
then
you
have
pointers
to
previous
message
and
so
on.
A
A
All
right
so
I
guess
yeah
there
was.
There
was
a
bit
of
a
a
it's
kind
of
called
selected,
this
interest
between
dead
mix
and
embedded
labs
and
and
swim,
and
it
seems
like
both
products
were
not
super
familiar
with
each
other.
So
we
we
said
we
talked
in
two
weeks
again
when
people
have
a
chance
to
read
up
on
stuff
mister
yeah,
who
would
like
to
start
like
how?
How
are
people
change
their
thinking
more
so
yeah,
so.
B
So
yeah
I
took
the
weekend
to
do
a
bit
more
reading
into
into
things.
So
thanks
for
for
sharing
all
the
materials
Victor,
that
was
that
was
quite
helpful,
and
it's
it's
still.
Some
things
are
a
little
bit
unclear
to
me
still
with
regards
to
to
swarm.
Maybe
you
could
help
me
clarify.
So
basically,
you
have
this
incentive
mechanism
that
is
baked
into
swarm
right.
Is
that
already
part
of
the
current
version
of
swarm?
Is
that
for
the
future
to
come.
D
Took
some
time
to
find
them,
I
knew
better
thanks
for
the
question
so
yeah.
So
it's
not
currently
working
it
already
had
at
POC
2
version.
We
already
had
a
working
version
of
the
of
the
swap
protocol
with
the
with
the
checkbook
contract
and
everything.
So
it
was
working
with
with
the
new,
with
the
new
network
layer
and
with
the
rewrites.
D
We
still,
we
still
have
it
outstanding
and
it's
going
to
be
finished,
I
guess
in
in
q2
or
q3,
and
at
the
moment
we
so
we
have
the
protocol
accounting
mechanism
done
implemented,
but
it's
not
hooked
up
with
the
with
the
most
WAP
mechanism
for
for
what
tit
for
tat
and
then
for
for
check
payments
and
its
etc,
so
that
the
incentive
mechanism
is
quite
thoroughly
worked
out
and
also
the
smart
contract
for
that
for
a
new
child
book
contract.
This
is
what
so
done
so
so.
B
Not
ok,
I
had
the
feeling
cool
and-
and
the
second
question
is
you-
you
basically
allow
mice
yeah.
You
basically
have
this
gradient
right,
so
it's
more
a
spectrum
kind
of
of
having
a
more
whisper
like
approach
or
having
something
that
is
pretty
targeted
towards
the
goal
directly
towards
the
recipient
directly.
So
if,
if
you
go
in
the
moral
whisper
like
approach
where
there
will
be
many
notes,
relaying
messages,
how
do
you
prevent
spam
and
such
a
setting?
Because
that's
where
we're
yeah
we're
whisper-
brings
in
their
their
proof-of-work
right.
D
And
basically,
the
incentive
mechanisms
has
to
has
to
take
into
account
count
in
the
price
of
I,
saw
partial,
addressed
messages
that
that
that
you
have
you
have
flooding
which
is
basically
exponential,
so
that
the
in
the
in
the
what
for
303
vertical
there,
the
luminosity
radius,
if
the
luminosity
radius
is
is,
is
a
small.
So
then,
then
you
basically
give
just
a
small
portion
of
the
of
the
of
the
address.
Then
you
have
to
pay
so
one
one
step
in
the
in
the
luminous.
B
B
D
D
D
B
C
But
exposing
different
things
because
in
in
the
SS
at
the
moment,
so
you
can
set
the
message
and
you
it
can
have
some
luminosity
radius
based
on
the
hash
of
the
node,
so
know
the
dress,
right
or
recipients
of
restlessness,
with
whisper
those
beautiful
does
they
work
differently.
So
there
is
a
little
bit
more.
C
Flexibility
because
it's
a
big
thing
he
leaked
up
again
and
the
topic
this
is
something
like
this
is
topic-
is
four
bytes
of
arbitrary
data.
So
the
topic
is
completely
arbitrary.
It's
not
bound
to
address
or
whatever
else-
and
you
can
change
it
on
the
fly
which
is
not
possible
with
with
node
address
for
them
and
yeah.
So
it's
it's
a
bit
different
mechanism.
B
C
Basically
implemented
the
cold,
but
so
I've
written
most
of
the
code,
but
I
didn't
write
the
documentation
now,
the
guy
who
is
in
charge
of
whisper
at
the
moment
his
name
is
Guillaume,
so
apparently
he
is
working
on
on
documentation.
Now
I
was
talking
with
with
the
guys
in
in
Berlin,
and
they
say
yes
or
they
expect
is
the
connotation
to
be
ready
soon,
I
don't
know
so
I'm
I
I'm
working
in
swarm
team
now,
so
I'm
not
sure
how
it's
there.
Maybe
you
should
ask
him
directly.
C
D
Yes,
oh
so
the
actual
network
layer
protocol
is
similar
to
similar
because
it's
using
RFP
and
the
same
area
PX
based
that
peer-to-peer
protocol
here.
So
it's
similar
in
many
ways
to
the
current,
if
you're
in
protocols
so
a
DSN
and
if
but
yeah,
it's
it.
So
this
there's
no
no,
no
proper
documentation
for
that
level.
It's
I
mean
if
you,
if
you
using
go,
then
it's
very
very
easy
to
to
see
it
in
the
clinic.
D
D
B
For
us,
it
would
certainly
be
be
helpful
to
understand,
because
that
would
then
help
us
to
understand.
Okay,
what
sort
of
you
know
unanimity
guarantees,
especially
with
regards
to
metadata
privacy?
Do
we
actually
have
in
there
and
yeah?
So
what
be
actually
interesting
for
me
to
look
at
if
you
have
a
please,
please
feel
free
to
share.
Please.
D
D
A
B
A
E
B
Cool
okay,
because
that
goes
in
a
similar
direction
and
they
have
basically
kind
of
a
a
fork
of
swarm.
If
you
want
to
say
so,
and
they
add
some,
they
basically
have
a
have:
a
mechanism
for
detecting
kind
of
good
peers
and
and
not
so
good
peers,
and
then
they
have
some
staking
it's
it's
pretty
heavy.
What
they
do,
though,
so
they
basically
have
a
fork
of
the
manera
blockchain,
and
they
you
knew
yet
they
use
some
assets
on
the
Monaro
chain
in
order
to
a
incentivize.
B
A
The
same
I'm
draya's
not
here
because
it's
been
doing
some
some
prototyping
in
in
based
on
some
of
this
work
in
look
and
so
on.
So
you
could
talk
a
bit
more
about
that.
I
can
send
this
link.
I'm
gonna
talk
to
him
on
first
day,
and
then
we
can
invite
him
to
the
next
call
as
well.
I
may
be
in
sync
see
if
there's
any
overlap
is
on
yeah
I.
Think.
D
A
B
F
My
general
feeling
that
what
you
probably
want
there
is
the
data,
so
the
you
well,
okay,
there's
there's
a
couple
ways:
you
could
do
things
but
yeah.
So
the
main
constraint
is
that
we
don't
want
to
be
forwarding
around
the
the
actual
mixed
messages,
so
the
data
Sinclair
is
is
probably
dealing
much
with
something
much
closer
to
the
actual
content
to
the
messages
and
as
far
as
so
I
mean
one
design.
A
F
The
main
consequence
of
doing
that
is
going
to
be
I,
mean
I,
think
it's
perfectly
fine.
There
are
consequences,
but
there
so,
for
example,
you
know
if,
if
the
message
is
expired,
then
you
actually
have
to
get
it
from
some
other
participant.
You
can't
just
get
it
from
some
server
and
that
that
seems
perfectly
fine.
You
know
people
and
say
this
was
asked
for
in
a
recent
message:
I'm
sticking
it
into
my
message
and
you
know,
and
everybody's
okay
yeah.
That's
history,
I,
remember
that
history,
but
I
don't
need
it.
So.
A
I
think
one
way
to
think
about
is
in
terms
of
more
pluggable
pieces,
so
we
don't
build
them
on
leaf,
but
we
actually
build
separate
components
that
could
be
plugged
according
to
what
the
end
user
needs,
and
this
is
also
a
I-
think
it's
something
like
they're,
how
we
want
to
think
about
swarm
and
make
sense
and
so
on
how
how
the
interface
would
be
really
interesting,
because
this
is
trade-offs
and
different
strings
and
so
on.
So
how
do
I
make
things?
That's
maximally
pluggable,
and
and
so,
how
can
we?
F
A
F
D
The
same
problem,
but
so
so
PSS
allows
you
to
mount
that
peer-to-peer
protocol
over
it,
which
means
that
it
can.
It
can
kind
of
pretend
that
it
runs
a
TCP
like,
like
stream,
assumption
based
protocol
over
it
so
which
can
which
can
then,
then
we,
then
we
handled
SS.
Of
course,
any
protocol
like
we
did
with
the
decrypted
content
and
it
shields
shields
order
to
be
SS
traffic
and
everything,
and
also
also
mainframe.
A
B
D
F
F
Yes,
those
things
yes,
so
there's
whatever
yeah
the
double
ratchet
lets
you
keep
around
cryptographic
material
for
package,
yeah
I
think
it
might
and
that
that
handles
that
and
MLS
should
be
explicitly
I
mean
that's
one
of
its
main
things
is
making
that
work
in
a
group
context,
yeah
that
that
shouldn't
be
a
problem.
Sorry
I've
missed
some
contacts.
There.
A
I'm
curious,
a
fixture,
never
from
this
one
team:
how?
How
are
you
guys
thinking
about
like
mixing
the
sign
like?
Do
you
imagine
incorporating
some
lessons,
not
in
this
Swami
initial
design?
What
do
you
met
in
like
having
a
good
interface
where
you
would
use
a
mixer
to
then
access
for
more
like
how
do
you?
How
do
you
look
at
these
pieces?
From
your
point
of
view.
E
F
Someone
should
actually
be
polishing
some
kind
of
broadcast
design
they're
just
accepting
that
you're
not
going
to
be
using
it
for
messages
between
individuals
and
figuring
out
when
you
should
really
be
pushing
a
broadcast
type
thing
and
I
haven't
thought
about
that
really
at
all,
because
the
broadcasts
don't
provide
the
same
kind
of
privacy
day
assurances,
but
but
I
don't
think
it
it
doesn't.
You
know
it
it's.
It
does
make
sense
at
some
point
to
have
a
broadcast
layer
for
certain
kinds
of
things
and
I'm,
not
I,
don't
know
exactly
when.
F
C
F
F
What
I'm
saying
is
the
the
swarm
type
design
where
it's
basically
a
broadcast
system,
or
maybe
so
currently
swarm
as
a
broadcast
system.
I
guess
the
goal
is
actually
to
have
if
I
may
be
say
this,
as
goal
is
actually
to
have
the
swarm
things
be
more
modular
over
a
broadcast
design
or
some
more
targeted
mix
net
like
design.
Yes,.
A
Yes,
something
like
this,
but
possibly
not
having
pure
broadcast
just
in
terms
of
it's
kind
of
difficult
to
have
strong
reliability
guarantees.
So
so
it
might
be
that
you
have
it
closed
so-called
closed
group
like
more
of
a
group
like
multi
message,
multi
costing
group
messaging,
but
but
sort
of
with
different
guarantees
for
large
groups.
A
My
question
on
the
swim
stuff
was
just
in
terms
of
understanding
how
how
these,
how
them
sort
of
hopper
design
and
mixes
research,
how
it
plays
into
this
form
product
as
a
whole,
in
terms
of
both
PSS
and
feed,
feeds
and
swarm.
If
we
sort
of
what
we
see
as
the
most
useful
zuv
way
of
combining
these
or
not,
that
was
the
original
question.
B
Jeff,
could
you
maybe
comment
here
a
little
bit
on
these
under
you
mentioned
a
site
sentence
before
on
the
privacy
guarantees
of
mixed
nuts
versus
kind
of
this.
This
broadcast
scheme,
which
is
was
introduced
by
whisper
and
which,
in
a
slightly
different
form,
but
still
kind
of
similar,
as
used
by
swarm
today,
I.
F
Mean
you're
not
gonna,
get
much
privacy
guarantees
from
a
broadcast
system
at
the
end
of
the
day,
you're
not
gonna,
get
I
mean,
but
for
certain
kinds
of
messages.
Certainly,
there
will
be
system
related
messages
where
you
need
something
like
that,
and
there
will
be
other
kinds
of
messages.
You
know
we're.
F
Okay,
so
there's
another
thing
you
can
do,
which
is
you
can
send
a
message
through
the
mix
net
and
then
have
it
broadcast
via
via
PSS
or
whisper
or
whatever
yeah
so
and
then
the
recipients.
Well,
they
may
think
they're
getting
some
privacy
they're,
probably
not
getting
that
much
if
they
react
in
any
way,
if
they're
only
receiving
it
sort
of
passively
offline
and
not
even
showing
it
very
soon
to
the
user,
then
maybe
because
the
users
not
even
paying
attention
to
their
device.
F
Maybe
the
privacy
guarantees
better,
but
yeah
I
mean
I'm
starting
to
realize
I
mean
I.
We
may
still
need
something
like
that
for
some
kinds
of
things:
I
don't
know,
yeah
you
we're
not
gonna,
be
able
to
give
privacy
guarantees
there,
but
we're
gonna
be
able
to
say,
but
we
may
say
be
able
to
tell
applications
that
you
know
I
mean
we
may
be
able
to
tell
application
developers
that
hey.
If
your
thing
is
really
looks
like
a
broadcast,
then
there's
a
scheme
to
do
that.
That
is
somewhat
efficient
and
you
know
I
have.
A
A
question
on
so
we
have
surveillance,
you
guys
have
hopper
and
then
Harry's
not
here,
but
if
they
have
this
new
product
and
and
you
guess,
prevention,
you're
doing
their
sort
of
mix
and
research.
What
do
you
guys
see
us
often
most
if
this
is
gonna,
be
end
up
being
the
same
design?
And
if
so,
so?
What's
the
first
milestone
and
the
first
sort
of
useful
thing
where
there's
a
we.
F
Have
at
some
point
we
want
it
to
have
some
point
end
up
being
the
same
design.
We
yeah
basically
I,
need
to
spit
and
sit
down
with
Robert
and
get
both
of
us
onto
the
same
page
about
the
design
at
some
point.
So
will
Robert
and
I
at
some
point.
Well,
we'll
do
that
and
hopefully
come
up
with
designs
that
that
I,
like
a
lot
and
I'm
fairly
sure
I,
can
sell
everybody
from
NIMH
on
whatever
changes.
I
like
the
only
really
controversial
thing
will
be
the
validity
labs
incentivisation
scheme.
A
Look,
what
do
you?
What
do
you
think
are
the
things
that
are
not
up
in
the
air
like
what
are
the
things
that
we
it's
anything
that
we
can
roughly
finalize
already
simplices
of
the
faces
new
film
to
Quinto,
where
you
have
face
here
where
people
agree.
So,
for
example,
the
data
sing
stuff
is
something
we
think
we
confirm
eyes
over
the
next
few
months
as
on.
But
it's
something
similar
in
terms
of
simple
face:
stuff.
F
E
Comment
I
think
before
we
finalize
things,
we
have
to
have
a
clear
understanding
of
the
design
of
the
system
in
terms
of
routing
anonymous
routing,
but
I
think
that
is
still
up
in
the
air
and
I.
Think
if
you
we
don't
finalize
that
other
stuff
are
going
to
be
impacted,
then
we
have
to
change
them
again.
They
talk
yeah.
F
A
E
F
Yeah,
thank
you.
Yes,
so
more
of
the
problems
most
of
the
problems
that
the
mix
net
layer
are
actually
on
the
PKI
side
and
the
way
to
do
that
is
going
to
be
sort
of
precisely
as
you
suggest.
We
have
some
phases
where
we
say
we
implement
this
and
then
we
move
on
to
this
other
thing.
The
problem
is:
is
that
what
happens
on
the
PKS
side?
F
Is
we
very
quickly
move
from
what
we
sort
of
can
implement
and
then
we
we
somewhat
quickly
start
hitting
what
our
research
problems
and
we
don't
know,
actually
know
how
to
solve,
but
at
the
same
time
those
are
for
like
really
aggressive
scaling
things
and
I'm.
Actually,
I
have
been
making
some
progress
on
those,
though
whether
we
walk
the
things
I've
been
doing
or
not
is
complicated.
A
A
A
A
D
A
B
A
actually
a
short
question
of
of
talking
about
events
is
anybody
gonna
be
around
New
York
for
the
New
York
madness
week,
they're,
probably
maybe
date
when
is
it?
Second,
we
love
May
around
10
to
18th
of
May.
That's
also
8th,
New,
York
and
and
a
bunch
of
things
really
a
lot.
So
that's
why
I
will
I
will
be
around
there.
I
think.
B
F
For
the
call
guy
yeah
glad
to
know
about
the
secret
of
glad
to
know
this
one
meeting
so
I
guess
maybe
Fatima
or
one
of
us
should
be
thinking
about
going
to
this
as
well,
I'm,
not
sure
well,
wolf,
wolf
like
try
and
figure
that
out.
Maybe
I'll
just
ask
you
questions
about
it.
Ask
her
and
figure
that
out
yeah
sure
right.