►
From YouTube: Status Townhall #27 - February 4, 2019
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
B
Cool
yeah
thanks.
Yes,
those
are
basically
some
updates
on
my
side.
Unfortunately,
there
isn't
too
much
to
get
into
in
terms
of
the
project
future,
but
I
didn't
want
to
keep
you
guys
in
the
dark.
Carl
and
I've
been
pretty
busy,
basically
flying
around
and
meeting
with
different
potential
investors
and
other
ways
of
seeking
how
to
raise
funds
we're
considering
putting
more
of
our
own
personal
money
into
it.
We've
been
talking
about
doing
some
private
TC
trades
with
some
other
individuals
that
would
probably
come
out
of
colonize
team
for
allocation
in
S&C.
B
We've
also
been
revisiting
how
equity
maybaby
makes
sense,
and
if
it
does
make
sense,
then
that
also
might
be
a
potential
way
to
to
raise
more
funds.
It
didn't
make
sense
in
the
person,
but
you
know
it's
unlikely,
but
it's
it's
option
and
we
continue
to
define
other
other
so
that
having
news
and
that
sort
of
thing,
but
just
giving
you
an
idea
that
we
are
working
on
this
other
stuff
and
then
going
through
town
and
meetings
and
I.
B
B
But
basically
we
start
talking
about
means
that
people
can
put
the
tokens
up
for
things
and
provide
you
know
in
the
creation
or
like
hosting
of
nodes
or
hosting
of
the
network,
and
even
the
user
acquisition
engine,
the
user
acquisition
engine
as
something
we
haven't
talked
about
in
a
while.
It's
pretty
much
like
one
of
the
last
of
the
S&T
things
we
should
do
and
I'll
talk
about
that
in
a
video
in
a
moment,
yeah
next
one,
please
so.
B
Basically
the
reason
we're
focusing
on
on
all
the
S&T
white
paper
promises
is
because
that's
really
what
I
focus
should
have
always
been.
The
reason
we
haven't
is
because
scaling
seemed
a
lot
further
away
and
we
assumed
we
had
the
wrong
way
to
get
things
arrive
and
really
consider
these
things.
Of
course,
it
makes
a
lot
more
sense
for
us
to
refocus
now
under
the
current
market
conditions.
B
On
top
of
that
now,
the
other
one
is
the
user
acquisition
engine
and
no
I
basically
already
went
over
this.
But
all
the
S&T
use
cases
that
we're
doing
are
essentially
supportive
functions
of
the
acquisition
engine
and
yeah.
We
were
actually
working
on
this
last
because
we
need
scalability
for
this
to
really
work.
The
idea
behind
this
is
essentially
actually
I'll
go
ahead
into
that
next
slide,
please
yep,
so
Carl
and
I
had
before
we
started
status.
B
We
run
a
few
different
businesses
together,
but
one
of
the
the
more
interesting
ones
that
so
we
did
was
essentially
paid
user
acquisition.
So
the
idea
here
is
essentially
there's
a
publisher
who's
either
created
some
software
or
some
kind
of
service,
and
they
essentially
want
to
outsource
their
marketing.
So
they
find
people
who
have
a
distribution
network
or
capable
of
acquiring
installs
or
acquiring
a
cost
for
actions.
B
So
the
idea
is
that
they're
willing
to
pay
a
nominal
amount
say
like
anywhere
between,
like
20
cents,
to
$5,
for
an
install
or
for
an
action
and
they're
willing
to
pay
for
that,
because
they
expect
the
lifetime
value
of
that
acquisition
is
going
to
be
more
so
there
will
be
some
kind
of
margin
of
return
on
that
acquisition.
So
the
problem
we
identify
with
this
is
it
tends
to
lead
to
fraud.
B
Basically,
the
marketer
essentially
can
as
long
as
the
market
convinced
the
publisher
to
pay.
Then
they
they
win
all
right.
So
there's
these
problems
of
like
creating
like
fake
installs
and
there's
these
problems
of
like
yeah,
basically
fake
installs,
and
the
reason
this
comes
up
is
because
also
the
payment
cycles
are
a
bit
different.
So,
for
example,
marketers
might
be
expected
to
be
paid
net
thirty,
three
thirty
days,
whereas
the
lifetime
value
of
that
user
might
not
be
realized
entirely
until
like
90
days
or
even
even
longer.
B
So
that's
kind
of
interesting
about
this
is
we
can
actually
democratize
this
whole
model
and
we
can
kind
of
marry
the
idea
of
organic
growths
in
an
application
with
paid
growth,
and
we
do
this
by
essentially
allowing
users
themselves
to
get
rewarded
by
S&T
or
get
paid
for
onboarding,
their
friends
or
family
or
whoever
they
want
to
do
and
what's
how
this
essentially
works.
Well,
the
actual
details
of
this
will
work.
A
little
undefined
I
did
have.
B
B
There
is
a
margin
between
the
amounts
that
somebody
say
say
like
I
want
to
make.
Iii
have
you
as
a
friend
and
I
want
to
bring
you
into
status,
but
I
know
that
I
can
increase
the
value
of
your
yuan
jumping
into
status
by
giving
us
an
T,
because
that
means
or
s
or
any
other
token,
because
that
means
you're
more
likely
to
use
the
application.
You're
gonna
have
less
hurdles,
so
you
can
imagine
that
I
could
essentially
bring
you
into
into
inter
status.
B
I
tell
these
validators
and
the
people
who
are
player,
who
are
buying
the
installs
that
you're
worth
this
amount,
and
the
reason
in
addition
to
to
securing
my
bet
on
you.
I
can
also
send
you
funds,
because
it
works
out
that
you're
going
to
stay
around
the
application
a
lot
longer,
because
you
have
a
smaller
barrier
of
entry,
because
I
basically
gave
you
tokens
upfront
and
the
reason
I
do.
B
That
is
also
because
I'll
get
compensated
for
forgive
for
that
install
and
what's
also
kind
of
interesting
about
this
is
there
might
be
an
opportunity
to
actually
make
the
rewards
that
to
the
person
who's
staking
or
buying
these
these
installs
to
actually
mint
tokens.
So
it
could
be
inflationary,
which
is
kind
of
interesting,
because
in
that
means,
as
the
network
grows,
there's
going
to
be
more
tokens
to
to
service
the
entire
community.
B
But
then,
if
we
create
all
that
other
S&T
use
cases
to
be
deflationary,
then
you
can
imagine
the
use
of
the
actual
network
is
going
to
cause
the
the
the
token
the
individual
tokens
to
be
worth
more.
So
basically,
you
get
this
kind
of
really
interesting
sort
of
curve
where,
like
the
more
people,
are
on
board
the
more
tokens
there
are.
But
then,
if
it's,
if
growth
starts
stopping,
then
the
the
you
but
the
network
is
continued
to
be
used,
then
the
token
to
go
and
value
which
then
drive
further
growth
potentially
anyway.
So
so.
B
Sorry
engine
and
the
other
part
here
I
wanted
to
mention
the
the
points.
Eluding
me
that's
right,
what's
kind
of
interesting
about
this.
Is
that
then
this
drive
is
the
featured
development
of
status
itself,
because
people
are
going
to
only
want
to
bring
more
people
into
status
is
if
they
can
get
high
retention
rates
of
the
people
that
are
they're,
bringing
in
so
they're
more
profitable.
It
is
to
bring
people
into
the
status
then
the
more
people
are
going
to
do
it.
B
So
you
can
imagine
that
okay,
like
onboarding,
there's
going
to
be
after
a
really
really
tight.
The
way
of
retaining
users
within
the
application
has
to
be
is
gonna,
become
tighter
and
tighter,
and
a
lot
of
our
focus
will
end
up
for
be
put
on
to
this
and
the
other
reason
well.
The
other
issue
here
is
scalability
in
general.
B
B
So
another
thing
that
we've
been
in
a
bit
of
a
hot
topic:
you
know,
between
JB
Carey
in
the
bill
we've
been
trying
to
figure
out
ways
that
we
can
basically
do
performance
management
of
peer
review.
We've
been
thinking
about
this
for
a
while.
We've
had
a
few
different
discussions
come
up
and
discuss.
There's
this
idea.
You
know
there
was
ideas
of
stack
based,
ranking.
There
was
ideas
of
having
some
kind
of
leaderboards
based
on
whatever
metrics,
but
a
lot
of
the
feedback
hasn't
been
that
great.
B
But
what
did
kind
of
work
for
us
was
the
kudos
in
slack.
If
you're,
you
can
recall
that
so
we're
looking
to
essentially
do
a
similar
thing
like
that
we're,
but
with
a
difference
in
status
that
you
can
essentially
earn
SNT
from
people
giving
you
kudos
and
there's
going
to
be
a
discussed
post
coming
out
by
JP
and
Carrie
it'd
be
great
to
get
your
feedback
on
that
when
it
does
come
out,
I
started
writing
the
the
contract
based
on
some
of
their
specifications
on
the
weekend.
B
C
Super
cool
bring
back.
Who
knows
thanks,
Jared,
so
great,
to
know
what
we're
thinking
about
for
the
next
stage
of
the
project,
especially
given
the
fast
progress
that
the
teams
are
making
on
all
of
these
S&T
utility
features,
so
we're
gonna.
Do
a
quick
update
on
each
swarm.
I
think
first,
up
is
tribute
to
talk,
is
Eric
on
the
line.
D
Can
you
hear
me
yep,
okay,
so
yeah
so
far,
the
screens
for
sitting
UI
with
the
screen
for
setting
the
tribute
to
talk
in
the
UI
are
in
peer
review
and
soon
the
contracts
it's
gonna
be
connected
to
the
contracts
for
for
testing.
So
then,
what's
left
is
the
chat
part
of
the
flow.
So
let's
sit
for
tree
to
talk
itself.
E
So
we
made
the
grid
for
race
on
the
UI
implementation
and
essentially
all
the
flows
are
implemented.
What
we
did
recently
is
implementing
as
a
purchase
walkthrough
so
that
you
can
buy
and
install
new
sticker
packs
and
also
implemented
the
recent
stickers
features.
So
actually
there
is
some
small
video.
Maybe
it
could
be
interesting
to
to
start
a
start
it
right
now.
The
video.
E
Right,
okay,
so
yeah,
you
can
see
that
now
we
start
with
an
empty
panel,
then
you
can
browse
the
existing
packs
and
you
can
decide
to
install
whatever
backs
you
want
and
then
once
your
pack
is
installed,
then
you
will
get
access
to
the
sticker.
So
that
was
like
an
extra
item
on
the
bottom
part
and
you
can
then
trigger
whatever
sticker
that
you
want
to
send
now,
each
time
you
send
a
sticker,
it
will
appear
in
the
recent
panel.
E
But
then
we
also
worked
on
the
contract
and
it's
essentially
implemented
and
unfinished
so
yeah,
it's
just
a
matter
of
having
Ricardo
well
this
like
validate
that
the
contract
is
correct
and
publish
it
and
unwrap
stone,
and
then
we
will
be
able
to
start
experiment,
and
that's
essentially
that
the
the
last
step
that
we
have
is
to
well
first
make
sure
that
the
the
UI
is
completely
finished
and
then
to
integrate
with
the
contract.
Because,
right
now
everything
is
kind
of
up
curly
they're,
all
using
like
default
cat
packs.
E
But
we
want
to
make
sure
that
yeah,
of
course
it
has
to
work
with
the
real
contract.
So
that's
the
next
step
and
the
yeah
the
last
bit
is
that
where
we
are
in
contact
with
the
artist
to
create
sticker
packs
and
I
think
yeah.
So
here,
Richard
is
talking
with
different
artists
and
I.
Think
we're
pretty
mean
we're
making
great
fluorescent
to
to
have
those
nice
sticker
pretty
soon.
C
F
So
tell
Network:
we
make
a
very
good
progress
web
prototype
the
duck
we
have,
the
prototype
tab
deployed
to
rinkeby,
which
can
be
accessed
through
get
up
pages,
we're
currently,
basically
change
of
the
UI
of
the
tab
to
comply
with
the
design
that
were
provided
to
us
by
the
design
thing
and
we're
also
adapting
the
workflow
of
the
duct
work.
With
that
we
discussed
a
lot
a
specification
for
a
better
arbitration
protocol
and
one
of
the
requirements
is
a
communication
layer
that
that
we'll
need
and
so
for
for
diner
purpose.
F
We
we
made
murmur
web
compatible,
which
is
really
amazing
and
we'll
see
later
Natonal.
Richard
will
talk
about
this
and,
besides
all
that
there
was
lots
of
improvements
and
bug
fixes
here
and
here
and
there
and
it's
actually
ensuring
that
the
dump
is
solid
from
the
get-go
and
not
something
we
look
later
than
so.
It's
fully
tested.
We
make
sure
everything
is
acceptance,
tests,
internal
safety,
support
and
so
on,
and
you
can
access
that
link
to
the
app
as
well
as
the
storybook.
F
G
G
Right
yeah,
so
we
moved
from
for
these
two
weeks.
We
moved
from
like
nine
to
nineteen
percent
of
our
backlog
completion,
but
we
will
also
have
a
back
walk
refresh
meeting
this
this
week,
so
this
number
will
most
likely
change
so
yeah,
as
you
can
see.
Almost
you
see.
What's
epic
sign
progress.
These
are
mostly
similar
that
were
there
before,
but
yeah.
We
did
a
progress
in
each
of
them,
but
we
are
still
like
we
finding
some
issues
with
like
message:
reliability
for
instance
and
we'll
try
to
fix
them.
C
Nice
work,
thank
you,
and
so
next
up,
Oscar's
gonna
give
a
bit
of
an
update
on
the
protocol
work,
but
first
I
just
want
to
leave
time
for
Q&A
on
swarm
updates
and
it
looks
like
there's
a
question
in
here
from
yong-sik
yacek
asked
the
swarm
cost
so
far.
Why
is
it
not
in
'if
may
be
Nabil
or
jarred?
One
of
you
could
speak
to
that.
H
Yeah
right
now,
salaries
aren't
Paden.
Sorry,
look
at
salaries
right
now
paid
in
the
the
cost
is
in
US
v.
One
salaries
are
once
compensation
move
to
bet.
Then
it's
much
easier
tackle
a
death.
We
could
calculate
that
just
by
doing
a
conversion,
I
think
that's
preferred
I.
Just
think
right.
Now
it's
easier
people
to
to
grok
USD
than
death.
Considering
the
price
fluctuations
that
we
have.
B
C
Make
sense
thanks
guys
there
are
no
questions
from
YouTube.
So
if
there
are
no
other
comments
or
questions,
then
I
think
we
can
move
on
to
the
protocol
update
Oscar.
I
Yeah
sure
so
this
week
we
had
SS
with
last
week
we
had
their
workshop
in
Brussels
bunch
of
people
with
her,
yes,
ik
Corey,
Uhrich,
Amy
and
Anthony,
as
well
as
people
from
web
foundation,
about
it
laughs,
matrix
and
a
few
others
and
the
move
down
to
the
corner.
We
worked
on
a
secure
messaging
evaluation.
I
They'll
talk
a
bit
about
later
next
slide,
so
so,
first
of
all,
like
whoa,
what
is
cells
trying
to
do
and
we're
trying
to
create
the
kind
of
secure
communication
tool
that
that
pulls
human
rights
and
enabling
community
based
money
could
meet
this
law
and
so
through
privacy,
preserve
culture
and
empower
individuals
to
make
choices
and
sort
of
socially
coordinate
and
love.
This
is
sort
of
what
jad
alluded
to
in
the
beginning
as
well.
I
guess
next
slide.
I
I
So,
if
our
set
of
protocols
actually
do
because
we
want
to
be
secure,
you
need
to
be
able
to
reason
about
that
as
a
standalone
thing,
and
it
also
allows
other
people
to
look
into
it
and
analyze
it.
That
is
not
stuff
tied
to
in
specific
code
base,
and
this
was
it'll
be
cultural
shift
in
terms
of
thinking
about
the
protocol
as
its
own
being
and
having
it
be
language,
independent
and
clear
specification,
and
this
also
doesn't
just
touch
still
up
there.
I
Whisper
or
whatnot
it
sort
of
covers
the
whole
stack
and
it's
sort
of
multiple
layers-
and
this
includes
doesn't
a
use
case.
This
and
sort
of
everything
that
powers.
This
does
not
work
and
also,
if
clarifies,
are
all
a
bit
more
similar
to
the
film
foundation
and,
if
you're
room,
that
we
are
security
of
this
protocol,
that
runs
the
cells
network
and
we're
building
a
reference
implementation
and
in
a
way
you
can
say,
sir,
that
the
protocol
is
an
embodiment
of
our
principles
in
various
ways,
as
well
as
sort
of
thoughts
with
this
example.
I
So
when
talking
about
secure
messaging,
it's
useful
to
try
to
break
it
apart.
One
way
of
doing
this
is
in
terms
of
so
for
g''l
concerns,
and
this
is
this
a
few
ways
you
can
slice
this,
but
a
fairly
common
methodology
is
to
big,
apart
into
at
least
three
parts,
but
the
first
one
is
trust,
establishment,
that's
sort
of
about.
If
I'm
talking
to
you,
how
do
I
know
I'm
actually
talking
to
you
and
how
does
the
initial
key
chains
work
and
all
of
that?
I
The
second
part
is
sort
of
when
you
more
in
terms
of
safeguarding
messages
over
the
wire.
So
making
sure
mrs.
are
encrypted
in
various
ways
and
so
on,
and
then
you
actually
have
to
move
to
domestic
somewhere
and
in
a
common
sort
of
secure
security
goal
is
to
look
at
privacy,
assess
charity,
feature
and
and
to
a
large
extent,
hiding
metadata
such
as
who's
talking
to
who
and
so
on,
is
security
goal
and
that's
dealt
with
at
a
transport
price
layer
which
right
now
is
whisper
for
us.
And
finally,
you
can
also
look
at
this.
I
Maybe
a
bit
less
observes,
but
it's
also
because
peer-to-peer
messaging
is
fairly
new.
So
it's
not
as
well
studied,
but
you
can
look
at
the
peer-to-peer
or
section
overlay.
We
have
a
sort
of
a
way
to
achieve
since
resistance
and
robustness
and
continuous
and
so
on,
because
if
you
look
at
something
like
signal,
it
dates
they
still
have.
There
might
be
secure
messaging
in
some
ways
missionary
so
some
ways,
but
if
their
service
disappeared,
then
you
can't
talk
right
so
by
having
people
hosting
obstruction
by
themselves.
I
It
allows
they're
just
a
function
to
persist
throughout
time
and
it's
more
robust
to
bears
the
types
of
attacks.
That's
why
it
appears
extremely
important
to
us
and
it's
sort
of
its
own
category
anyway.
Next
up,
so
this
is
roughly
what
a
sex
looks
like
for
those
who
don't
know
at
the
bottom.
We
have
CP
protocol,
which
is
what
F
and
whisper
and
so
on,
runs
on
a
whisper
and
then
thought
about
that.
I
Next
up
so
starting
at
the
bottom,
we're
running
at
deputy
P-
and
it
has
sort
of
doesn't
have
to
go
into
too
much
detail
about
this,
but
it
has
sort
of
a
way
to
discover
nodes
deals
with
how
missus
look
over
the
wire
when
you
sign
them,
has
these
different
sub
protocols
and
like
eff
and
whisper
and
swarm,
and
it
flows
as
well
as
the
basic
session
management.
Next,
one.
I
I
This
will
be
bought
in
public
Wi-Fi
send,
and
this
would
have
no
path
to
things
like
mesh
networks
and
something
like
tor,
which
uses
pluggable
transports
to
masquerade
as
a
different
type
of
traffic
in
order
to
run
when,
when
your
blog.
So
that's
something
that's
very
useful
if
we
get
some
more
of
a
target
as
well
as
sort
of
to
distribute
hash
table,
that
it's
maintained
is
kind
of
expensive
and
slow
on
mobile
and
that
there's
a
lot
of
workarounds
that
we
do
with
that
in
terms
of
turning
on
and
off
for
service
protocol
and.
I
I
So
one
thing
is
that
the
anonymity
claims
are
not
very
rigorous
and
it's
not
something
that's
taken
seriously
by
by
researchers,
people
who
are
looking
into
private
or
serving
messaging
and
for
good
reasons,
and
that's
sort
of
both
for
the
sender
and
namida
side,
as
well
as
receive
an
image
size.
The
rules
of
no
machine,
readable,
spec
and
it's
incompatible
compatibility
is
between
gift
and
parity.
I
The
wispy
is
a
troupe
of
work
which
is
really
bad
for
hedge
against
Isis,
and
that's
something
that's
been
known
in
the
religious
literature
for
a
very
long
time
and
we're
not
even
using
it
so
we're
using
like
1%
of
work,
which
means
that
we're
not
part
of
the
larger
whisperer
network
that
barely
exists,
and
it's
also
sort
of
easy
to
spam
and
so
on.
There's
also
no
incentive
you're
on
a
node,
which
is
the
reason
that,
even
though
it's
poly
different
stack,
it's
not
something.
I
People
generally
allow,
and
finally,
this
extreme
scaling
difficulties
that
make
it
so
that
it's
it's
not
it's
not
very.
It's
simply
not
it's
simply
not
scalable.
It's
a
bit
naive.
You
know
it
does
things,
nothing
we're
storing,
notes
it
now,
even
though
we
don't
have
that
much
of
that
many
users.
So
that's
a
big
problem
mixed
up
in
terms
of
we
use
whisper
service
in
practice
like
we
are
we're
actually
running
like
a
centralized
cluster.
So
we're
not
that's
often
it
it
doesn't
work
right.
I
We
are
not
relying
relaying.
We
just
do
partial
relaying
on
mobile
phones,
whereas
in
order
to
have
the
security
and
ease
as
I
sort
of
specified
in
the
whispers
vilification,
you
need
to
relay
all
traffic,
but
that's
just
not
possible
on
the
mobile
device
and
actually
save
you
having
reasonable
bandwidth
and
battery
consumption.
It
won't
touch
them
to
proof-of-work
and
mail.
I
We
noticing
this
now
as
well
when
there's
a
mail
server
issues,
and
that
means
that
we
have
reliability.
I'll
talk
a
bit
more
about
that
later.
I
won't
go
into
topics
and
chat,
because
I
think
yeah
people
are
fairly
familiar
with
that
or
it's
not
super
relevant
to
people.
So
we
can
move
on
so
in
terms
of
this
sort
of
messaging
evaluation.
So
this
is.
This
is
an
excerpt
from
a
survey
paper
where
essentially
did
take.
This
is
specifically
about
Transport
privacy
and
a
way
to
think
about
that
and
think
about
this,
this
layer.
I
So
this
is
what
we've
also
done.
There's
a
huge
document
where
we
have
done:
corn
I
still
work
on
it.
We
still
need
to
finished
up,
but
we
have
essentially
this
type
of
evaluation,
but
for
whisper
as
well
as
for
how
we
do
trust
assessment
and
conversation
security
and
it's
it's
a
long-ass
document,
but
ideas
that
that
becomes
part
of
our
psmith
guarantees
and
the
things
that
we
know
that
we're
providing
and
we
can
communicate
it
clearly
and
and
not
just
wave
our
hands
and
say
well
whisper
is
dark.
I
So
we
trust,
establishment
and
much
time
to
have
right,
so
we
essentially
use
this
HT
keys,
which
has
whisper
and
wallet
keepers
and
then
in
terms
of
quantity
in
question,
its
kind
of
trust
and
first
use.
So
you
find
some
user
in
a
public
chat
or
you
message
them
or
you
might
use
a
QR
code
or
DNS
and
so
on.
But
we
don't
really
have
any
kind
of
verification.
It
happens
out
of
ban.
I
But
it's
something
that
we
should
be
more
rigorous
about
what
sort
of
counties
we
have
and
what
we
had
volleyball
when
it
comes
to
manual
attacks
impersonation,
is
on,
and
those
are
an
opportunity
here
in
terms
of
having
multiple
keys,
Ford,
apps
and
many
identities
and
so
on,
which
is
also
bigger
debate
in
the
film
community
of
that
privacy
going
on
moving
on
when
it
comes
to
come,
safe
security,
we're
not
doing
too
bad.
Actually
a
lot
of
pranks
to
love.
I
The
work
that
Andrea
and
others
have
been
doing
is
down
ratchet
this
sort
of
there's
not
a
lot
of
pieces
of
this
signal.
Open
whisper
system
stack
and
it
provides
us
with
things
like
perpetual
secrecy
and
integrity
of
message,
authentication
and
so
on.
There's
some
sort
of
small
things
that
we
still
have
to
figure
out
so,
for
example,
the
fact
that
perfect
world
secrecy
requires
back
and
forth
messages.
I
So
one
thing
working
on
is
sort
of
in
terms
of
introducing
a
data,
synchronization
layer
and
ideas
so
that
we
are
creative
application
and
we
don't
want
to
have
any
central
service
service
and
with
the
date
synchronism
synchronization
layer,
which
is
to
an
architecture
that
is
very
common
in
debate,
would
use
the
form
of
it.
It
uses
it
Breyer
and
so
on,
made
cheeks
as
well
scuttlebutt.
I
So
the
idea
is
that
you
have
sort
of
nodes
that
have
some
state
sort
of
keep
a
kind
of
log
of
events
that
happen,
and
then
they
synchronize
with
other
nodes
and
and
these
other
nodes
in
terms
of
the
messages
they
sort
of
have
some
to
communicate
message
dependency.
So
the
message
they
obscene
and
then
you
you
take
this
together
and
that's
all
you
sort
of
get
some
kind
of
history.
That
makes
sense
for
you
and.
I
The
idea
that,
by
having
this
type
of
states
interest
rates
in
later,
it
gives
us
more
options.
So
it's
kind
of
a
in
in
between
layer
where
you
have
the
state
successes,
synchronization
layer,
and
then
you
have
allows
you
to
sync
clients
on
top
of
this
which
semantics-
and
this
is
where
you
can
sort
of
go
wild
and
be
creative
about,
is
in
the
use
cases
and
tribute
talk
in
these
types
of
coordination
mechanisms,
because
this
provides
with
the
reliability
layer.
Anderson
deals
with
with
with
data
sync.
I
Essentially,
it
also
allows
us
to
swap
out
the
transport
underneath.
So
perhaps
we
aren't
gonna
use
whisper
forever
and
and
then
this
would
make
it
easier
to
do
that
type
of
thing
and
dealing
with
that,
as
opposed
to
being
pod
coded
whisper
and
tied
to
it.
And
it's
assuming
that
that's
the
underlying
layer,
so
it's
unreliable
delayed
and
offline
cells.
I
So
if
you
have
problems
with
mail
service
being
online,
all
the
time,
that's
hopefully
would
be
less
of
an
issue
because
it
deals
with
the
system
of
offline
files
and
it's
synchronized
this
one
since
online
and
the
sort
of
shoot,
peer-to-peer
communication
and
next
slide-
and
this
is
just
an
example
of
what
it
might
look
like.
So
you
could
just
sync
with
another
mobile
phone
and
in
order
to
sync
in
the
default
mode,
and
both
people
have
to
be
online
at
the
same
time.
I
Essentially,
next
one
so
in
terms
of
a
visualization
of
what
a
few
sec
might
look
like
talked
about
data
synchronization
layer,
adding
that
and
that's
likely
going
to
look
something
like
rambo
sync
protocol,
which
is
from
Beyer,
and
this
was
a
prototype
of
that
we'll
work
on
on
extending
that
and
then
I
guess
in
terms
of
layer
below
metadata
protection.
That's
something
that
might
be
something
like
a
mixing
based
approach,
and
that
was
also
largely
what
the
workshop
was
about.
I
So
it
makes
like
yeah
so
first
rough
to
use
priorities
when
introduced
this
date,
synchronization
layer
into
that
which
allows
us
to
experiment
with
transfers
below,
as
well
as
having
other
types
of
on
top
of
it.
It
shouldn't
require
many
changes
right
now
and
try
to
gracefully
introduce
it,
so
it
doesn't
disrupt
people
and
making
sure
it
actually
works
before
we
turned
on
killing
our
cluster
and
and
actually
making
sure
that
everything
works.
I
People
can
communicate
peer-to-peer
without
it
and
then
sort
of
working
on
creating
these
sort
of
sets
of
opus
over
specifications
with
these
counties
that
enables
multiple
implementations
of
such
plants
and,
as
part
of
this
is
also
stuff.
Fine,
finalizing
their
secure
messaging
evaluation,
I
mentioned
earlier
next
one.
I
So
for
the
workshop,
a
lot
of
it
was
based
on
on
mixed
networks,
and
mixed
networks
are
not
going
to
dive
into
it
in
a
detail,
but
essentially
a
rigorous
way
of
actually
ensuring
privacy
and
there's
been
lots
of
exciting
research
in
the
last
decade
or
so
that
makes
them
more
viable.
A
lot
of
them
are
not
based
on
peer
to
peer
Association,
architectures
and
that's
lower.
The
challenge
is
about,
and
we
had
lots
of
discussions
about
what
those
components
are
rough
areas.
I
What
solutions
might
look
like
considerations
like
how
you
deal
with
mobile
churn
and
how
to
pick
what
legacy
as
well
as
sort
of
accounting
for
resources,
sort
of
incentives
and
making
sure
these
are
sore
privacy-preserving
and
the
mechanics
make
sense
and
it's
efficient
and
so
on.
We
covered
a
lot
of
ground
and
we
there's
a
long
document
about
notes.
I
There
and
there
will,
you,
know,
sort
of
clean
up
and
do
a
collaborative
post,
which
the
other
participants
of
that
and
and
I
think
it's
worth
mentioning
that
this
is
very
much
a
long-term
effort
than
implies
Lourdes
almost
basic
kind
of
research.
So
it's
not
something
that
we're
going
to
introduce
in
time
soon,
but
it's
it's.
I
If
we
are
serious
about
privacy
and
actually
want
to
provide
these
countries
for
users,
it's
it's
it's
necessary,
and
this
is
something
that
people
hopefully
from
space,
but
also
the
web
free
general
space
are
excited
about
the
interested
in
so
we've
seen
lots
of
interest
and
sort
of
wanna
work
together
with
liquid
to
be
and
so
on
in
terms
of
building
out
this
protocol
and
yeah
it's
exciting.
So
that's
it
for
me.
D
And
so,
if
we
can
specifically
say
how
we
do
these
things,
whether
or
not
we
provide
them
the
guarantees
around
them
and
why
we
don't
provide
other
things.
Then
we
can
engage
the
academic
community
because
that's
kind
of
a
decent
standard
set
of
things
that
people
look
for
when
trying
to
evaluate
secure
messengers.
D
If
we
can
start
engaging
the
outside
academic,
community
and
security
researchers,
then
we
can
really
start
to
have
real,
strong
feelings
about
why
we
do
things
we
can
be
included
in
the
conversations
and
I
could
help
you
can
help
us
grow
quite
a
bit.
So
if
you
know
things,
please
help
me
an
Oscar
contribute
to
that
document
can
get
kind
of
a
better
outward
discussion
on
how
we
work
and
why
we
work
the
way
we
tale
and
then
fixing
things
that
don't
work.
H
D
H
J
Hello,
so
hello,
everyone
I'm
really
excited
to
present.
Today
what
we
have
been
working
on
for
the
last
few
weeks,
a
couple
of
months
wax
we
decided
to
work
on
a
state
to
GES
javascript
library
that
would
let
anyone
create
their
own
state
to
swine's
product
offer.
That
effort
we
build
to
node.js
application,
stay
to
X
and
stay
to
CAS
desktop.
The
next
step
was
to
make
a
state
to
use
JavaScript
best
available
inside
the
browser
which,
with
the
potential
increasing
adoption
that
this
would
bring.
J
However,
there
was
a
big
challenge:
whisper
availability,
because
most
what
three
providers
do
not
have
whisper,
enable
meaning
that
taps
are
not
able
to
make
use
of
this
protocol.
So
more
more
is
her
approach
to
solve
this
problem.
We
have
a
script
based
whisper
node
with
power
curve.
We
have
two
main
goals.
The
first
one
is
to
be
valid.
Web
tree
provider
fork
whisper,
that
any
idiot
can
connect
to
via
WebSockets
and
send
messages.
J
Basically,
you
can
run
a
more
more
node
and
have
access
to
whisper,
RPC
methods
and
broadcast
your
messages
be
at
that
p2p,
which
is
the
standard
interim
protocol
and
Lippi
to
P,
which
is
the
transfer
protocol
used
by
ipfs.
As
you
can
see
here
on
this
image
using
the
mobile
client
NPM
package
in
less
than
ten
lines
of
code,
you
can
start
and
use
a
whispered
out.
J
J
So
we
found
some
technical
challenge.
While
doing
this.
You
see
here
on
this
image.
They
output
of
Marmora
in
chrome,
console
well
up
until
last
week.
This
was
this
was
not
possible
since
more
more
was
a
node.js
application
with
no
browser
support.
We
successfully
made
more
more
ready
for
browsers
by
replacing
all
the
native
packages
for
pure
JavaScript
implementations
and
now
developers
are
able
to
import
the
NPM
package
and
Spain.
They
are
the
app
functionality.
J
That's
a
slide,
please!
So
another
technical
challenge
we
have
is
related
to
peer
discovery,
as
you
can
see
here
on
this
diagram.
We
we're
currently
using
web
RTC
with
a
signaling
server
client
will
connect
to
the
server
and
will
discover
new
peers.
This
is
similar
to
Italian
what
notes
this
works.
Put,
introduces
a
centralized
element
and
possible
point
of
failure.
We
are
researching
on
different
discovery
mechanisms
and
there
are
some
ideas
on
the
table
so
respect
to
solve
this
soon.
Next
slide.
J
So
a
bridge
are
no
tedious
application
that
can
act
as
a
whisper
node
as
well
notice
in
this
image.
How
there
are
listeners
for
both
protocols
and
the
peers
are
being
added.
This
appears
on
the
upper
side
of
the
screen.
Are
the
deputy
peers
on
below
you
can
see
a
liquid
to
peer
peer
being
added
next
slide,
please.
J
So.
Finally,
the
next
logical
step
in
our
quest
was
to
convert
the
state
to
CAS
the
application
to
the
web,
which
was
also
done
successfully.
So
we
now
have
a
web
page,
JavaScript
client
state
to
JavaScript
web.
This
doesn't
use
plugins
at
all.
It
doesn't
use
browser
extensions,
it
just
works.
You
can
play
with
it
in
the
URL.
J
You
see
in
the
screen
and
see
all
the
interesting
features
we
have
like
a
smiley,
select
or
being
able
being
able
to
send
images,
YouTube,
videos
and
all
kind
of
cool
stuff,
but
keep
in
mind
that,
but
murmur
and
status
are
angling
to
swipe
our
what
proof
of
concept,
so
it
is
suspected
to
find
bugs
due
to
their
experimental
nature
for
the
next
steps.
There
is
a
still
a
lot
of
stuff
that
needs
to
be
done,
like
retrieving
messages
from
the
server
storing
them
into
a
local
database,
as
well
as
making
things
more
efficient.
J
D
F
We
did
put
the
warning
that
if
we
put
like
a
very
bold
warning
that
there's
a
proof
of
concept
and
alpha
it
that
would
be
bugs,
and
if
someone
wants
something
more
stable,
they
should
use
status
text
or
app
and
the
you
know
the
stuff.
That
is
that
for
putting
the
key.
That's
just
for
that
purposes,
and
we
may
we
try
to
make
it
clear
in
adapt
that
that's
the
case.
Yeah.