►
From YouTube: Status Core Dev Meeting #4 - September 17, 2018
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
Bernard
livestreaming
full
item
all
right.
It's
it's
now
streaming
live
all
right.
Welcome
everyone
to
the
fourth
quorum
call
with
a
bunch
of
topics
to
go
through.
The
first
section
will
do
some
hot
topics,
which
are
things
that
people
are
explicitly
to
talk
about,
and
small
will
be
more
general
if
there
are
some
things
that
I
missed
or
maybe
just
minor
optics
and
just
a
reminder
to
try
to
keep
su
awesome.
A
All
right,
that's!
Okay!
Now
it's
gone
so
just
reminder
to
try
to
keep
it
to
things
that
are
relevant
to
at
least
a
sizable
minority
of
people
here
and
things
I'm
just
like
discussed
and
the
outcomes
that
are
a
future-oriented
s
person
just
giving
updates
about
fast.
With
that,
let's
start
with
item
number
one
which
is
whisper
key
and
other
status
reserved
derivation.
Perhaps.
A
To
focus
in
the
context
of
this
is
that
we
have
to
fix
the
relation,
perhaps
with
the
whisper
key,
and
they
never
keep.
It's
a
range
of
irritation
perhaps
were
reserved
for
other
stairs
used
in
the
water
in
the
future,
and
it's
a
proposal
post
and
discuss-
and
this
is
about
the
hot
wallet
and
the
goal
is
to
check
if
we
agree
with
the
proposal
in
order
to
proceed-
and
you
wanna
add
something
else
to
this.
That
may
be.
B
Yeah,
can
you
hear
me
yeah?
Okay,
good
I
was
saying
that
there
is
the
issue
of
not
only
about
who
we
choose
the
path.
There
is
also
the
issue
of
how
we
decide
to
implement
this
and
wrote
everything
in
the
discussed
topic
and
I
want
to
clarify
that
if
we
store
the
keys
in
completely
different
parts
instead
of
using
a
sibling
of
the
account,
but
then,
of
course,
we
need
to
modify
the
applet
to
do
more
te.
Caching
because
it's
impossible
from
a
performance
point
of
view
to
switch
from
from
completely
different
paths.
B
B
B
B
Memory,
or
is
it
also
problem
with
memory
or
other
stuff
in
their
hardware
the
to
having
in
the
cards
we
use?
It
is
possible
because
we
don't
use
all
the
memory
wheels.
We
use
a
small
part.
We
have.
We
can
implement
caching,
it's
just
more
complex
to
depend
on
people's.
We
will
have
to
detect
the
special
case,
so,
okay,
this
key
part,
is
important.
B
I
have
to
cache
it
separately
from
every
everything
else,
because
I
always
have
to
keep
it
and
then
I
have
the
account,
but
it
also
has
to
be
cashed
so
that
so
we
got
a
lot
of
cold,
a
lot
of
complexity
which
and
we're
not
complexity.
You
are
potentially
adding
some
security
issues
because
they
are
poor.
Their
pledge
is
not
as
readable
as
as
it
is
now
and
so
it's
possible,
but
is
we
have
to
deal
with
the
complexity?
B
D
B
The
only
modification
that
we
need
is
to
the
export
key
command,
because
the
export
government
now,
as
not
coded
list
of
keys
which
can
be
exported.
This
is
great
from
a
security
point
of
view,
because
you
cannot
export
anything
else,
but
if,
if
we
think
that
we
might
need
to
add
keys
to
this
list
in
the
future,
then
this
does
not
work,
because
the
applet
is
not
a
breathable,
and
so
then,
every
time
we
decide
to
add
the
new
key.
B
B
The
real
problem
is
that
if
we
use
the
cap
with
several
clients
that
every
client
they
maybe
they
might
not
agree
on,
what
is
the
the
list
of
parts
that
needs
to
be
exported,
so
is
the
little
bit
contest.
The
second
alternative
is
to
recognize
a
range
of
is
which
can
always
be
exported.
So
it's
my
proposal
with
Sylvie
row
traffics
to
the
to
the
key
index,
and
the
third
alternative
is
to
keep
it
are
coded.
B
B
C
A
E
C
A
So
we
have
a
we
limited
the
setting
around
the
proof
of
work,
which
is
what
we
is
in,
which
is
what
whispers
based
on,
but
it
makes
our
sort
of
mobile
notes
volume
both
to
denial
of
service
attacks
and
spam,
and
so
on,
and
some
questions
on
yeah
what
to
do
about
it
and
the
goal
around
this
is
to
spread
information
about
the
idea
as
it
touches
chats
and
also
collect
feedback.
Do
you
want
to
elaborate
another.
F
Yeah
item
that
so
so,
as
you
know,
we
have
this
peel.
The
abuse
approval
for
work
value
sets
to
super
low
number
on
mobile
devices,
with
creating
one
troop
of
work
and
whispers
like
the
first
lick
defense
against
the
normal
service
attack
or
spam
attacks
and
yeah.
We
cannot
really
use
it
because
when
you
send
a
message-
and
you
have
high
proof-of-work
value,
then
you
need
to
compute
a
lot
of
liquor
and
a
lot
of
computation
on
your
mobile
device
and
that,
of
course,
leads
to
battery
draining
battery
and
high
make
increase
your
usage.
F
So
that's
why
we
disabled
it
take
a
long
time
ago
and
that's
how
it
is
since,
since
they
can
over
a
year
and
and
yeah
like
doing
that,
put
our
cluster
and
also
mobile
nodes
in
some
kind
of
a
position
where
it
can
receive
a
lot
of
traffic
because
they
another
mean
that
that
can
reduce
traffic
is
blump,
blump
filters
and
blue
filters.
We
are
using
them
and
they're
cold,
but
yeah.
We
thought
that
we
need
someone
something
more
and
some
another
layer
of
Defense
and
here's
like
the
rate,
limiting
that
the
Dmitry
is
right.
F
Now
working
on,
so
it's
pretty
much
like
another
layer,
so
we
get
proof
of
work,
bloom
filters
and
now
we
can
rate
the
meeting
and
it's
just
gonna
like
drop
peers.
Who
that,
like
sent
you
the
most
messages,
also
what's
what's
very
interesting
for
for
people
from
working
on
the
chart.
Is
that
we'll
be
able
to
tell
you
until
they
tell
user,
which
chatroom
or
topic
uses
the
most
traffic?
F
So,
for
example,
if
you
joined
some
public
channel
and
it
just
just
like
a
lot
of
people's
chatting
there
and
you'd
like
to
remit
the
number
of
messages
that's
received,
number
of
the
traffic
you
can
just
you
could
just
unsubscribe
from
that
channel
for
a
bit.
For
example,
if
you
I
know
you're
abroad-
and
you
don't
want
to
use
your
data
plan
or
something
like
that,
so
I
think
that
could
be
super
cool
and
yeah.
F
F
D
Yeah
I
guess
something
else
that
came
up
from
F
Berlin
is
we
wanted
to
see
I
kept
asking
what
the
point
of
our
like
Alf
our
whisper,
can
scale
on
like
a
fundamental
level,
and
there
doesn't
seem
to
be
any
literature
involved
with
scaling
tests
on
how
how
far
whisper
can
go,
even
without
even
like,
with
or
without
bloom
filters,
with
or
without
group
of
work.
So
on
and
so
forth,
and
so
we've
started
up
a
I.
Don't
know
if
you
were
to
call
a
swarm
but
a
group
to
try
and
simulate
and
test
this.
D
So
we
can
have
hard
numbers
on
what
the
limits
of
whisper
can
actually
do
and
I
think
we
have
some
Jeremy
the
implementer
from
whisper
involved
in
this.
He
guard
knows
more
about
his
involvement
in
regards
to
this,
and
because
that
way
we
need
to
know
it
like
put.
The
limits
of
whisper
are,
which
are
basically
the
what
the
limits
of
our
communication
like
how
many
people
we
can
on
board
to
use
status
is
because,
if
that's
that
small,
we'll
have
to
start
figuring
out
a
way
to
either
make
that
larger
or
or
change.
A
E
D
Well,
the
mate,
I
guess.
The
main
motivation
for
this
that
I
wanted
to
make
clear
is
that
there
were
some
copy
that
wanted
to
go
out.
That
was
basically
pinning
us
to
whisper
like
we
are
special
because
we
we
use
whisper
and
I
wanted
to
move
away
from
that
a
little
bit,
because,
if
whisper
can't
scale,
then
that
did
and
we're
tied
to
it
I,
don't
that's
bad
for
us.
I
would
rather
say
these
are
our
principles
and
we
use
whisper
because
of
that.
D
F
Yeah
I
think
also
what
we
started
discussing.
Click
I,
guess
using
basil
ribbons
even
earlier
was
the
idea
that
we
would
like
to
build
a
protocol
that
work
well
kind
of
decentralized
peer-to-peer
layers,
but
not
necessarily
on
whispered
might
be
something
different,
PSS
or
whatever
so
yeah
I
I'm,
like
100
percent
or
not
it's
like
being
tied
to
whispered
in
anyway.
Then
it's
just
the
best.
What
what's
now
available,
but
it's
not
does
not
necessarily
mean
that
we
are
gonna
use
whisper
like
forever.
C
Yeah
and
so
yeah
Marcus
Marcus
started
documents
it's
in
the
channel
as
well.
That
basically
frames
the
issues.
There's
a
metaphor
in
the
whisper
series
channel
was
a
hack.
It
comes
in
document
which
is
beginning
of
introducing
those
ideas,
rather
than
the
whole.
The
three
document
series.
If
anybody
wants
take
a
look
at
that.
A
C
A
A
All
right,
cool
and
I
guess
we
can.
We
can
revisit
ists
at
the
end
because
it
wasn't
about
PSS,
mic
and
five,
but
before
that
we
have
also
paid
to
a
team
on
rethinking
mail,
so
it
with
long
based
communication,
I
guess
allowed
terms
of
educator.
I
just
want
to
make
sure
that
we
cover
items
listed.
So
we
don't
miss
anything
like
that.
Adam.
Do
you
want
to
go
through
that
yeah.
F
So
I
think
I
I
put
a
link
here
later,
but
on
discuss
there
was
a
couple
of
great
posts
that
started
Lee
and
also
slack
that
we
started
talking
about
flogged
based
communication.
So,
instead
of
like
spreading
that
you
sent
a
message
to
the
network
marketing
whisper,
it
would
work
this
way
that
you
append
your
messages
to
a
log
and
publish
this
log
at
some
intervals.
That's,
of
course,
is
from
there
like
communication
perspective
of
you.
It's
it's
increased
latency.
F
So
it's
not
like
super
real,
like
it's
basically
not
real-time
communication,
its
kind
of
mean
real-time,
something
like
that
where
you
need
to
synchronize
some
some
some
logs
and
that's
how
like
primary
it
works.
So
there's
like
there's
thing:
that's
called
skull
bot
and
it
pretty
much
worked
this
way
dressed
like
a
peer-to-peer
network.
F
F
How
would
it
work
so,
instead
of
having
main
servers,
each
peer
would
would
append
and
their
messages
to
a
log
and
publish
publish
them
like
periodically
to
swarm,
because
that's
that's,
probably
the
the
best
solution
in
this
dream
world
and
in
this
way,
other
peers
would
just
sing
those
logs.
Instead
of
asking
my
servers
for
historic
whisper
messages,
the
huge
point
here
is
that
we
would
remove
mail
servers
as
kind
of
third
parties.
F
That
right
now
are
necessary
for
history
for
collective
communication
like
offline
communication,
but
they
are
not
really
playing
well
with
the
centralization
principle,
where
you
want
your
mouse
to
be
self-sufficient
and
just
remove
all
third
parties.
So
so
the
idea
is
that
maybe
we
could
be
like
POC,
where
we
could
remove
my
servers
and
try
to
try
to
build
this
offline
communication
based
on
log
log
communication.
F
The
huge
point
right,
like
a
huge
drawback
here,
is
that
we
do
not
have
like
lights.
One
client
I
think
that
touches
like
one
more
topic
that
we
have
leg
below
about
lights:
warm
client,
but
that's,
that's
huge
missing
part
in
in
my
opinion,
and
until
we
have
it,
it
could
be
hard
to
achieve
other
goals
like
darkness
and
other
things.
So
that
would
probably
need
to
use
some
HTTP
warm
gateways
which
I
think
is
still
mad.
F
It
still
might
be
a
better
solution
using
my
server
so-
and
this
is
this-
is
somehow
that
kind
of
was
put
together
based
on
different
ideas,
but
we
haven't
really
started
working
on
that.
So
so
the
goal
here
is
to
like
what
like,
if
you
guys
have
like
some
some
feedback
about
it,
like
like
some
farts
and
and
yeah
like
pretty
much
figured
out
if
we
should
start
working
on
this
more
seriously
so
based
on
POC
and
whether
we
want
to
yeah
it
kind
of
depends,
for
example,
this
lights
one
clients.
D
With
the
origin
guys
at
etheral,
and
they
have
a
new
I
guess
they
used
whisper
but
only
I,
think
to
initialize
communication
or
they're
doing
something
with
a
community
counselor
called
they
wanted
us
to
look
at
that
they
just
released
on
their
platform.
You
initialize
it
like
the
initialization
step,
is
basically
signing
a
message
and
then
everything
else
is
off.
Blockchain
and
I
wasn't
sure
how
they're
relaying
the
message,
but
they
said
it
work.
It
worked
pretty
well
and.
F
Talk
about
origin
proto,
yes,
yes,
uh-huh
yeah
I
mean
from
what
I
what
I've
seen
like
I
seen
their
blog
post.
So
they
are
using
IP,
FS
and
yeah
they're
pretty
much
based
on
IP
FS,
but
that's
what
they
are
doing
is
I
guess
like
pretty
close,
so
you
log
based
communication,
so
they
are
actually
posting
this
message
to
ipfs
and
they
use
pops
up
from
my
PFS
to
distribute
the
information.
Something
changed.
So
so
yeah
I'm
not
sure
about
like
details
of
this
pops
up
and
ipfs
in
store.
F
But
there's
there's
this
sound
again
in
swarm.
There
is
mutable
resource
updates
that
that
that
guys
are
improving
right
now,
so
there's
open
for
requests.
That
might
be
great
for
that,
because
I
in
metal
resource
updates,
you
just
have
one
hash
and
you
just
update
word
points
to
like
date,
the
content,
but
the
heart
stays
the
same.
So
that's
pretty
cool.
Also
I
think
that
even
if
we
decide
to
put
this
whisper
would
still
would
be
used
for
like
reaching
the
like
the
low
intensity
and
and
discovery
of
other
peers,
and
so
on
so
forth.
F
So
I,
don't
think
like
logs
by
its
communication
at
this
point,
without
some
additional
features
of
swarm
or
mutant,
ipfs
could
could
support
like
real-time
communication.
It
can
be
useful
for
offline
communication,
but
still
something
like
whispers
needed
to
to
achieve
this.
This
idea
that
and
the
Technic
user
experience
that
you
communicate
with
others
in
real
time.
So.
A
Few
questions
one
have:
is
it
possible
like?
Can
we
create
like
a
generic
interface
for
single
access,
email,
server
or
swarm
or
ipfs,
or
anything
like
this,
and
then
we
could
even
have
like
it'd
be
redundant,
so
we
can
subtract
all
of
them,
but
they
sort
of
have
roughly
correspond
to
the
same
API
that
one
one
question.
A
second
question
is:
oh
yeah,
maybe
start
with
that.
One.
F
F
C
That
so
I
met
with
trend
carnegie
this
week
right
before
I
left
Berlin
and
what
digitalocean
is
planning
on
doing
is
load-balanced
their
digital
answering,
really
soon,
load,
balancing
across
AWS,
swarm
and
I've.
That's
where
you
can
basically
choose
on
an
algorithmic
basis,
how
the
load
balancing
is
done
across
traditional
infrastructure
and
then
swarm
an
IP
FS,
which
is
really
interested.
A
And
then
the
second
second
question:
so
if
we
were
to
move
to
something
like
swarm
like
how
do
we
have
a
and
4e
move,
you
skate
waves
because
right
now
mind?
Is
that
it's
not
you
can't
guarantee?
That's
actually
gonna
stay
online,
the
content,
so
that
could
be
like
a
pretty
big
breaking
news
experience
if
we
can't
guarantee
that
content
stays
online.
So
if
we
sort
of
moved
to
something
at
that,
considering
the
states
from
seeing
right
now
what
we
have
a
plan
for
how
to
deal
with
that
I.
F
Think
it's
pretty
similar
to
my
servers
right
now,
so
maybe
not
pre
similar,
but
for
my
service,
we,
like
we
run
this
on
our
DC
and
we
make
sure
that
the
data
are
not
gone
right,
but
this
is
pretty
pretty
poor
in
terms
of
like
the
centralization,
because
we
run
all
my
servers
and
now,
if
we
decide
that
we
want
other
people
to
run
my
servers,
we
have
no
really
means
to
check
it.
Does
my
servers
run
way?
Other
people
are
like
storing
all
the
messages,
so
it's
super
hard.
F
That's
why
initially
we
started
thinking
about
so
incentivization
and
validators
that
can
make
sure
that
that
what
my
servers
automates
every
store,
it's
actually
like
full
history
of
all
messages
and
not
like
some
just
you
know,
partial
partial,
historic
matters,
so
I
think
that
this
is
pretty
similar
to
this.
So
as
long
as
swarm
removes
data
like
the
protocol
level,
I
mean
I
think
it's
not
on
protocol
level.
Some
messages
can
just
disappear
because
no,
it's
like
no
one
really
stores
the
data
right.
That's
that's
the
I'm,
not
exactly
sure
about
this.
F
So,
for
example,
if
we
run
some
swarm
notes-
and
we
keep
this
data
this
just
for,
for
example,
let's
assume
that
we
can
decide
that
we
keep
data
for
those
nodes,
and
if
we
keep
this
data
around
I
guess
there
we
will
be
available
as
long
as
they
are
somewhere
stored
like
permanently
so
we
could
make.
This
is,
in
my
opinion,
it
seems
like
the
same.
The
same.
H
This
was
a
submission
to
consider
that
I've
seen
that
that
would
be
mentioned.
So
if
you
move
to
a
lock
based
sort
of
like
communication
protocol,
mind
it,
you
know,
like
you,
move
into
sort
of
like
it's.
It's
a
pool
model,
so
I
asked
for
what
I'm
interested
in
right,
yeah
so
currently
is
not
very
what
you
know.
The
scuttlebutt
was
meant
to
do
to
be
social
network.
We
in
works
fine
for
a
social
network,
because
you
follow
people
and
you
put
their
messages
for
a
chat
that
doesn't
really
work.
H
Well,
so
you
can
look
around
a
block
people
done
with
scatter
BOTS
and
they
try
to
with
charts,
and
none
of
them
was
very
successful
minded
you
know
like
as
well
to
provide
darkness
it's
very
difficult,
because
when
you're
boom
order
it
burden
of
discoverability
is
on
the
publisher
and
that's
what
Whisperer
protects,
mainly
the
publisher
and
what
also
the
receiver.
But
you
know
like
the
publishers,
primarily
it
you
know,
you
don't
know
where
the
message
originated
from
mm-hmm
with
them,
with
a
log
base
model,
it's
very
difficult
and
currently
like.
F
H
E
F
H
F
H
C
F
Yeah
I
think
I'm
gonna
describe
what
we
talked
so
far,
but
yeah.
We
we
also,
we
also
discount
this
solution
where
everyone
can
run
my
server
and
there
is
like
a
set
up
validators
that
tried
to
measure
all
the
peers,
like
the
main
servers
and
decide
which
of
them
is
keeping
all
the
data
as
it
should,
and
based
on
that,
you
could.
You
know
somehow
figured
out
which
node,
which
made
server
is
the
best
and
and
also
if
the
incentive
is
a
tional,
is
implement,
and
you,
of
course,
can
charge
a
bit
more.
F
A
To
me
to
me,
the
primary
problem
with
the
mail
service
is
just
the
fact
that
the
reason
you
want
to
have
one
of
the
reasons
you
won't
have
this
entire
system
is
because
you
want
to
have
availability
and
it
ends
up
days
and
with
mail.
Server
is
actually
worse
that,
but
I
would
be
with
the
more
centralized
set
up.
A
So
the
main
problem
is
just
the
fact
that
you're
tied
uniquely
to
one
mail
server,
so
if
we
can
just
find
a
way
to
spread
the
data,
so
it
doesn't
matter
which
males
have
you
connected
to,
but
they
provide
that
service
and
the
data
is
everywhere.
Then
maybe
you
could
still
have
that
kind
of
personification
display,
yeah.
F
Maybe
a
solution
is
to
actually
have
some
hybrid,
like
into
the
figure
out
some
hybrid
solution,
where
my
servers
could
use
pretty
much
swarm
or
IP
at
has
to
store
this
data
instead
of
instead
of
uploading.
This
responsibility
on
the
user's,
so
that
was
also
like
described
in
a
you
know,
one
of
the
discussed
post
that
I
posted
before
this
log
based
messaging.
F
That
was
pretty
much
like
like
something
like
this,
that
my
servers
would
still
be
running,
but
they
would
use
IP,
FS
or
swarm
as
a
storage
data,
and
there
would
be
set
of,
like
my
servers,
who
proposed
some
like
envelopes,
that
I
gonna
to
store
and
validator
that
do
that
and
kind
of
there's
like
one
index
of
messages,
all
of
all
whisper
messages
like
one
index
where
you
can
efficiently
search
for
them.
That's
also
like
super.
You
know
using
weekends
playing,
but
yeah
would.
A
It
be
possible
to
have
like
similar,
like
BitTorrent
in
some
ways
that
you
have
multiple
mail
servers,
that
all
that
all
stores
of
the
same
data
and
any
theories
every
several
times.
You
have
a
random
set
of
some
number
of
males
over
a
few
of
them,
and
then
whoever
gets
that
fast
as
they
that's
the
one
you
get.
So
it's
it's
not
tied
to
any
individual
machine
and
I
mean
denote
someone
that
does
be
reasonable.
I.
C
F
That
that
could
work
yeah.
So
in
this
idea
that
the
subscribed
on
discussed,
it
was
said
that
it
might
be
like
that
there
might
be
like
multiple
in
indices
of
many
competing
indices
of
master
or
historic
messages
stored
on
swarm
and
yeah.
It
would
be
they
would
compete
with
like
like
which,
which
log
is
the
most
like
it
like,
collects
the
most
messages
and
so
on
and
so
forth.
So
there's
also
such
an
idea
yeah.
It's
still
like
super
unknown
about
the
implementation,
but
but
yeah
yeah.
That
might
be.
That
might
be
something
something
possible.
F
E
Because
then
you
never
know
if
this
node
actually
posted
a
message
or
it
wrote
someone
else's
message
on
IP,
but
in
general
you
in
this
case,
like
you,
always
trade,
traffic
and
darkness
right,
because
you,
if
you
want
to
use
as
little
traffic
as
possible,
you
will
lose
darkness.
If
you
want
to
have
darkness,
then
you
will
use
much
more
traffic
like
x
times
more
traffic
than
otherwise,
and
it's
it's
a
good
trade-off
to
think
of
what
what
do
we
want
to
achieve
in
the
end?
E
D
This
is
something
that
came
up
that
I
had
a
question
about
in
terms
of
like
our
overall
goal.
Do
we
had
do
we
need
to
provide
full
darkness
by
default?
So
can
we
have
certain
modes
that
give
certain
types
of
communications
and
then
different
protocols
that
then
route
those
messages
appropriately
like
if,
if
99%
of
everyone's
traffic
or
communications
on
status,
it
doesn't
need
to
be
dark,
then
we
don't
need
to
use
a
completely
dark
protocol
to
do
it,
but
all
but
enable
completely
dark
messaging.
A
So
it's
a
good
question
and
we
also
need,
in
terms
of
thinking
about
darkness,
more
rigorously,
that's
something
we
haven't
done
really,
and
it
also
impact
things
like
push
notifications.
Why
are
you
actually
tying
to
specific
machines
together
and
it's
a
momentum,
L
server,
so
I
think
previously
we
have
said
that
will
be
alright
with
having
two
modes,
but
that
hasn't
been
sort
of
certain
one.
Yet,
but
one
way
of
simple
kind
of
problem
I
mean
like
they.
J
Did
yeah?
Maybe
if,
if
some
some
users
use
the
relaying
of
messages
in
some,
don't
how
someone
in
the
networking
can
figure
out
if
they
set
up
this
configuration
that
uses
more
data
or
not
so
maybe
just
by
having
the
option
enabled,
but
for
some
users,
then
someone
looking
the
network
cannot
know
if
they
are
using
or
not
this
feature,
because
all
the
messages
are
encrypted.
So
I.
J
A
Certain
representative
of
time,
just
on
the
topic
like,
should
we
start?
It
is
something
that
we
can
work
it
up,
working
on
a
proof
concept
regarding
this
of
mail
server,
log
based
Commission
and
thence.
They
come.
We
think,
moving
to
log
based
communication,
it's
like
a
completely
separate
thing,
or
is
it
something
that
we
can
do
gradually
as
part
of
making
sort
of
mail
service
better
in
some
way,
yeah.
F
H
H
F
On
so
from
my
perspective
that,
like
from
like
my
and
and
peer-to-peer
team
perspective,
that's
that's
what
we
like
to
do.
You
want
to
get
rid
of
get
to
your
main
servers,
because
it's
very
hard
to
surprise
and
you
need
to
have
some
validator
sensitization,
allow
you
to
actually
make
other
people
to
run
it.
F
A
We
it
gives
in
terms
of
those
apps
me,
can
we
store
it?
So
oneness
may
be
around
you
just
using
existing
interface
in
improving
the
backend
data
storage
thing,
because
I
think
that
was
also
one
of
the
pain
points
and
then
the
separate
discussion
around
sort
of
overall
log
based
and
these
things
and
removing
it
altogether,
because
we
still
have
constraints
that
Pedro
mention
on
push
navigation
provider
and
these
things
doing
it
sort
of
gradually
just
broke
a
plate.
Overall,
everything
we're
doing,
because
that's
a
logical,
no
discussion
board.
F
Yeah
I
think
like
if
we
maybe
try
right
now
to
to
solve
this
one
problem,
which
is
maybe
we
could
leave
main
servers
but
just
remove
like
switch
the
storage
layer,
so
maybe
try
to
use
swarm
or
IP
at
first
for
my
service
as
a
storage
layer
and
maybe
make
those
many
servers
to
work
together
on
a
single
offline
history
of
whisper
message.
That
could
be
super
cool
and
that
might
solve
like
huge
part
of
the
problem.
So.
I
K
Just
mentioned
two
things:
one
is
Andre:
I
started
actually
trying
to
make
the
mail
server
work
with
storing
its
database
in
swarm.
We
were
in
Ghana,
so
I
don't
know
how
far
he
got
with
that,
but
it
might
be
interesting
to
investigate
the
other
thing.
I
wanted
to
say,
though,
is
that,
like
log
based
communication,
it's
not
like
we
do
what
we
keep
doing,
what
we're
doing,
and
then
we
just
throw
a
log
in
there.
It's
a
ground
up.
K
It's
it's
a
completely
different
way
of
looking
at
the
problem
from
the
ground
up
and
the
most
important
thing
about.
It
is
not
whether
the
data
goes
into
swarm
or
not.
The
most
important
thing
about
it
is
that
all
the
clients
are
equal,
they
all
or
you
store
your
own
log
on
your
device
or
on
somewhere.
K
First,
that's
like
the
first
principle
according
to
which
these
are
crates,
and
then,
on
top
of
that,
in
order
to
make
communication
easier,
you
might
add
something
like
swarm
or,
you
might
add
something
like
ID
FS
or
you
might
add
something
like
mesh
network
communications
through
Bluetooth
or
whatever
it
might
be.
But
all
these
are
sort
of
add-ons
in
the
system
that
you
that
you
throw
in
there
so
said
said
so
it's
good
to
think
about
it.
This
way,
right
if
we
want
to
go
down
that
route,
it's
it's
also
from
a
product
perspective.
K
And
what
happens
with
those
systems
is
basically
that
you
can
get
to
normal
latency
of
a
messenger
bar
by
doing
tricks,
and
then
it
degrades
gracefully
as
your
connectivity
options
degrade.
So
if
you're
not
able
to
get
in
touch
with
one
of
those
helper
servers
and
then
it's
a
little
bit
slower,
but
if
that
works,
then
you
can
make
it
more
or
less
as
fastest
as
whisper.
K
H
Yeah
I
feel
latency
is
actually
I
care.
Yeah
I
agree
with
you.
A
thing
later
see
that
you
can
get
it
work,
I'm
fine.
They
are
the
problem
with
the
love
based
communication.
Then
it
fails
to
be
a
cool
model.
So
if
you
look
at
the
track
implementation
they
ten
twelve,
you
know
it's
difficult
to
discover.
New
contact
now
discovering
your
contact,
but
sort
of
like
I,
won't
communicate
with
that
user
and
that
user
doesn't
know
me.
A
This
someone
who
wants
to
dive
deeper
into
debts
and
maybe
write
AB,
some
idea
of
like
what
trade-offs
are
and
what
the
first
debts
of
this
would
be
I.
Don't
just
think
it's
already
done
like
lots
of
legwork
this
one,
but
just
a
water
pour
yeah
in
terms
of
moving
this
forward
and
seeing
what
we
need
to
do
on
a
product
level
and
so
on.
If
that
business,
which
we
want
to
make
in
terms
of
how
we
fundamentally
think
about.
J
I
always
see
the
major
cyber
problem
as
being
read
that
you
use
swarm
to
host
the
message,
but
it
may
serve,
it
would
look
like
more
just
notification
servers
but
of
course,
then
music
will
have
to
download
from
swarm.
Maybe
they
would
have
to
pay
some
tokens
I,
don't
know
how
the
situation
the
station
will
be
with
swarm.
A
F
Mean
it's
not
super
important.
We
can
skip
it.
I
didn't
just
maybe
without
discussion,
so
I
just
can
I
can
say
that
we
still
kind
of
working
on
a
solution.
How
to
convince
like
how
to
allow
more
people
to
run
status
knows
right
now.
We
want
to
go
with
documentation
first,
because
we
currently
do
not
have
any
documentation.
If
someone
like
to
run
status
stuff
it
just
it
just
it's
hard
to
figure
it
out
without
asking
us
it's
so
we
start
with
documentation.
F
We
have
some
some
contacts
to
guys
bleeding
mesh
box
the
where
we
could
run
status.
So
that's
a
little
another
ping
and
as
I
mentioned,
for
example,
my
first
we
tried
to
figure
out
how
to
do
to
solve
this
whole.
My
server
issue
and
right
now
we're
status
desktop
is
like
kind
of
purity
that
might
also
be
very
interesting
in
terms
of
headings
just
nodes
around
the
globe.
F
That's
for,
like
past
whisper
messages
around,
so
we,
if
we,
for
example,
get
to
the
point
when
there's
like
being
on
hundreds
of
subtle
desktops
running
around
the
globe.
Maybe
we
don't
have
our
whisper
notes
at
all.
In
our
see
because
they
could
handle
the
traffic,
it's
still
unknown,
but
that's
maybe
their
feature
so
so
yeah.
That
is
the
photo.
F
Mean
yeah,
if
you
yeah
I'm
yeah,
if
each
node
could
start
it
that
that
so
it's
like
even
like
another
step
but
but
yeah,
and
that
would
be
super
crazy
and
also
super
great,
but
even
like
getting
creed
of
our
whisper
knows
that
we
need
to
keep
around,
for
example,
and
just
rely
on
status
desktop.
That
would
be
super
nice
super
great,
but
that
requires
a
lot
of
status,
desktop
instances
running
around
the
globe
right,
but
I,
guess
that's
the
goal
right
here.
We're
building
this
hoping
well.
F
Not
really
yeah,
it's
also
on
our
other.
It
just
like
step
by
step.
Just
and
maybe
I
can
talk
to
you
I
cook,
to
maybe
get
someone
else
who
could
do
that,
but
yeah
I
just
wanted
to
focus
on
the
docs,
because
that's
the
missing
part.
So,
even
if
we
get
it
to
the
up
node,
we
still
would
there
would
be
no
documentation
for
like
people
who
would
like
to
run
it
on
on
digital
ocean
or
whatever
I
thought
that
that's
there
at
first.
F
C
A
C
A
A
L
E
K
K
I
Now
they
want,
they
want
our
help,
but
you
know
for
us:
it's
actually
much
harder
to
to
be
in
to
provide
some.
You
know
initiative
here,
because
we
we
don't
understand,
swarm
the
way
they
understand
bike
or
developers,
so
I'm
kind
of
I'm
trying
to
keep
you
know,
keep
reading
their
conversations
and
follow
their
work
to
find
like
where,
where
the
best
the
best
place,
we
can
jump
in
with
development,
but
I
think
at
least
for
me.
A
So,
in
a
guess
in
terms
of
what
Eric's
initial
point
it
seems
like,
maybe
we
don't
have
resources
to
develop
like
a
gym.
It's
for
my
client
right
now,
but
I
might
change,
but
but
more
immediately.
Like
Jeff's
point,
we
can
start
experimenting
with
on
desktop
already.
Do
people
think
that
seems
like
a
reasonable
thing
to
do
and
and
many
folks
from
around
how
we
would
go
about
that.
So
we
don't
break
existing
things
and,
and
all
of
this.
I
A
L
There
is
really
three
different
things,
so
the
p2p
thing
is
this:
ongoing
effort
in
a
theorem
go
to
now
for
a
theorem
to
switch
from
P
to
P.
2
will
appeal
to
P
and
whisper.
Is
the
test
subject
for
it?
So
we're
just
so
I
think
that
we
can
leave
aside
for
now
and
see
how
Guillaume
is
doing
with
with
it.
Maybe
we
will
be
asked
to
try
it
out
on
one
of
our
test
fleet
at
some
point,
then
the
second
thing
thing
is
swarm
itself,
so
I
guess.
A
A
H
D
Yeah
in
terms
of
the
breaking
whisper
stuff,
anyone
who's
interested
in
trying
to
help
figure
out
the
necessary
metrics
needed
to
to
test
scalability
of
whisper.
Please
join
the
channel
and
help
out
and
there's
a
there's,
a
living
document
there
that
it's
listing
the
types
of
things
we
need
from
the
simulator,
as
well
as
the
types
of
tests
we're
going
to
have
and
how
I
like
how
we're
going
to
test
this
stuff,
and
so
we
get
I'd
like
to
get
that
underway
as
fast
as
possible.
So
that'll
help
us
make
these
types
of
decisions.
C
A
I
Yeah
I
actually
wanted
to
briefly
discuss
area
for
a
producible
builds
for
status
go,
but
I
can
give
the
higher
level
overview
for
the
whole
problem.
So
there
is
that
we
want
to
have
reproducibility
in
our
builds,
so
the
path
from
the
source
to
the
binary
for
users
is
kind
of
reproducible
and
verifiable.
It's
kind
of
two
separate
problem,
reproducibility
and
verifiability,
and
we're
trying
to
solve
first
one.
There
are
a
few
reasons
for
that
security.
One
of
them-
and
we
also
this-
is
requirement
for
I,
think
f-droid
project.
I
I
Can
you
can
you
see
them?
Can
you
see
the
my
scream
yeah?
Okay,
so
this
is.
This
is
actually
very
specific
topic
for
for
Google
developers,
so
we
I
will
be
as
quick
as
possible.
It's
describing
the
issue
in
issue
thousand
hundred
eighty
five,
so
the
problem
is
that
goal
gives
us
reproducibility
almost
out
of
the
box.
So
most
of
the
problems
already
solved
you
can
build
if
you
have
the
same
go
version
and
dependencies
and
everything
it
gives
the
same
binary
almost
the
same.
I
If
you,
if
you
happen
to
build
in
a
different,
absolute
source
path
directory,
which
is
probably
true
for
most
users,
the
binary
will
be
different
and-
and
that
was
the
problem
and
we
could
solve
it
by
Joker.
This
is
our
current
thinking,
which
will
be
the
best
solution
for
that.
But
doctor
feels
too
heavy,
for
you
know
just
faking
their
path
right.
So
I
was
thinking
about
change
route.
I
Some
highlighter
I,
can't
organization
like
Alexei's
may
be
using
LD
pre-war
talks
like
your
septum
gets
in
the
syscall,
etc,
etc,
but
they
too
high
can
require
route.
So
I
took
I,
took
a
closer
look
to
the
binaries
and
actually
they
differ
in,
like
30
bytes
only
I
mean
the
binaries
built
with
the
same
set
of
tooling,
but
in
the
different
directories
and
I
noticed
that
the
difference
is
the
built
ID
value.
The
lady
is
is
60
60
bytes
that
go
stamp
each
binary
and
exist
from
super
store.
I
I
In
inputs
are
different,
but
we
actually
care
about
outputs,
so
my
thinking
was
maybe
we
can
just
on
the
build
on
like
when
you
run
make
go
status,
go
release.
We
can
analyze
analyze.
If,
if
the
output
is
correct
from
build
ID
and
if
it's
true,
so
we
can
just
replace
replace
the
whole
like
overwrite
the
binary.
With
with
this
simple
comment,
it's
a
bit
hockey
approach,
but
it
results
in
the
byte
by
byte,
invite
equal,
binaries
and
I.
I
Don't
know
I
like
it
I
actually,
and
it's
actually
like
saying
hey,
we
don't
care
what
your
import
path
for.
Yes,
your
absolute
source
directory
is,
but
as
long
as
go
to
and
provide
the
same
binary,
the
same
output.
We
consider
this
exactly
the
same,
so
I
posted
I
posted
the
issue
on
in
the
status
go
rapport
with.
You
can
read
for
more
details,
and
if
you
have
some
questions
towards
this,
let's
talk
I
mean
I.
I
just
want
to
hear
that
their
throats
and
and
feedback
on
this
okay
yeah,
that's.
It
sounds.
C
D
C
I
So
this
content,
ID
for
binary,
will
be
different
between,
for
example,
Mac
OS
builds
and
Linux
builds,
but
this
one
third
third
value
will
be
equal,
even
if
one
user
builds
in
the
Linux
and
other
user
build
builds
on
the
Mac,
OS
or
Windows,
and
this
is
really
cool,
because
we
can
actually
verify
if
it's
the
same
binary,
even
in
cross-platform
fashion.
We
don't
need
this
probably,
but
this
still
very
cool
I
mean
it's
better
than
checking
the
hash
of
the
binary
right.
H
I
D
A
key
point
that
I
want
to
make
sure
is
that
this
needs
to
be
incredibly
easy
for
people
to
do,
because
they're
bare,
like
most
people,
are,
have
a
hard
time,
just
checking
a
hash.
So
if
we
need
to
be
able
to
publish
something
and
allow
people
to
check
that
easily
when
they
build
it
themselves,
make
sure
that
gets
what
we
publish.
If
that's
not
an
easy
process,
and
it's
it's,
it's
not
gonna.
It's
not
gonna
work.
I
Yeah,
you
know,
but,
as
I
said,
the
major
problem
is,
reproducible
builds
for
us
is
the
status
riad,
and
we
had
a
call
on
this
two
days
ago
and
we
agreed
that
the
biggest
problem
is
that
we
probably
don't
have
anyone
who
truly
understand
the
whole
complex
series
of
the
status
really
build
pipeline.
So
it's
like
four
or
five
different
stack
technologies
with
different
sets
of
dependencies
of
tooling,
and
it's
many
of
them
apart
from
perfect
and
so
yeah
I
I
think
we
agree
that
we
need
to
break
down
this.
A
So
yeah
yeah,
obvious
victory,
serves
like
one
one
thing.
This
initial
step
is
to
just
on
one
system,
run
the
entire
paper
lines
of
printout
or
like
the
file
tree
structure
with
MD
Phibes
whatever
and
then
do
the
same
thing
again
and
there
see
which
kind
of
folders
are
moving
out
around
a
lot
and
then
sort
of
start
from
there,
and
you
can
probably
get
like
some
parallel
work
in
terms
of
well.
A
Here
are
five
libraries
that
have
some
kind
of
time
thing,
or
these
seem
to
be
in
deterministic
in
their
order
because
of
node
whatever.
So
that's
probably
one
way
of
doing
it,
because
if
you
can
get
to
the
state
where
you
can
run
this
same
thing
on
the
same
machine
and
it
produces
the
same
sort
of
merkel
route
or
whatever
or
folder
structure,
that's
a
pretty
solid
start,
and
then
you
probably
need
to
go
deeper
than
that.
A
C
I
I
The
project
I
started
to
work
in
April,
because
the
my
main
goal
was
to
be
able
to
simulate
whisper
behavior
on
different
networks
play
with
these
networks
and
visualize
it
to
you
know
to
have
intuition
about
how
it
behaves,
because
we
have
very
pure
intuition,
with
pure
pure
and,
and
we
almost
have
no
data,
not
no
actual
data.
So
the
project
that
she
currently
is
just
the
set
of
three
three
repositories:
three
comments
and
they
buggy
they.
You
know
I,
just
I'm,
only
user
of
this,
so
you
needed
like
instructions
to
make
them
work.
I
You
need
to
copy
output
of
one
program
to
another
program,
but
in
in
the
essence,
it's
the
generator
of
network
topologies
different
random
grabs
generators.
It's
a
simulator,
the
trance
runs
real
whisperer
v6
code
on
top
of
this
network
topology
using
the
p2p
simulation
framework
from
Goa
tyrion
created
by
swarm
team
by
the
way
and
and
then
it
provides
well.
I
This
is
enough
to
to
get
some
data
from
this
simulation,
but
it
also-
and
the
third
project
is
the
vision
3d
visualization
of
this,
and
so
probably
many
of
you
have
seen
already
so
it
looks
like
this.
This
is
the
nether
I
think
it's
three
hundred
networks
and
in
red
you
see
how
message
is
propagated
so
and
I
started
to
work
on
this
visualization,
but
they
will.
There
were
two
problems
with
that.
First
of
all,
there
was
not
enough
interest,
so
I
always
had
other
tasks
to
do
so.
I
It
kind
of
stalled
the
project
and-
and
the
second
one
is-
is
a
JavaScript
and
I'm
not
trying
to
start
language
words
right,
but
it
was
easy
for
me.
I
wrote
a
few
huge
project
projects
in
JavaScript
and
frameworks,
but
I
still
can
write,
can't
write
the
like
testable
maintainable
JavaScript
code,
so
it
was
hindering
me
from
from
advancing
so
what
I
started
doing
now.
This
is,
this
is
sounds
crazy,
probably,
but
I
started
rewriting
front-end
in
bhakti.
It's
a
gopher
JS
based
framework,
it's
similar
to
reality.
I
Button
go
and
govern,
J's
go
to
JavaScript,
transpiler
I
know
this
is
sounds
stupid
and
crazy,
and
we
probably
should
have
shouldn't
find
some
experienced
JavaScript
developers
and
if
you,
if
you
are
one
of
them,
just
join
me
if
you're
interested,
but
actually
I,
was
surprised
how
cool
this.
So
this
is
a
example.
What
I
have
now-
and
it
just
spent
a
few
days
on
this,
so
I
I've
been
able
to
add
you
know.
A
lot
of
configuration.
I
can
now
applaud
this
and
run
actually
physics
algorithm.
I
This
is
this
graph
is
based
on
force-directed,
algorithm
and
I
did
and
I
able
to
run
this
in
the
browser.
I
can't
imagine
right
and
this
myself
in
JavaScript,
so
anyway,
I
mean
this
allows
me
to
advance
much
quicker
than
with
JavaScript,
and
so
there
is
not
a
single
line
of
code
in
JavaScript
in
this
front-end.
So
this
is
really
cool
I'm
doing
in
the
framework,
and
it's
and
it's
readable,
it's
actually.
I
It's
built
it's
in
in
in
some
parts
it's
very
similar
to
the
JavaScript,
because
sometimes
you
just
call
raw
JavaScript,
but
you
know
it's
readable.
It's
extensible!
You
can
use
all
the
power
of
go
tooling
for
that,
so
I'm
going
to
continue.
Work
on
this
and
yeah
join
us
in
breaking
whisper
channel.
If
you
interested
and
if
again,
if
you
have
use
cases
for
these
simulations
for
visualizations,
please
talk
to
me
because
and
just
doing
what
what
I
want
myself.
I
But
if
you,
if
you,
if
I
get
more
inputs
that
will
help
to
to
build
a
proper
tooling
that
other
people
can
use
to
verify
ideas
and
one
thing
that
I
really
want
to
work
on
as
well
on
the
basis
of
this
to
win
is
simulation
and
visualization
of
different
discovery
protocols.
So
we
we
right
now
using
two
protocols
disk
we
five
and
ran
around
the
pool
and
I
mean
I
read
the
code.
I
read
the
specs,
but
it's
still
like
mostly
the
black
box
for
me.
I
C
A
C
Actually,
that's
a
specific
exit
from
continue
working
on
desktop,
optimization
it
will,
during
this
week
and
for
blocking
part
from
North.
We
are
almost
there
and
there
are
PRS
which
emerged
without
testing.
We
don't
want
this
to
be
so.
We
will
introduce
him
this
one.
A
street
like
100%
pass
tests
for
everything
can
will
block
each
PR
from
merge.
If
they're,
not
the
automated
test
results.
Just
that's
it
from
our
side.
G
G
As
a
bicycle
as
a
substitution
for
mail
server
notes-
or
something
like
that-
and
this
goes
quite
together
with
some
performance
and
security
issues
that
we
still
have
for,
for
instance,
currently,
application
runs
two
processes.
One
is
the
Qt
native
application.
Another
one
is
a
node.js
server
process.
They
communicate
the
airport.
G
There
was
no
encryption
on
that
channel,
so
this
potentially
could
be
hijacked,
and
this
I
think
is
a
bit
of
a
problem
that
we
should
focus
on,
even
given
that
one
of
the
first
item
in
our
principles
that
we've
all
scientists
that
we're
not
compromised
on
security.
So
this
is
important
and
also
what
the
regards
to
performance.
We
are
having
some
issues
that
we'll
need
to
fix
and
I'm
thinking.
G
Some
of
these
items
are
believe
in
Oh
max
is
that
is
the
Guru
here,
so
he's
been
trying
experimenting
with
them.
Embedding
that
GS
server
process
into
the
QT
application
so
that
we
get
rid
of
this
double
process
conundrum
and
therefore
increase
our
security.
Somehow
this
would
also
help
us
with
performance,
because
then
we
would
not
have
an
inter
process
communication
pipeline
working
each
time
you
get
some
UI
updates
or
vice-versa,
but
anyway
there
will
be
still
some
more.
G
So
what
would
your
thoughts
be
like
at
least
my
line
of
thinking
is
that
we
would
definitely
need
to
work
hard
on
improving
performance
and
also
the
security
point,
but
whether
this
is
even
then
whether
this
is
a
must
for
the
Oh
point.
10
I
know
it's
a
it's
a
question,
I'm
not
sure,
but
this
is
very
important.
G
No,
not
even
mentioning
the
other
items
like
private
charts
and
UI
stuff,
like
markdown
support
itself,
etc,
right
now,
I'm,
just
not
that
so
what
we
have
right
now
and
this
the
weapons
to
do
most
of
these
things
are
native
stuff,
low
level.
Specific
code
very
little,
if
any
from
the
items
still
left,
where
the
open
10
release
pertains
to
closure
side
of
things
or
ghost
sort
of
things.
G
There
are
items
that
are
on
the
are
the
to-do
lists,
which
might
also
include
the
threads
markdown
and
all
that
other
stuff,
but
they
will
take
time
and
for
now
we
have
decided
to
postpone
them.
So
the
initial
release
will
not
include
those,
so
this
might
mean
obviously
that
they
will
have
to
engineer
some
shortcuts
to
bypass
those
missing
features
temporarily
yeah.
Oh,
so
that's
the
current
state
of
the
art
with
regards
to
desktop
chanted
fewer
on
the
line.
Please
correct
me
I,
provide
your
input
to
this,
but
yeah.
G
K
Just
wanted
to
mention
that,
although
the
Nimbus
team
has
given
a
long
list,
I
think
we're
maybe
one
of
the
less
important
teams
a
lot
of
our
project
problems.
Can
we
solve
a
bridging
but,
like
we
have
a
community
outreach,
but
the
thing
the
thing
with
the
Nimbus
team
is
that
we
really
don't
have
the
means
to
fix
any
bugs
we
find
so
like
in
dog
food
and
that's
typically,
what
you
want
to
achieve
right,
that
you
use
the
item
and
then
you
can
fix
it.
K
But
for
the
numbers
team,
which
we
see,
we
see
it
as
like.
We
see
it
almost
as
a
as
an
external
third-party
communication
tool
and
we
have
to
pick
one
amongst
many
like
it
or
like
whatever
right,
because
it's
harder
for
us
to
shift
context
and
six
things
that
are
critical
for
our
works
and
a
lot
of
our
work
is
also
a
little
bit
about
community
outreach
with
the
other
parent
developers.
We're
kind
of
tied
a
bit
more
like
that.
A
D
Curious,
too,
because
there
is
what
are
the
solutions
for
private
group
chats
there's
a
few
chats
that
need
to
be
private
with
regards
to
status,
specifically
like
Corvo
abilities
discussions
and
things
like
that,
we
don't
want
those
to
be
automatically
public
but
eventually
public
once
the
result.
So
what
are
what
are
current
ways
of
dealing
with
private
group
chats
or
just
me
to
take
it
out
off
of
status
altogether?
Maybe
move
to
something
like
email
for
that
I.
A
J
But
what
you
told
me
it's
really
interesting.
We
can
make
a
group
chat
that
is
that
can
be
released.
The
message
like
we
have
the
messages
in
the
indeed
in.
If
we
do
the
it's
private
group
and
we've,
if
you
even
approve
in
the
future
day
this
communication,
we
can
just
show
the
question
decryption
key.
D
J
I
think
yes,
because,
as
I
imagine
that
the
group
chats
these
keys
that
are
used
to
encrypt
message
will
be
changing
once
a
while.
So
you
can,
if
you
want
to
release
the
message
that
this
group,
what
they,
whatever
the
contents,
you
just
release
this
this
path
keys.
We
can
elaborate
Avella
pinking
on
this
in
mind
when
we
develop
the
group,
shots
I
think
it's
something
interesting
to
keep
in
mind.
H
H
Know
there
are
some
group
shots
that
do
that?
Basically,
so
there
are
two
strategies:
it
depends
on
what
the
product
your
product
is
so
many
charts.
If
you
want
new
users
who
read
the
history
of
the
chart,
then
some
charts
do
exactly
that.
So
basically
what
they
do
they
you
wrote
they
you
know
they
ratchet
the
key
using
a
hash
ratchet,
but
then
the
heap
of
keys
so
that
when
a
user
comes
in,
you
know
they
can.
H
H
J
G
I'd
like
to
get
started
with
security,
I
think
because
I've
been
also
thinking
over
some
possible
scenarios
like
let's
say,
team
moves
to
status
desktop
and
we
don't
have
any
protection
against.
Let's
say
spammers
joining
public
channels
and
ruining
them
with
an
empty
messages
or
stuff
like
that.
Oh
maybe
I'm
missing
something
like
a
way
to
block
users.
If
we
don't
have
private
chat
capability,
I,
don't
know
I.
D
J
D
K
K
To
move
to
slack,
basically
that
there
was
no
way
to
protect
things
and
I
guess
I.
Think
the
private,
the
private
group
chat.
It's
it's
more
of
a
priority
question
than
a
technical
one
like
group
chat
can
be
solved
if
we
need
to.
But
one
way
to
protect
against
spam
is
to
use.
Only
private
is
to
only
use
private
group
chats
and
then
have
some
sort
of
channel
where
people
can
get
invited
bye-bye.
K
L
But
you
can
ignore
their
messages.
Well,
I
mean
we
can
implement
something
that
allows
us
to
ignore
their
messages,
but
I
think
we
probably
will
have
a
few
weeks
before
spamming
starts
being
a
thing,
and
we
can
think
about
that
when
it
starts
happening
I
would
maybe,
for
instance,
we
have
a
an
easy
way
to
share
a
list
of
banned
users,
or
maybe
a
whitelist
like
if
someone
post
also
there
was
the
discussion
about
filters
based
on
content.
L
J
So,
some
some
month
ago,
I
written
description,
it's
discussed
a
to
say
on
about
visibility,
stake
or
public
chat
room
governance.
This
is
an
idea
that
users
that
spam
the
network
or
send
messages
that
are
offensive.
Something
like
this
could
get
deposit
that
they
need
to
deposit
to
have
these
their
message
visible
so
which
they
want
to
think
they.
The
UI,
could
configure
itself.
If
you
want
to
see
users
that
don't
deposit
this
is
visibility,
stake
or,
if
you
want.
J
So
if
you
want
to
cut
off
the
spam
and
bad
messages
to
just
rise,
your
visibility
minimal
to
display
the
message
in
the
in
this
group,
so
I
can
post
the
link
in.
We
can
further
discussion
on
this,
but
it's
a
really
interesting
think
to
discuss,
and
there
is
a
lot
of
potential
of
snad
users
on
top
of
that,
also
in
a
way
that,
if
you
like,
if
you
do
a
good
work
in
voting,
what
is
a
bad
message
or
if
it's
a
good-
and
you
also
will
do
a
earnest
enmity
for
doing
that.
J
D
H
D
H
Devon,
look
it's
just
basically
like
essentially
like
it's
its
proxy
encryption.
Is
when
it
what
it
says
it
is
you're,
basically
allowing
someone
else
to
ring
clip
the
message
for
you.
So
in
our
case,
if
you
look
at
our
system
and
compare
it
across
encryption,
it
doesn't
really
help
us
then
much
you
get
the
same.
If
you,
because
you
still
have
the
student
click
times,
you
know
they
can
just
like
the
nice
things
that
allows
us
allows
another
proxy
to
encrypt
it
for
you
for
another
recipient
and
you
need
to
including
with
your
public
key.
K
C
D
D
K
J
Also
I,
my
house
also
when
I
didn't
knew
about
the
solution.
I,
always
thinking
about
using
a
common
shared
key
among
everyone
there.
So
everyone
encrypts
that
their
message
to
this
shared
key
I
know
there
is
this
problem
in
synchronization.
Then
then
we
might
use
the
blockchain,
but
then
the
of
course,
the
news
you
sign
the
message
which
are
public
key
and
then
you
encrypt
it
that
shared
key
and
then
only
people
that
knows
this
singer
see
see
this
key
can
decrypt
it.
K
A
Sorry,
it
was
difficult
to
rejoin
as
a
host
with
zoom
and
everything.
I
agree
with
what
Jessica
said
and
I
guess.
We
should
start
to
finish
it
up.
So
maybe
if
we
do
something
we
can
just
it
just
if
you
there's
some
brief
question,
we
can
talk
about
it
more,
but
this
is
something
specific
that
we
should
ensure
that
you
wanna
ask
people
right
now,
literally.
G
J
Think
the
most
important
thing
is
about
when
you
have
these
public
chats
users
that
register
their
usernames,
they
might
want
to
have
their
username
that
they
registered
displaying
in
this
public
chat,
but
also
they
might
want
to
not
show
it
every
time.
So
we
first
we
need
to
solve
how
we
can
reverse
find
the
the
username.
So
there
is
two
problems.
First
of
all,
the
the
in
the
ENS,
the
usernames
are
hashes
so
and
you
don't
register
the
public
key
to
resolve
to
hash
of
the
username.
J
Of
course,
we
could
implement
that,
but
I
think
the
most
simple
solution
to
so
that
would
be
when
the
users
want
you
to
indentify
themselves,
with
their
usernames,
with
in
a
public
chat
to
people
that
don't
know
him.
Those
doesn't
have
the
in
the
context
they
send.
Also
among
the
with
the
message
in
the
play:
payload
the
username
in
the
in
plain
text,
so
everyone
came
in
then
the
clients
can
check
in
the
in
the
inia
ENS
registry.
J
What
is
a
resolver
in
the
entire
to
resolve
the
public
key
to
see
if
it
matches
with
the
user?
Thats
ended
that
message.
So
we
just
verify
ourselves
to
be,
or
some
users
could
try
to
like
same
oh
I'm
thinking
there
there
I'm
there
I'm
Jared
but
I.
Doesn't
then
this
message
will
be
fake.
It
doesn't
even
be
displayed
at
all
because
the
clients
who
will
be
seeing
that
the
users
trying
to
do
something
that
it's
illegal.
J
A
J
A
J
So
there
is,
they
did
this
problem
of
resolving
the
the
usernames
in
the
public
chat
messages
and
also
about
enabling,
in
the
same
link
the
user
names.
So
they,
the
goal,
is
actually
to
discuss
and
find
how
the
team
wanted
a
implements.
Ens
use
earnings
in
the
state
to
sub
application
in
desktop
in
mobile
everywhere.
J
A
J
Okay,
there
is
about
the
resolving
and
then
about
enabling,
in
the
same
in
the
username
in
the
public
shots.
So
maybe
we
can
have
to
kiss
or
just
doesn't
don't
send
that
that
payload
with
the
like
act
like
you,
didn't
registered
so
then
only
wherever
we
have
you
in
the
context
will
see
your
username
in
the
public
shots
so
that
yes,
so
it's
Irenaeus
all
your
names.
That's.