►
From YouTube: Sequel Core dev call #32
Description
Follow up of Core Dev Call #32 as some agenda items were not discussed due to time restrictions
A
A
One
specifically
started
them
probably
best
to
start
with.
It
is
a
mitigation
process
for
Tobias
to
walk
us
through
what
that,
what
looked
like,
we
have
a
security
team
and
then
a
second
point
that
we
couldn't
get
to
yesterday
is
discussion
on
stemis.
Basically,
following
up
on
the
discuss
posts
that
are,
there,
I'll
walk
through
like
that
process
a
little
bit
more
in
detail,
but
in
a
way
understanding
what
what
is
Simba's.
Why
should
we
be
working
on
it?
Why
now?
What
are
the
responsibilities
or
expectations
of
anyone
who
starts
working
on
that
deliverables?
B
A
A
C
C
C
One
ideas,
basically
from
every
core
contributor
towards
the
security
team.
I
would
like
to
propose
thinking
of
us
as
a
team
as
security
as
a
service
facility,
which
is
probably
done
partially
or
ready,
but
basically
also
wanted
for
me,
okay,
like
whenever
something
is
happening
where
you
think
security
should
have
an
eye
on
tech,
either
me
or
security,
internal
or
Rajiv.
You
can
also
like
direct
messages.
Obviously,
so
that's
basically
only
to
tell
you
okay,
Corey's,
already
available,
read
no
for
anyone.
That
has
not
noticed
then.
C
Basically,
this
this
part
is
also
meant
to
give
you
some
idea
of
how
we
could
perform
threat
modeling,
while
you
guys
are
writing
specs
or
PRS.
So
I
had
a
brief
chat
with
Esther
before
the
call
and
I
think,
especially
like
as
an
example.
The
dad
mention
issue
that
we
discussed
in
yesterday's
call
was
interesting
because,
basically,
we
reviewed
it
collaborate
collaboratively
and
yeah
I
think
at
some
point
there
were
some
security
considerations
discussed,
even
though
there
was
actually
no
security
issues
identified
beforehand.
C
That
could
be
some
indication
that
you
might
still
want
to
take
us
because,
usually
when
people
think
everything
is
fine,
this
is
when
stuff
goes
wrong,
so
punchline
being
taggers
more
often
also
to
get
security
out
of
the
way.
Basically,
you
can
use
this
as
a
catalyst
for
your
own
development
process,
so
you
can
focus
on
the
stuff
that
you
actually
want
to
produce
so
yeah
again,
punchline
take
us
makes
our
lives
a
lot
of
easier
because
we
don't
have
to
go
through
like
everything
that
you
guys
produce,
which
is
yeah
a
lot.
C
On
the
other
hand,
I
can
see
this
this
problem
when
it
comes
to
threat.
Modeling
writing
appears
to
github
issues
which
might
end
up
being
not
read
or
are
not
really
being
interacted
with.
So
I
think
it
makes
a
lot
of
sense
if
you
also
give
the
security
team
some
feedback
on.
Ok,
is
this
stuff
too?
Generic
for
you?
Is
it
actionable
enough
because
in
my
Edessa
straight
off
between
stuff
being
too
educative,
too
generic
not
pretty
interesting
and
then
again
too
specific?
C
C
Then
the
second
direction
is
bug
bounty
program
mitigations
that
we
kind
of
wanna
speed
up
a
little.
So
as
a
first
bullet
point
here,
there's
a
time
frame
of
30
days
for
everything
that
we
get
reported,
I
think
there's
a
video,
ambitious
goal,
but
yeah
as
a
security
and
privacy.
Our
project
I
think
it
makes
a
lot
of
sense
for
us
to
kind
of
aim
high
here
and
go
for
a
timely
resolution
of
any
of
any
findings.
I
think
the
rest
of
the
of
the
bug,
bounty
related
issues
is
known
to
you.
C
C
One
of
the
reports
and
I
want
to
like
personally
dedicate
my
my
kudos
budget
to
people
that
contribute
to
mitigating
security
issues.
Reasoning,
being
security
and
privacy
is
one
of
the
core
values
of
the
status
project,
so
I
think
it
makes
sense
if
we
have
some
sort
of
direct
mechanism
to
incentivize
people
to
work
on
that
and
yeah
give
them
more
credibility
than
for
your
day-to-day
issue.
I
hope
this
makes
sense
to
you
guys,
but
I
think
it's
it's
important
that
we
have
a
differentiation
between.
C
Let's
say
a
regular
bug
fix
and
something
which
is
security
relevant
then.
The
other
thing
I
wanted
to
propose
and
ask
is
if
we
can
paralyze
testing
for
vulnerabilities
together
with
the
mitigation.
So
the
idea
here
is
that
I
think
from
testing
and
tests
being
in
the
CI.
We
can
potentially
also
have
some
cross-project
insights
on
vulnerabilities,
so
the
issue
is
usually
that
if
only
a
few
people
are
involved
in
mitigating
things,
the
same
issue
might
be
reintroduced
either
in
the
same
project
or
in
another
project
and
I
think,
especially
between
react
and
test
subjects.
C
A
lot
of
shared
exposure
and
vulnerability
surface
so
I
think.
If
we
can
get
potentially
Tatiana
and
say
get
to
write
some
tests
which
reproduce
vulnerabilities,
then
we
could
share
these
tests
or
at
least
testing
goals
or
the
testing
paths
that
we
did.
We
implement
there
to
kind
of
be
safe
when,
when
making
changes
to
not
reintroduce
any
bugs
so
yeah,
that's
what
this
was.
Basically,
what
I
wanted
to
wanted
to?
Let
you
guys
know
and
yeah
I
mean
it's
a
process
sketch.
So
whatever
comes
out
of
it,
let's
see
yeah.
A
Toby
huss,
maybe
to
summarize,
for
those
who
are
only
listening
to
the
call
on
youtube
I,
might
not
be
able
to
okay,
okay,
direct
access
to
the
link
of
the
process,
doc.
What
are
the
key
takeaways
that
you
want
people
in
the
different
teams
to
know
when
it
comes
to
dealing
with
security,
relevant
fixes,
testing
or
automated
testing,
and
whenever
they
submit
a
PR.
C
To
make
it
to
make
it
very
brief:
I
mean
you
get
you
get
kudos
whenever
you
fix
a
security
vulnerability,
so
there's
some
some
sugar
and
Wolof.
Definitely
the
other
thing
is,
if
you
think
something
is
not
security
relevant,
it
probably
is
so
tag
us
if
you
think
there
is
a
security
issues,
but
also
tag
us.
C
If
you
think
everything
is
perfectly
fine,
so
we
can
basically
verify
that
I
think
whenever
there
is
a
discussion
around
the
security
year
around
the
security
of
a
feature
or
PR,
that's
a
good
thing
and
then
then
we're
basically
where
we
want
to
where
we
want
to
be
at.
But
in
the
moment
when
there
is
like
yeah,
everything
is
good.
This
is
usually
when
stuff
goes
wrong,
so
yeah
basically
has
a
border
of
some
taggers.
If
you
think
everything
is
fine
and
of
course,
if
you
have
whatever
open
question.
A
A
Okay,
great,
thank
you,
I,
don't
think
we
we
actually
had
an
established
process
of
managing
security
we'd
like
along
with
product
development.
So
that's
a
good
step
forward.
Moving
to
our
next
agenda
item
stim
bus
working
me
suggestions,
welcome.
Does
anyone
feel
inclined
to
give
an
introduction
or
should
I
go
ahead
and
kind
of
like
brief
on
what
it
is
that
we
want
to
talk
about.
D
A
A
There
we
go
but
yeah
what
it
comes
down
to.
Let
me
when
we
walk
through
the,
why
of
this,
there
are
features
and
development
for
Nimbus,
and
that
are
like,
especially
for
a
Nimbus
goes
without
saying,
but
also
for
a
are
developed
for
Nimbus
integration,
specifically
they're
not
being
developed
for
sales
go.
So
if
ever
we
want
to
make
use
of
these
features,
we
need
to
somehow
integrate
with
Nimbus
on
the
client
site.
A
That
also
means
we
would
be
able
to
experiment
with
features
could
be
anything
from
from
messaging
protocol
to
handling
transactions
giving
early
confirmations,
but
in
any
case,
experiments
are
something
that
we
we
cannot
do
if
we
rely
on,
let's
say,
I
believe
it's.
It's
guess
that
is
also
responsible
for
supporting
the
current
aetherium
network.
A
A
A
Another
topic
here
that
is,
that
bundling
different
features
in
Nimbus
and
FAC
will
allow
for
better
Network
scaling
and
eventually
network
scaling
will
help
us
move
away
from.
If
you
that
don't
know
like
what
time
frame
to
be
thinking
out,
that
I
know,
there's
like
an
ongoing
discussion
of
how
do
we
handle
pending
transactions?
Well,
if
we
want
to
be
able
to
do
that,
if
whoo
it
doesn't
offer
it,
then
we
would
need
to
set
up
our
own
sort
of
like
yeah
broker
network
incentivize
routing
network.
A
A
Why
now
is
that
we
could
you
we
could
start
with
messaging
as
a
simple
use
case,
so
this
is
more
to
like
sort
of
the
softening
factor.
It
doesn't
have
to
be
everything
all
at
once,
but
messaging
is
a
is
a
relatively
cut
out
case
that
we
can
start
with.
We
could
slowly
introduce
more
complex
scenarios
that
that
go
more
in
the
direction
of
transaction
syncing.
A
At
sea
start
starting
now
with
again
allow
for
into
including
more
upon
services
later
on,
so
we
could
control
better
rollout,
whereas
if
we
were
to
start
like
four
or
five
months
from
now,
we
would
still
need
to
transition
at
some
point
in
time.
So
there
is
more
of
an
issue
of
if
we
start
later,
deleting
it'll
need
to
happen.
Eventually,
if
we
want
to
become
an
independent,
Network
yeah,
let's
see
what
else
is
here,
I
think
most
of
those
we've
touched
on
missing
anything
you
in.
D
Well
long
term,
the
future
the
future
of
aetherium
is
at
2.0
anyway.
So
there's
something
we
need
to
move
at
some
point
anyway
and
if
we
move
it
sooner
rather
than
once
already
too
late,
it's
better
and
I
think
the
Limbaugh
steam
they
mention
things
like
there's
value
on
the
table
because
are
things
that
get
evolve,
but
then
they
sit
there
for
four
months
while
it
needs
to
be
implemented
in
in
stars.
C
To
ask
I
would
like
to
understand
relationship
between
this
issue
that
we
have
within
furor
and
stimulus
a
little
bit
better.
Maybe
it's
just
some
some
shared
planning
or
submit
the
connection
between
the
two
things.
Hey
Vitaly
I,
wanted
to
give
you
kudos
for
the
for
the
security
issue
that
you
joined.
B
D
So,
okay,
so
take
what
I
say
with
a
grain
of
salt
I.
Don't
want
to
speak
for
them
too
much.
But
on
that,
although
that's
not
what
me
we're
being
asked
to
do,
first
I
think
that
we're
being
asked
to
the
first,
the
Waku
stuff.
What
is
that's,
what
Oscar
mentioned
was
the
priority,
but
it
says
of
in
furor,
there's
the
delight,
clients
and
other
produce.
You
know
we
like
to
get
the
data
that,
and
this
way
will
not
need
to
rely
on
if
you're
there's.
D
The
question,
though,
and
that's
why
I
want
to
just
on
from
Limbaugh's
like
how
much
is
that
actually
develop?
How
much
is
ready
to
use
what
is
the
actual
bandwidth
that
it
uses?
Oh
yeah
and
you
know
of
practical
overall
it
is
and
what
are
the
different
approaches
in
at
one
point
or
an
entry
point?
Oh,
but
in
our
in
a
you
know,
in
our
ideal
world,
if
the
light
clients
work
as
well
as
it's
supposed
to
to
work,
then
we
would
not
need
in
pure.
E
A
lot
of
these
reasons
that
I
think
are
very
valuable,
technical
reason.
A
lot
of
these
technical
reason
I
just
you
know
one
guy's
opinion
and
especially
I,
make
an
example
when
they
stated
features,
they
prevented
faster,
a
name
in
status
curve.
That's
based
on
Waku
excuse
with
water,
where
the
water
fermentation
that
currently
every
name
is
not
complete
in
lots
of
mail
servers
rate
limiting
and
it
was
implemented
by
developer,
who
shortly
after
left
the
company
and
basically
drop
the
ball
or
the
point
on
or
be
prevented
it.
E
E
I
am
involved
in
business
of
ZF
people,
feedback
will
be
say.
Let's
go
and
I've
been
working
with
me,
something
that
other
people
have
not
been
working
with,
but
is
edited
this
video
development,
it's
I'm,
not
sure
with
it.
You
know
it's
it's
difficult
to
compare
a
finger
to
leave
it
whatever
damages
under
the
neck,
but
that's
why
I
push
for
people
that
are
technically
obese
before
giving
opinion,
because
clearly
there's
a
lack
of
understanding.
Some
exist
yeah.
E
D
E
E
Organization
of
quite
a
body,
the
ones
that
probably
book
right,
you
know
it's
nice,
it's
worth.
You
know
that
synergy
between
on
the
TV
2.0
is
good.
The
rest.
It's
I
think
this
question
over,
like
it
experimenting
with
Dean
Wilson
and
as
it
is,
go
back,
it's
it's
again
t
:.
If
you
want
to
further
project
that
you
can
just
work
it
there's!
No,
you
know,
there's
a
beautiful
culture
shock,
it's
just
like
we
own
Nimbus
and
that
make
sense.
You
know
to
push
for
our
implementation
from
a
comedy
strategy,
but.
F
E
E
Is
you're
deciding
if
you
didn't
eat
2.0,
okay
check?
Is
that
if
you
are
the
placement
we're
already
thinking
about
getting
one?
Is
it
just
go?
So
it's
not
something
that
cannot
be
done
as
much
about
can
be
done
much
easier
because
of
this
is
who
is
already
there?
It's
nimble
at
integrating?
It
is
proven
to
be
very
difficult
because
well,
it
was
work
on
it,
provided
that.
A
E
D
B
E
To
you
know,
but
you
can
also
eat
the
great
we.
Don't
it
there's
gonna,
be
a
sages.
You
know,
go
client
for
you
know,
I,
don't
know
you
know,
I'm
not
really
familiar
with
development,
but
there's
going
to
be
a
go
plan
for
that.
Yes,
again,
you're
like
it
would
be
a
bit
silly
for
us
to
use
a
go
client
when
we
have
a
need
to
kind
in-house.
So
but
it's
possible.
B
Hi
Steph-
this
is
rocky,
maybe
maybe
using
the
context
here.
Oh
this
in
any
way.
So
when
we
are
talking
about
dimples
here,
you
are
be
referring
to
numbers
to
point
to
which
the
team
is
already
working
on,
or
are
we
also
referring
to
other
name,
implementations
of,
let's
say
baku
and
eat
1.0
and
others,
which
I'm
not
sure
what
the
statuses,
because
otherwise
I
mean
able
to
point.
Members
is
really
the
point
to
you
know,
become
validated
all
the
stuff
that
we
are
already
working
on.
A
Most
importantly,
Ninh
Waku,
as
new
Oahu
features,
are
being
developed
in
him,
as
well
as
other
Nimbus.
Ethan
features,
Nimbus
2.0
and
like
in
my
mind,
is
something
that
we
would
then
be
like
have
an
easier
time
integrating
afterwards,
but
what
we're
looking
at
for
Ana
coming
half
year
with
being
in
Waku
and
Nimbus
for
each
one
dot.
Oh,
is
that
right
in
your
commentary,.
B
G
B
G
B
D
Not
that
you're
here
Jessica,
so
the
question
asked
about
like
reduce
our
reliance
on
the
if
you
are
and
if
there's
a
Nimbus
code
is
related
to
that
in
a
way
in
the
sense
that
it
could
reduce
that
reliant.
So
in
terms
of
like,
like
lot
clients
in
the
context
of
at
1.0
and
2.0.
Well,
what
is
like
the
state
of
that,
and
and
in
particular
we're
taking
to
you
considerations,
things
like
bandwidth
in
considerations
in
mobile
yeah.
G
G
The
idea,
however,
is
that
by
starting
integrating
the
Nimbus
code
now
and
by
starting
to
use
it
yeah,
we
learn
more
precisely
about
what
API
need
to
be
supported
and
implement
those
with
priority
right.
So
one
pretty
exciting
development
in
these
fun
places
is
being
think
right,
sometimes
like
never
mind
the
implemented
scream
sings.
That
originally
was
made
for,
like
the
client
sitting,
yeah
Trinity,
and
it
seems
your
way
of
thinking.
G
Is
one
stayed
on
the
flies
that
are
going
through
all
the
dachshund
ratings
on
and
that's
the
kind
of
feature
that
we
would
prioritize
if
we
saw
a
concrete
way
of
using
it
in
the
app,
for
example,
and
this
would
also
mean
that
we
can
tailor
this
feature
to
our
use
case
by
basing
it
on
these
real
world
use
cases.
But
that's
not
ready.
G
However,
it's
being
worked
on
and
in
order
to
have
it
ready
or
like
in
order
to
be
in
the
position
where
we
can
use
it
when
it
will
get
ready
what
pieces
far
away
from
now.
The
idea
with
starting
this
in
this
project
right
now
that
we
take
care
of
a
lot
of
other
integration
issues
that
have
to
be
solved
before
we
can
start
using
these
more
efficiency
protocols
anyway
and
in
terms
of
readiness.
G
G
In
terms
of
e
to
light,
clients
don't
exist
yet
not
even
as
a
specifications
can
I
take
a
year
or
two
before
that
happens
in
terms
of
least
one.
There
are
proposals,
and,
and
what
we'd
like
to
do
is
implement
one
of
the
proposals
and
go
with
that
flow,
and
that
would
be
like
a
step
on
the
journey
to
become
less
dependent
on
hormones
and
external
services.
A
G
We
don't
know
I,
look
at
it
right
and
then
what
those
other
services
are
we
could
go
with
with
no
right
and
I.
Think
we
even
try
that
at
something.
So
what
I
mean
the
general
argument
for
going
with
Nimbus
is
that
we
want
to
develop
original
solutions
to
these
problems,
because
existing
solutions
are
not
adapted
to
use
to
a
sufficient
degree,
and
if
we
replace
I.
B
G
I
think
the
whole
idea
of
building
our
own
client
is
because
we
have
a
use
case
which
is
not
sufficiently
served
by
other
vendors,
and
we
believe
that
by
giving
it
focus
its
of
the
sufficient
importance
to
the
organization
that
we
want
to
develop
ourselves
and
at
that
point
make
the
Oscars
is
the
question
like:
where
do
we
go?
And
then
this
is
just
a
natural
answer:
mm-hmm.
A
Okay,
so
I'm
gonna,
try
and
summarize
the
arguments
I've
heard
so
far
like
just
to
close
off
the
agenda
item
on.
Why
should
we
be
implementing
nimbus?
The
two
strongest
arguments,
I
heard
was
yeah
synergy
between
teams
like
just
synergy
as
an
organization
by
using
each
other's
infrastructure
and
using
each
other's
resources,
and
then
the
second
was
preparing
for
each
2.0
by
starting
with
integrating
Nimbus.
A
Now
so
that's
like
preparing
for
what's
going
to
happen
two
years
down
the
line
and
then
there's
like
I
would
want
to
add
C
of
nin
Waku
currently
only
being
developed
in
them.
So
we
need
to
figure
out
a
way
to
integrate
with
them.
Regardless
are
there
any
other
arguments
to
start
implementing
Nimbus?
Now
that
I
missed
there.
C
We
have
some
sorry
I
would
like
to
add
that
we
have
some
inherent
demand
for
the
case
of
in
fewer,
which
I
think
is
a
very
cool
and
motivating
thing
to
work
on
extreme.
So
if
we
have
like
some
project
internal
demand-
and
we
have
some
project
internal
supply,
it's
just
a
question
of
much
making
to
make
it
work.
Okay,.
A
A
D
So
we
can
discuss
this
and
see
if
everyone
agrees
the
edge
a
section
suggested
that
we
start
by
by
doing
a
sort
of
like
a
shame-
and
you
might
call
this
several
things-
I
think
some
people
called
service
layer
proxy
whatever.
But
the
idea
is
that
the
you
just
have
a
very
simple
name,
library
that
all
it
does
is
that
it
mirrors
the
the
exact
API
that
that
let
status
is
is
doing
and
it
then
it
just
redirects
everything
to
starts
go
and
back,
and
this
way
we
could
start.
D
You
know
sooner
rather
than
later,
integrating
with
the
desktop
and
the
mobile
clients
without
any
loss
or
break
of
functionality.
I,
and
with
this
we
could,
then
you
know,
do
certain
changes
and-
and
it's
really
what
the
changes
are
and
it's
here
to
test
them
and
not
to
to
great
things.
Instead
of
becoming
this,
like
a
huge
project
that
you
know
at
certain
date,
we
have
to
do
all
this
gigantic
transition.
D
A
D
E
Considered
I
mean
I,
think
I,
like
the
integral
approach.
I
think
it
is
worth
it
working
to
consider
is
that
the
other
way
around
allows
you
more
granularity.
So
if
you
do
as
well
as
Petrova's,
do
it
so
integrated
into
status
code?
The
reason
being
is
that
you
know
taken
endpoint,
like
our
send
chat
message
right.
You
know
if
you
replace
that
with
name,
then
anything
the
leaves.
You
know
anything
that
leaves
under
the
endpoint
needs
to
be.
E
A
bit
that
doesn't
that
may
be
an
issue
because
there's
a
lot
of
stuff.
That
means
that
you
need
to
before.
You
can
actually
integrate
that
endpoint.
You
need,
you
need
PFS
to
be
implemented
or
the
protocol
stuff.
You
know
they
can't
answer,
install
and
support.
So
you're,
not
really-
and
let
me
know
like
you
and
the
dependence
is
because
then
that
handsome
is
the
unsubtle
score.
So
you
might
end
up
having
to
basically
do
a
big
bang
if
you
could
ever
the
other
way
around.
E
If
you
integrated
the
input
state,
let's
go,
then
you
can
sort
of
replace
whisper
right.
You
start
from
the
bottom
here,
you're
really
starting
from
the
top,
and
so
that's
a
bit
more
cat,
but
I,
don't
know
whether
it
is
a
good
oppression
of
men
approach
because
it's
difficult
to
say
very
stinky
bleah.
The
other
way
around
is
better.
You
know
the
thing
is
nothing
better.
The
problem
is
the
complication,
because
we've
seen
tendrils
work,
that's
complex
is
not
as
many
people
yeah.
D
Well,
I
I
can
see
obvious
point
that
it
does
seem
to
be
easier
and
the
starters
go
for
serving
some
things,
but
especially
connected
things
to
be
signed
and
so
on.
So
there's
a
lot
of
components
in
there
that
others
another
hand
are
things
like
say,
doctor
places
stuff
like
that
that
could
be
done
on
the
under
knee,
so
it's
more
easily
and
the
api's
replaced.
G
G
One
thing
is
that
when
we
create
a
shim
layer
in
same
name
or
really
and
go
as
well
typically
done
in
two
layers,
one
is
a
shim
layer
for
native
name
applications
and
then,
on
top
of
that
puts
typically
ability
to
see
interface.
Just
like
not.
The
go
code
now
exposes
the
see
interface
into
the
go
world,
which
is
a
little
bit
more
simple
than
what
you
would
have
if
you
were
using
go
natively.
G
Is
that
status
go
in
order
to
implement
that
as
name
we
can
selectively
choose
bits
and
pieces
from
status
goal
exports
and
from
that
is
going
input
them
into
Saturday
as
well
expedition
later
so
I,
don't
think
it
that
big
of
a
difference
which
way
you
go
bottom
up
found
in
this
case,
you
can
still
pick
out
parts
of
stars
go
and
use
them
in
name.
That's
that's
like
not
an
issue
importing
going
in
is
just
as
easy
as
exporting
into
go.
There's
no
really
I,
don't
feel!
G
However,
if
the
end
like,
if
we,
if
we
believe
in
the
fact
that
we're
going
to
be
using
a
majority
of
Nimbus
code
at
some
point
in
the
future
and
I,
think
that
we
will
end
up
there
not
too
far
from
now
because
work,
who
is
already
significantly
diverging
from
whisper
and
there
is
more
divergence
coming
likewise
with
research
developments
in
lis.
In
the
p2p
itself,.
E
Just
don't
think
you
think
about
so
I
mean
that's
disagree
with
you,
I
mean
so
J
I'm,
just
thinking
practically
speaking.
So
this
is
the
I
will
call
meme
right
name
will
call
some
status
girl,
for
example,
PFS
right,
because
we
need
to
do
that
and
then
status
girl
needs
to
call
name
again
but
whisper
right.
E
G
Yeah
yeah,
the
girl
code
would
have
to
expose
a
few
things.
So
let's
use
an
example
here,
so
that
it's
concrete
right.
So
let's
say
I
want
to
send
a
message
right
right
now
we
I
don't
remember
how
much
of
the
status
check
message
format.
You
know
the
thing
that
used
to
be
transit
and
now
is
protocol.
That
stuff
isn't
gone
out
right
now,
yeah
yeah!
Exactly
so
like
the
user,
wants
to
send
a
message
right
and
that's
some
kind
of
text
string
and
we
pass
that
deciders
go.
G
If
I
was
approaching
this
problem
with
with
the
status
name
similar
driving
it,
the
way
I
would
probably
do
is
that
I
would
intercept
the
raw
message
and
in
version
one
I
would
just
call
status
go
and
have
it
do
everything
then
in
version
two
I
would
maybe
ask
status,
go
to
perform.
You
know
the
transformation
to
protobuf
and
the
encryption
steps
and
give
me
you
know
the
payload
and
I
would
send
it
to
whisper
myself
from
if
I
was
status.
E
Us
pointing
out
is
not
that
it
cannot
be
done.
I
mean
that's
the
obvious:
how
to
do
it
and
to
say
that
you
know
that
when
considering
the
work
unit
also
taking
consideration
that
at
status
code,
we
need
to
be
changed.
That's
what
I
said.
Basically
like
these,
you
know
things
going
really
you
know
like
if
you
do
it
the
other
way
around
you,
you
catch
the
least
amount
of
status
grow,
and
if
you
do
the
other
way
around,
you
need
to
factor
that
they
dig
into
its
direction.
B
From
what
I
understand
you're
saying,
basically
that's
that
farts,
like
the
FSM
and
maybe
the
portable
parts,
is
currently
in
the
statistical
codes
to
interwined,
but
it's
a
risk
for
the
one
who
code
so
there's
no
clear
API
cut
between
those
and
what
user
gets
proposing.
You
could
not
currently
do
because
of
that,
but
you.
E
E
You
have
two
separate
steps,
so
you
go
all
the
way
down
to
encryption
layer
and
then,
instead
of
stacking
it
going
to
the
next
layer,
which
is
the
transport
layer
whisperer,
you
bring
it
back
up
and
use
a
different
implementation
because
you're
not
we're
not
gonna
plug
it
in
directly
as
it
is
now
at
the
moment.
It's
just
like
it's
a
stack
right
and
so
like
each
day
passes
to
the
other
layer.
So
in
this
case
you
need
to
be
a
bit
different
right.
You
know.
Instead,
it's
a
baby.
You
like
more
system
bath.
B
E
A
So
I
mean
this
is
obviously
a
very
technical
discussion
and
I'm,
not
sure
who
exactly
needs
to
be
involved.
But
it
sounds
to
me
like
there's.
There
is
a
proposal
to
use
its
young
layer.
It
has
pros
and
cons
that
also
have
implications
for
resources
or
how
work
is
required
to
work
on
each
layer
by
respective
team.
So
some
of
the
work
might
land
more
on
the
core
team,
whereas
if
the
implications
for
sales
go
or
minimal,
it
could
be
an
independancy.
A
So
we
avoid
like
a
like
a
big
bang
where,
where
risk
is
high,
that
proposal
would
then
need
to
be
approved
by
whoever
we
appoint
as
stakeholder.
Does
that
sound
workable,
because
it's
a
it
seems
to
me
like
this
yeah
we
can
go,
we
can
go
in
circles
and
what's
the
best
implementation
approach,
I
think
at
some
point
some
people
will
need
to
start
the
work
to
understand
what
the
actual
implications
will
be
and
I
know
Peto
already
an
Adam.
D
Well,
I
mean
it
was
those
people
and
III
want
you
to
want
to
make
a
case
as
well
the
desktop
claim
who,
from
the
desktop
payments
exactly
I,
cannot.
Why
do
we
say
we
cannot
use
any
board
to
work
on
different
things
at
the
same
time,
so
we'll
probably
which
rotation
of
people,
depending
on
the
on
this
particular
iteration.
B
F
F
And
yeah,
we
did
my
only
concern
that
this
has
already
been
brought
up
with
URI
and
we
kind
of
agreed
on
this
is
that
you
know,
as
long
as
we
have
a
like
a
solids,
no
way
to
work
where
we
respect
things
and
we
you
know,
we
make
sure
that
all
the
code
that
ends
up
there
as
tests
and
and
you
know
we-
we
kind
of
take
on
a
consideration-
what
what
core
and
mobile
needs
as
well
I.
Don't
care
who
leads
a
team
I,
don't
care
like
what
the
what
the
org
structure
is.
D
Yeah
so
I
say
I
second,
that
so
that
I
mean
basically
means
that
we
haven't
course
like
some
cold
I,
guess
that
could
be
used
as
a
library
but
will
not
use
that.
Instead,
we
will.
It's
better.
Just
to
you
know,
to
start
over,
like
just
facts,
make
sure
that
a
perfectly
as
attractive
as
possible
as
possible,
but
we
expect
to
make
sure
that
both
of
the
requirements,
mobile
and
desktop,
are
are
taken
into
account,
namely
performance
and
bandwidth,
and
all
that
is
very
important.
D
D
A
F
B
F
F
F
B
A
Like
I,
don't
want
to
get
caught
up
on
on
dates
or
like
exact
deliverables,
but
I
do
want
to
make
sure
that
alia
who
works
on
go
and
him
on.
They
are
saying
Samuel.
It
should
be
involved
because
he's
working
in
go
and
specifying
a
lot
now
I
do
want
to
make
sure
that
we
don't
leave
it
to
chance
that
people
can
review
to
work
so.
A
F
D
That's
our
I
propose
like
we
can
this
week
quite
like
a
proposal
for
the
the
shame
that
can
be
like
discuss
and
analyze
sort
of,
like
a
initial
drafts
back
for
that
and
then
next
week
to
look
at
starting.
It's
like
an
iteration
with
that,
so
to
be
fun,
Fanbox
and
with
some
goals
for
that
I
know,
I
can't
do
a
meaning
for
that.
I'll
just
have
to
see
everyone.
You
know
time
zones
and
schedule
to
see
what
David's
best
okay.