►
From YouTube: Dynamic Data and Capabilities WG — November 13th, 2018
A
A
A
B
So
during
the
the
last,
the
last
meeting
until
now,
I've
been
working
on
the
world
map
as
many
of
us,
but
the
first
iteration
iteration,
so
I
think
we
finished
it
so
in
terms
of
success,
if
I
I
finished
an
issue
that
Andrea
created,
which
was
related
to
small
issues
around
things
related
to.
Why
and
like
things
that
were
mispositioned
or
it
was
just
wrong,
so
it
created
an
issue
with
a
list
of
things
that
it
detected
and
I
fixed
them,
and
also
we
took
some
time
to
sterilize
some
components
around.
B
B
So,
of
course,
if
the
working
group
would
map
wouldn't
like
take
some
time
so
much
time
like
fall
off
my
time,
I
probably
already
have
finished
these
two
points,
but
fact
unfortunately
I
I
couldn't
so
hopefully
with
this
week.
I
will
finish.
The
next
week's
I
will
finish
a
lot
of
all
these
tasks
and
I.
Think
this.
That's
it.
B
You
I
got
one
question
which
is
related
to
the
world
map
and
because
Adina
I
think
we
factor
the
pier
pods
milestones
and
actually
made
them
ran
or
made
a
card
again
sorry,
they
made
more
granularly
like
when
Oh
points
milestones
and
but
discussed.
If
I,
just
like
a
big
list
of
points
and
in
a
big
milestone,
which
is
version,
one
should
shall
I
make
it
be
annular
to
be
consistent
with
the
rest
of
the
one
not
or
not.
A
C
A
There
is
a
lot
of
the
comments
on
this
classify.
Can
you
elaborate
a
bit
on
an
honest
strategy?
Is
it
related
to
visibility,
or
are
you
making
the
load
sequential
the
loads?
I
know
that,
for
it
is
just
for
everyone
to
be
on
the
same
page,
each
one
of
the
comments.
Is
it
just
a
link
to
appear
first
when
I
profess
object
that
contains
the
the
comment
itself,
so
they
have
to
be
loaded.
B
So
the
charity
data
or
the
value
of
to
see
ret,
just
like
an
index
of
pointers
for
ipfs
notes,
so
the
lazy
load
means
that
is
the
combination
of
a
visual
expect
and
also,
and
also
the
lazy
load
of
the
command
itself.
So
initially,
when
you
enter
the
discussion,
I
fetch
the
whole
index
of
the
Chateauguay,
but
it
doesn't
contain
the
comments
and
then
I
just
apply
display.
The
last
ten
comments
like
let's.
A
Let
me
paste
you
into
the
group
chat.
Thank
you.
So
you
can
you
can
if
you
want,
you
can
add
yourself
there
if
you
want
to
give
us
at
the
end
like
a
quick
intro
about
yourself
and
and
also
feel
free
to
to
at
the
end
of
it
now
we're
running
doing
the
individual
updates,
and
if
you
have
questions
by
depending
on
which
one
you
feel
free
to
post
your
questions
cool.
E
E
One
thing
that
is
very
necessary
actually
and
is
to
when
we
are
deploying
a
new
build.
We
need
to
have
the
source
map
in
roll
bar,
so
we
can
see
actually
where
the
error
is
happening
right
now.
We
just
the
line
one
column
4000,
which
is
super
helpful
in
general,
so
I'm,
gonna,
I'm,
gonna,
add
a
text
then
in
general
what
I
am
doing
right
now.
E
First
I'm,
looking
at
the
UX
proposal
that
has
has
been
working
on
and
I'm
going
through
that
and
making
sure
it
makes
sense,
it
seems
to
just
be
missing
some
things
might
be
the
states
of
the
collaborators
and
being
able
to
see
the
pinners
and
stuff
like
that.
So
I'm
gonna
make
some
suggestions
around
that.
E
The
other
thing
that
I'm
looking
into
right
now
is
everything
about
the
Pinner,
so
I'm,
mainly
looking
at
the
architecture
and
the
functionality
I
guess
of
the
Pinner
and
how
how
people
would
actually
use
the
Pinner
and
how
we
can
have
collaborative
pinning
I'm,
not
looking
so
much
at
the
code
itself
right
now.
There
will
be
a
next
X,
then,
after
that
I
will
start
looking
into
how
we
can
deploy
the
Pinner
and
how
we
can
have
it
sort
of
as
a
bundle
that
people
can
can
use.
E
So,
if
you're
in
a
local
network,
you
could
have
a
Pinner
that
maybe
ships
with
pier
pad
stuff
like
that,
so
we
can
have
it
usable
in
applying
scenarios,
then
probably
we
need
to
have
a
rendezvous
server
for
us
right
now
and
I.
Don't
know
we.
That
would
be
a
next
step
to
figure
out
exactly
what
room
and
also
having
metrics.
E
B
Yeah
I
got
a
question
which
is
related
to
insights.
Are
you?
Are
you
looking
into
collecting
like
performance
metrics
from
our
users?
Let's
say
yeah?
How
much
does
it
take
to?
You
know,
receive
a
delta
and
refi
the
sinusoids,
and
so
on
it
in
a
variety
of
devices
like
slow,
like
slow
devices
and
more
performant
devices,
yeah.
E
I
I
think
that
will
be
very
important
in
general
table.
I
think
that
the
first
step
is
not
so
much
thinking
about
exactly
what
what
metrics
we
need
just
to
make
it
easy
to
add
new
metrics
and
have
a
way
of
collecting
metrics
from
like
asynchronously.
So
if
you,
if
you
run
your
card,
you
know
in
our
disconnected
at
work
or
to
aggregated
metrics
and
then
send
them
to
us.
So
we
don't
just
collect
metrics
from
pure
badness
like
the
deployed
version.
E
But
even
when
someone
is
just
doing
ad
hoc
collaboration
on
a
train
and
I
see
something
is
funky.
They
should
be
able
to
like
same
a
bundle
of
the
metrics
and
then
send
us
an
email
when
they
are
online.
So
that
would
be
the
first
thing
I
would
look
into,
and
then
we
can
start
adding
these
like
performance,
metrics
used
search
retrieve
whatever
we
feel
like
collecting.
Basically.
A
That
was
also
a
question:
what
type
of
metrics
were
you
going
to
collect
and
so
just
just
a
caveat,
Pinner
yeah
if
the
the
how
the
Pinner
behaves
right
now
in
regards
to
private
collaborations,
so
they
were
encrypted
and
they
only
get
the
full
state
and
they
cannot
look
into
the
states.
So
there
is
not
much
but
not
much
they.
They
can
extract
from
that.
You
can
extract
from
the
pin
yourself
so
I
guess
that
you're
also
looking
into
of
course,
so
the
centralized
metrics.
So
that's
that's
yeah.
E
We
want
to
collect
these
metrics
in
case
something
goes
wrong
and
then
it
will
be
sent
to
another
host
on
the
local
network
and
and
collect
the
bundle
so
I
I,
don't
think
it
in
itself
will
be
tied
to
the
Pinner
but
I'm
I'm,
starting
to
think
that
we
need
some
sort
of
bundles.
That
kind
of
like
contains
a
pin
or
a
rendezvous
server.
The
serving
of
the
Pierre
pad
bundle,
metrics,
collector
and
stuff
like
that.
So
that
will
be
the
next
steps
to
figure
out
like
well.
How
would
a
bundle
like
that
look?
Ooh.
F
G
Okay,
so
basic
synchronizing
of
two
graphs
is
up
and
running
updates
seem
to
work.
Okay,
I'm,
not
sure
how
much
to
progress
on
like
having
a
more
sophisticated
algorithm
than
just,
for
instance,
sends
the
whole
graph
and
sync
him
up
locally,
because
some
of
that
may
or
may
not
overlap
with
IP
LD
selector
work.
I
sat
in
on
the
week
by
weekly
call
with
them.
A
If
not,
I
have
I
have
yeah
just
a
comment
regarding
watch.
Yes,
so
correctly,
it's
not.
What
is
that
that
is
the
Sierra
T
behind
the
pure
pad
it's
a
database.
Yjs
is
operation,
biased
and
it
uses.
There's
there's
generator
API
for
that.
Well,
it
can
experiment
with
that.
It's
fine.
There
are
also
other
charity
tools.
There
is
ones
that
we're
using
in
gear
pad
and
peer
start,
which
is
Delta
charities,
so
they're,
not
operation
by
solar
state,
state,
Delta
and
state
based
charities.
A
If
you're
you
want
to
look
into
that
that
you're
thinking
a
graph,
but
regarding
that
I
have
a
question
so
you're
doing
graph
change.
So
if
you
right
now
as
I
understand,
you
have
you're
seeking
the
entire
graph
to
the
other
side
and
then
doing
your
local
emerge.
Right
are
using
graphs,
Inc
itself
or
using
just
bit
swap
well.
G
Yeah
so
I
guess
sort
of
like
neither
on
what
I'm
doing
is
I'm.
Not
the
plan
is
C
is,
for
the
time
being,
just
sort
of
basic.
You
know
protocol
on
TCP
connections
to
share
the
the
graph
which
just
contains
the
not
the
data
itself,
but
the
CID
is
for
the
data
and
then
use
regular
bit
swap
stuff
to
get
data
because
grassing,
it's
I,
don't
like
the
I'm
having
we're
having
a
words
problem,
I'm
trying
to
synchronize
two
graphs.
You
have
graph
a
I
have
graph
B
and
we
synchronize
them
and
I.
G
G
Right
so
so
there
are
some
naming
problems.
So
that's
why
I
like
every
time
I
write
out
like
graph
synchronization
or
something
cuz?
That's
that's
just
the
simple
yeah
version
of
what
it
is,
and
the
idea
is
that,
hopefully,
that
data
shouldn't
be
too
Prabhat
very
large
graphs.
The
data
shouldn't
be
too
expensive,
too
son,
especially
since
I'm
not
planning
at
least
short
term
for
a
few
reasons,
not
planning
on
having
like
every
letter,
be
a
new
Delta
in
the
tree
and
it
sort
of
bundling
them
off
a
little
more
well.
G
There's
gonna
have
to
be
probably
until
grass
sync
is
running:
I'll,
probably
circumvent
it
with
some
sort
of
like
a
want
list
where
I
just
say,
I
want.
You
know
as
soon
as
I
find
someone
who
has
anything
asking
and
say
here's
the
whole
graph
that
I
want,
but
I
don't
want
to
think
the
people
who
are
working
on
like
graphs
Inc
as
a
way
of
fetching
lots
of
data
are.
G
I
should
let
them
do
their
thing
and
instead
work
on
the
the
naming
API
right
effectively.
What
I'm
thinking
synchronizing
the
graphs
is
about
naming
and
keeping
the
latest
version,
and
then
how
do
I
get
all
of
the
deltas
is
like
grass
is
like
a
bit
swap
crap
same
language,
exactly
I
get
back
core
people
are
happy
to
deal
with
that.
A
Was
the
approach
of
pure
star?
That's
that's
the
approach,
so
we
have
a
replication
specific
application
protocol
for
that
I'm.
Also
very
interested
in
how
well
I've
been
like
poking
around
in
a
pill,
the
graph
sync
people
now
and
then,
but
if
you're
interested
you
can,
you
can
like
like
talk
to
them
and
keep
that
flame
alive,
yeah,
no
one
there.
That
would
be
I.
Think
useful,
because
that
for
them
to
have
a
lot
of
use
case
in
mind,
would
be
helpful.
I.
D
Yet
yasaka
from
from
what's
the
protocol,
so
he's
got
a
library
that
that's
gonna
do
decentralized,
identity,
formatting
and
verification.
So
there's
a
lot
of
cool
stuff
going
on
and
I
really
I
don't
know.
I
guess
this
is
turning
into
an
update,
but
I've
looked
very
much
forward
to
collaborating
with
you
guys
on
this.
Have
you
had
those
and
what
is
your?
What
is
your
concept
that
you've
had
for
that
I.
G
Think
we
had
a
we
had
a
couple
of
discussions
about
this.
Some
smaller
groups,
I
think
is
part
of
the
roadmapping
talked
a
little
bit
about
this
I
think
what
what
I
see
is
that
like
Andre,
is
working
at
the
identity
management
and
the
identity
manager
and
like
in
that
sort
of
thing
and
there's
the
there's
I
have
my
idea,
there's
generating
my
D
ID.
D
D
G
D
G
Of
it
is
gonna,
be
synchronized,
but
some
of
the
user
stuff
I'm
gonna
want
to
keep
as
just
me
if
for
no
reason
other
than
there
are
systems
you
can
set
up
with
this,
that
you're
gonna
want
to
have
spam
filtering
for,
and
you
can
just
use
your
friends
list
as
spam
filters,
but
I
think
a
lot
of
it.
Just
is
it's
that
basic
step.
If
I
want
to
have
app
data
that
belongs
to
me,
which,
which
you've
been
talking
about
a
little
bit
yeah.
A
That's
the
part
of
beautiful
well
of
the
the
first
draft
of
the
2019
working
group
roadmap
to
have
private
private
app
specific
data
based
on
an
identity.
So
that's
that's
something
we
we
are
actively
going
to
pursue
not,
but
we're
address
right
now
is
is
working
on
on
finishing
that.
Well,
the
discuss,
if
I
some
features,
always
classify
and
then
I
believe
he
will
jump
into
an
editing
manager,
stuff
and
I
believe
that
we
well
having
this
discussion
here.
A
B
It
was
I
was
thinking
of
I,
think
it's
it's
better
to
because
I
mean
there's
also
a
Dean
now
and
David
also
wants
to
to
give
in
some
innings
and
views
on
these
subjects,
and
so
I
think
it's
a
good
approach
to
have
like
a
session
deep
dive
session
of
ideas
and
and
things
about
identity
and
perhaps
converging
for
something
together.
Yeah.
A
A
B
B
A
A
A
A
So
that,
in
a
large
new
collaboration
with
many
members
that
we
don't
try
to
connect
to
everyone,
there
were
some
bugs
in
in
there
and
that
there
wasn't
balanced
was
getting
in
balance.
That
was
one
peer
that
were
depending
on
the
position.
One
P
was
getting
more
connections
to
it,
then
all
the
other
peers
on
average.
So
that's
corrected,
there's
also
the
duplication
of
the
membership
addresses,
because
you
can
have
many
binding
addresses
right
and
there
was
some
some
duplication
because
we're
using
an
austerity
to
propagate
membership
and
I
said
it.
A
A
A
A
Then
there
would
be
an
accumulation
of
a
synchronous,
I/o
operations
and
they
would
fire
at
the
same
time,
which
would
then
broadcast
a
bunch
of
membership
messages,
depending
on
the
related
all
the
time
that
the
tab
was
was
light
installments,
and
so
hopefully
that
is
corrected.
So
it's
very
hard
to
detect
you
because
it
depends
on
the
the
chrome
throttling
the
tab
so
yeah
there
was
a
bunch
of
chores
like
linting,
putting
AG
r
and
linking
and
fixing
those
those
linking
errors
and
in
progress.
A
I
have
I'm
working
on
exposing
the
collaboration,
replication
events,
so
once
you
have
the
collaboration
and
we
replicate
the
changes
to
other
peers,
you
get
you
expose.
Some
events
like
is
that
peer,
up-to-date
with
my
changes,
I,
have
received
some
changes
from
the
other.
Peer
and
I
have
save
changes
to
a
certain
Pinner.
So
that's
that's
our.
Those
are
the
three
categories
that
I'm
working
on
and
if
you
have
time,
I
have
a
demo
like
a
very
early
demo
on
on
that
next
block
next
is
to
continue
this.
A
This
work
and
also
an
issue
that
I
opened,
is
to
support
different
replica
strategies
for
a
local
store,
so
support
teams
work
of
whether
to
persist
every
oppression,
synchronously
or
just
do
it
all
in
memory
or
have
a
strategy
that
balances
both
of
these
Friday
strategies.
And
that's
it
any
question
for
me.
F
Everyone
so
I've
been
working
closely
with
and
recruit
for
the
pre-alpha
version
of
this
cause.
If
I
do
we
see
this
week,
we
define
the
membership
convention
for
some
components,
such
as
the
comment
one,
and
we
have
a
updated
style
guide
with
all
the
changes
just
to
make
sure
that
the
design
and
off
and
everything
that
is
implemented
implemented
is
consistent
after
that
I
just
follow
Victor's
feedback
regarding
peer
path.
There
are
pixels
regarding
UX
and
UI
level
degrees,
with
greed,
with
with
80
columns
will
change
for
desktop
version
there.
F
There
are
now
some
dismiss
to
all
notifications,
I
created
the
empty
state
illustration
and
the
all
the
history
section
was
revised
as
well.
I
am
blogs
right
now
on
your
part,
mobile,
the
following
also
the
beat
feedback
just
to
make
sure
that
was.
First,
we
have
a
polished
person
for
desktop
on
that
mobile
version.
I
had
predicted
some
transitions
between
between
elements
start
defining
some
some
values
for
the
users,
and
this
will
be
an
backlog
until
the
next
version
is
completed.
F
So
in
progress
right
now,
I'm
working
on
a
quite
simple
exercise,
which
is
a
brand
manual
for
the
web
and
brand
guidelines
for
discuss
if
I,
just
to
make
something
really
simple
and
then
looking
for
for
new
features
for
a
few
parts
to
sync
with
Victor
and
our
colleague.
It
will
be
the
next
steps.
I
guess
and
that's
it.
H
That
specific
feedback
that
we
gave
you
on
the
mobile
mobile
and
that
actually
I
was
not
actually
happy
with
his
feedback.
I
am
a
strong
believer
in
mobile
first
development.
Having
done
this
for
clients
for
like
a
long
time
and
like
if
you,
if
you
leave
the
mobile
app
for
later
its
way
more
work
to
do
it
and
and
the
majority
of
usage
is
in
mobile,
so
you're
actually
like
making
the
desktop
one
first
is
not
but
yeah.
C
F
E
I
was
gonna,
say,
I,
agree
with
David
that
we
need
to.
We
need
to
appear
pub
working
correctly
for
us
before
we
can
focus
on
having
it
on
other
platforms
and
I.
Think
the
main
use
case
of
Pierre
pad
is
using
it
for
collaborating
of
notes
when
you
actually
have
like
a
productive
device
with
you,
which
are
mobile
devices,
usually
for
consumption.
So
we
can.
E
We
can
make
sure
that
the
various
working
for
mobile
device,
but
the
the
input
methods
of
mobile
is
gonna
take
so
long
time
to
get
right,
especially
because
it
be
the
centralized
stuff
that
it
just
makes.
Sense
first
focus
first
on
the
desktop
version,
rather
than
doing
it
mobile
first,
which
would
be
simpler
if
we
have
our
centralized
service
and
we
don't
have
to
deal
with
like
keyboard
inputs
and
focuses
of.
I
H
Sorry,
there's
a
bunch
of
little
things,
so
obviously
the
the
road
Mackworth
work
which
everybody
it
participated
in
going
back
two
weeks,
they
were
I,
found
some
issues
with
the
or
map
CDT
which
sent
some
fixes
for
that
petrol
integrated.
Those
Pedro
had
fixed
some
of
the
editor
binding
issues
that
I
was
seeing
those
causing
corruption
between
my
different
instances
as
I,
integrate,
those
in
and
tested
them,
and
that
they're
working
good.
For
me,
then
I
was
two
weeks
ago,
but
we
did
some
real-life
testing
with
the
pl
staff.
H
I
just
got
on
to
the
slack
and
tried
to
get
a
bunch
of
people
to
edits
the
things
together.
It
worked
okay,
but
then
there
was
just
like
the
whole
thing
collapsed
after
a
while,
because
of
gham,
because
I
I
ended
up
calling
them
zombie
gossip
messages
and
it's
a
different
gossip
problem
than
when
the
Pedro
was
talking
about.
So
this
is
this
is
actually
a
problem
that
is
probably
needs
to
be
fixed.
H
You
know
it
wasn't
like
I'd
shut
down
all
all
my
peers
and
the
messages
would
just
keep
going,
so
it
was
actually
just
in
this
GSI
PFS
nodes
that
were
out
there.
So
so
this
is
just
pub/sub
problem,
so
anyways
I
ended
up
making
my
own
branch
of
the
GSI
PFS
and
the
the
pled
sub.
Just
so
I
could
add
in
some
extra
features.
H
Where
did
where
did
the
mitigations
for
this
go
in
but
I
what
I
was
experiencing
it
like
I,
shut
down
every
single
note
that
I
controlled,
but
there
were
still
peers
out
there
that
had
participated
in
the
collaboration
that
you
know.
People
had
left
the
browser
windows
open
or
something
that
were
the
source
of
the
zombie.
These
gossip
messages
that
were
you
know
they.
H
They
were
like
six,
six
hours
old,
ten
hours
old,
you
know
and
they
just
kept
repeating
repeating
repeating
and
coming
in,
and
if
you
look
at
the
implementation
of
flood
sub,
it
only
squashes
messages
that
are
more
than
less
than
60
seconds
old.
So,
but
if
the
latency
on
got
pub/sub
message
is
more
than
60
seconds,
they'll
just
continue
forever
they'll,
just
ping
pong
back
and
forth.
So
this
seems
to
be
like
a
pretty
bad
problem.
H
It's
gonna
take
out
any
production
peer
pad
thing
that
we
said
that
we
don't
fix
it
so
so
yeah
so
that
took
a
few
days
it
was
I
was
I,
wasn't
surprised
at
how
at
that
problem
so
and
then
after
I
got
so
in
my
version,
I
got
a
little
version
to
fix
that,
or
at
least
I
can
kill
messages
that
I
have
time
stamps
or
more
than
say,
10
seconds
old
and
then
I
don't
modified.
The
pub/sub
doesn't
mean
lay
them.
H
React.Js
binding
was
sort
of
it
wasn't
really
done
right
so
on
Android
like
I,
couldn't
really
type
at
all,
because
so
I
did
a
few
fixes
to
that.
But
I
want
it.
I
think
we
have
to
decide
what's
going
to
happen
with
the
production
pure
pad
version,
because
those
changes
definitely
need
to
go
into
that.
H
The
child
get
the
Peter,
Pan
and
end
load
tests
working,
so
I
usually
started
with
just
the
end-to-end
tests,
but
apparently
those
are
just
nobody's
been
using
them,
so
they
did
their
there.
They
expect
things
to
appear
and
the
patient
have
been
removed
from
the
page.
So
the
end
and
load
test
worked
worked
pretty
neat
that
I
was
having
fun
with
that
side.
I
picked
it
up
to
ten
I
got
ten
nodes
going.
It
went
at
the
same
time
on
my
machine,
twenty
sort
of
lost
up.
H
My
machine
was
doing
it
took
up
my
yeah
so
that
that
seems
like
a
good
framework,
but
I
think
the
challenge
is
going
to
be
like,
like
we
have
one
test,
but
we
need
like
20
little
tests
to
to
focus
on
in
certain
areas
and
I.
Think
about
that.
The
smoke
test
should
be
fixed.
I
should
actually
add
that
so.
H
H
Like
DNS
mapped
to
it
is
like
a
feature
that
is
probably
something
that
could
be
discussed
as
a
feature
for
actual
Pierre
pad
or
have
like,
like
human
rememberable
names,
but
because
it
was
just
my
my
demo.
I
could
just
do
it,
but
for
I
think
there's
some
discussion
about
how
identity
and
yeah
having
like
say
list
of
documents
that
you
worked
on
could
go
into
the
pier
pad
production
version.
Having.
H
So
with
with
my
with
my
pier
pad
nano,
the
purpose,
the
whole
purpose
exercise
was
to
strip
all
the
features
out.
So
I
took
out
I'm,
not
using
any
of
the
the
encryption
capability
which,
which
means
that
any
document
that's
published
there.
Anybody
can
just
go
modify
so
you
know,
but
that
wasn't
the
area
I
wanted
to
focus
on.
I
wanted
to
focus
on
pinning
and
and
that
sort
of
thing,
so
it
was
easy
to
integrated
in
I.
H
So,
but
it
so
okay,
so
so
my
biggest
problem
right
now
is
peer
pad
Nano
is
diverged
too
much
from
the
production
version,
so
I'm
pretty
much
I
just
want
to
pat
down
to
be
stable,
so
we
can
dog
food
I'm,
that's
one
of
the
things
I
have
listed
for
blocking
is
I.
Think
because
we've
had
that
so
much
time
put
into
the
planning
process
for
the
corporates
stuff.
H
We
need
to
really
figure
out
what
the
story
is
for
that
so
and
I
am
willing
to
throw
in
indecent
work
on
that
and
then
I
want
to
characterize
some
of
these
issues.
I've
found
because
we
those
are
going
to
have
to
be
fixed
in
the
production
version
and
yeah
and
then
yeah.
Also
next
week,
I'm
going
to
go
to
been
invited
to
give
a
talk
in
Tokyo,
so
I'm
going
to
be
in
Japan
next
week.
So
I'm
still
gonna
be
working,
but
it
might
actually
work
better.
E
Victor
yeah
sorry,
I
I
have
a
question
regarding
because
none
of
us
as
a
whole
I've
been
unable
to
find
any
sort
of
description
or
requirements
or
roadmap
or
what
what
Nano
is
actually
supposed
to
to
be
and
when.
When
we
consider
it
to
be
like
finished
like
I
I
I
am
the
stanitsa
like
an
experimental
test,
but
for
the
Pinner
and
some
other
collaborative
features
right
now.
But.
H
E
H
Like
okay,
it's
difficult
because
it's
basically
just
when
you're
logging
on
the
Nano
you're
logging
onto
the
machine,
I'm
working
on
right
now,
like
the
desktop
that
I'm
talking
on
right
now,
like
it's
not
meant
to
be
a
big
thing
like
we
want
to
move
these
changes
into
pierpan
itself.
So
no!
No!
It's
just
my
working
copy,
yeah.
E
I
I
am
I,
don't
see
any
where
were
so
for
the
for
the
pin
or,
for
example,
as
far
as
we
have
been
talking
about
the
pin
or
in
general,
it's
always
been
that
it's
experimental
at
this
point
is
it's
not
ready
for
production,
but
I?
Don't
I,
don't
see
the
the
road
map
of
what
what's
missing
for
it
to
be
ready
to
production?
How
do
we
get
it
into
production?
There
seems
it
seems
to
be
a
continuously
effort
that
we
don't
currently
have.
We
don't
see
a
end
to
it.
Yeah.
A
I
I
My
only
comment
they
think
at
this
time
is
whether
it
might
make
sense
to
try
to
unify
the
codebase
and
just
have
like
a
dev
mode
switch
of
some
kind.
If
and
it
gives
you
what
you
need
in
terms
of
like
the
debugging
output,
but
it
doesn't
like
resultant
so
much
differences
of
the
code
basis,
at
least
haven't
be
in
the
same
repo.
You
know
just
about.
E
We
will
know
what
the
Pinner
would
be,
but
I
I
mean
that
feels
like
a
very
long
time
to
just
write
a
list
of
requirements
because,
like
right
now,
when
and
anyone
who
is
working
on
the
panel,
they
currently
are
just
freeform
guessing
what
would
be
cool
to
have
in
the
panel
I
say,
I
think
we
need
to
schedule
something
right
now
before
we
continue
working
on
the
Penner
to
actually
figure
out
what
what
are
we
doing
with
the
Penner
in
reality
like
before?
We
continue
to
work
on
the
panel.
I
G
Is
the
inner
question,
or
are
you
talking
about
like
a
specific?
How
do
we
get
you
know
ten
stuff
for
pure
powder?
Are
we
talking
like
more
generically
whether
it's
you
know
discuss
the
fire
or
anything
else
like?
How
do
we
want
to
go
about
pending
things
that
aren't
you
know
that
we
need
for
our
applications,
whether
it's
the.
A
A
Second,
s,
with
the
current
view
is,
is
pierced
our
generic,
so
you
can
you?
Can
it's
not
in
any
way
it's
specific
to
be
pure
path
and
I?
Don't
think
there
is
the
need
to,
but
it's
right
on
the
main
driver
for
for
for
that
people
at
is
my
driver
for
the
pinion,
because
people
need
to
to
persist
their
changes.
I
A
A
A
B
A
In
a
trustful
manner,
so
it's
just
the
pure
a
little
file
themselves
itself
as
a
pillar
I
think
I.
The
protocol
has
as
an
option
for
that.
So
you
could
say:
oh
yeah
I'm
opinion
I'm
on
the
pin
by
default
in
on
the
pin
here.
So
if
you're
at
European
or
in
terms
of
almost
educating
that
that
that
attributes
its,
we
could
have
like
a
well
known
list
of
pinners
in
the
future,
which
are
specific,
we
could
think
about
different
different
ways
of
of
educating
that,
but
its
current,
currently
it's
I
would
say.
A
A
You
could
do
some
limited
tests,
like
probably
the
denial
of
service
level
attacks
and
probably
some
timing,
attacks,
I'm,
not
I'm,
not
sure,
depending
on
on
how
the
the
other,
the
the
frequency
of
updates
I,
think
they're
well
I.
There
is
in
place
some
some
throttling
on
on
the
communication
to
the
Pinner,
so
those
attacks
could
be
minimized,
but
there's
it's
a
whole
discussion
that
happens
to
be
happy
in
terms
of
the
pinna
features.
A
B
I
A
A
C
You
can
update
yes,
yes,
what
I'm
working
on
I
mean
if
you
want,
if
you
want
to
ya,
know
if
you
want
to
yeah
sure
so,
I
am
working
on
Molly
on
the
research
side,
so
currently
I
am
kind
of
making
a
mix
of
CDG's
and
keep
the
structure
so
essentially
a
mix
of
course
on
trees
and
Merkle
trees.
So
it
is
like
a
CD
data
structure,
which
is
also
more
coal.
First,
on
everything.
C
C
C
Our
demo
plication
is
a
real-time
revision
control
system,
so
it
is
like
a
revision
control
system
which
you
can
embed
into
your
editor.
So,
basically,
when
you
press
control,
as,
for
example,
it
may
sent
exist
in
Russian
of
the
code
too
continuous
integration,
for
example,
this
kind
of
things
so
the
other
guy
who
is
working
with
me,
might
a
little
demo
based
on
Microsoft,
Visual,
Studio
code
and
well
and
sabbatical
protocol
works
are
data
types.
C
There
is
little
demo
and
the
difficult
part
now
is
actually
the
storage
engine,
the
storage
engine
which
must
do
see
he
is
like
a
salt
fin.
We
already
have
one
version
actually
couple
of
versions
working,
but
actually
mixing.
Cd
and
crypt
is
a
difficult
task,
so
you're,
mostly
struggling
with
that
now,
and
also
having
some
discussions
about
Jason
Bay
civilization
for
the
protocol
soul
is
smallest,
the
fins
we
are
working
on.
Yep
questions.
G
A
C
A
D
Please
pick
it
apart.
I!
It's
it's
basically
like
you
know,
alpha
alpha
and
I'm.
Building
this
thing
just
to
learn,
I'd
be
FS
and
to
figure
out
how
all
of
your
systems
work
extremely
interested
in
the
IP
of
Essen.
What
can
be
done?
You
know
and
I
I
kind
of
love.
A
lot
of
the
sort
of
you
know.
I,
don't
know
what
you
might
call
shortcomings
that
it
might
have,
because
it's
making
everything
very
creative.
So
that's
that's!
D
And
then
the
pure
application
goes
and
gets
your
proofs
and
verifies
them
so
trying
to
automate
more
of
that
and,
of
course,
it's
lacking
lots
of
tests
and
I'm
looking
deeply
into
the
ipfs
Identity
Manager
requirements
and
what
is
needed
to
really
so
glad
you
guys
have
put
all
that
thinking
out
there
and
I
there's
a
lot
of
stuff.
I'm
missing
and
I
would
like
to
work
on
that
as
well.
B
Cool
I
think
I
think
you
should
get
involved
in
a
deep
dive
discussion
that
will
have
about
you
know:
identity
in
and
everything
about
an
entity.
There's
a
lot
of
stuff
that
you
you
are
working
on,
that
we
will
be
working
on
so
I
think
wish.
We
could
just
you
know,
merge
together
forces.
You
know
in
order
to
do
not
do
not
have
duplicating
for
efforts
for
the
same
goals.
Absolutely.
D
B
Like
be
sorry
pitiful,
it
can
even
be
like
you
know:
probable
are
open
source
packages,
for
instance,
they
don't
have
stuff
related
to
proofs
and
silencers
and
so
on.
It
can
be,
you
know,
stand-alone
modules
it
doesn't.
They
don't
necessarily
need
to
leave
in
an
eidm
a
repository.
They
are
just
dependencies
off
IDM,
so
I
guess
I
guess
we
can.
Just
you
know,
make
make
a
list
of
all
the
requirements
and
make
make
a
possible
list
of
Records.
A
That's,
let's
get
everyone
that
is
interested
in
deep
dive
like
put
their
names
at
at
the
end
of
the
the
crib
pad
and
and
then
I
will
and
then
I
will
invite
you
to
a
doodle
so
that
you
can
share
your
availability
and
then
we
can
schedule
something
I
think
it
will
be
interesting
also
to
schedule
something
new.
Currently,
let's
let
this
cost
sorry
discussed
this
at
that
at
that
meeting.
If
it's
okay
with
you
because
we're
running
a
bit
well,
we
ran
out
of
time
already
like
11
minutes
ago,.
A
A
A
A
Yeah
I'm,
just
creating
a
collaboration
with
two
two
peers
and
the
Pinner
and
I
should
get
the
the
replication
cross.
Replication
events
right
now
from
from
that
I'll
show
the
codes.
After,
let's
see
if
it's
working
I
was
working
on
that
on
this,
like
30
seconds
before
the
meeting
started,
so
it's
probably
not
going
to
work.
A
A
B
Yeah
I
got
a
few
suggestions,
but
I
don't
want
to.
You
know,
have
discussions
on
that
because
we're
over
schedule
but
I
think
you'll
schedule
like
something
around
having,
especially
on
the
Pinner.
So
perhaps
I
give
some
some
of
my
thoughts
about
how
can
we
authenticate
the
spinner
and
identify
the
Pinner
themselves
within
within
within
the
within
your
circle,
because
I
think
having
a
pins
event,
you
know
I
am
not
really
sure
about
if
we
think
with
which
goods
differentiate,
normal
peers
and
year's
winners.
B
I
think
we
can
incorporate
identity
around
peers
and
pinners
so
that
we
can
identify
pinners.
Just
like
I
can
identify
pedro
as
as
net
beer
and
Adina's
at
peer,
because
you
know
pinners
can
have
identity
and
in
its
theory
of
works
we
can.
We
can
prove
that
the
preneur
is
a
and
you
know,
I
will
just
have
like
a
way
to
get
the
D
idea
of
the
pure
and
and
Bree
fights,
but
yeah.
A
I
in
general,
I
agree
with
you,
I
think
we
need
a
let's
say,
strict
or
a
twenty
five
pillars,
but
that
is
not
based
on
them.
The
planer
telling
us
they
were
thinner
and
we
trusting
that
that
that
that
that's
true,
but
also
peanuts,
are
different
in
terms
of
the
protocol,
because
the
Peter
doesn't
have
the
keys
if
there
is
a
private
collaboration,
and
so
we
have
to
be
careful
not
to
send
them.
Well,
not
careful,
but
we
don't
need
to
send
them
deltas.
A
We
just
just
send
them
entire
states
because
Delta
they
they
don't
need
to
use
Delta's,
because
I
cannot
reduce
the
Delta
to
the
state.
So
if
a
new
replica
comes
online
and-
and
there
is
a
document
that
it's
been
long-running
there's
a
lot
of
Delta's-
you
have
two
options
like
the
Pinner
stores.
The
entirety
know
the
the
toll,
all
the
history
of
a
document
or
only
stores
the
state
if
we
sort
just
storing
the
state
that
will
make
bootstrapping
process
of
a
new
replica
faster
right.
A
So
that's
that's
the
that's
justice
to
explain
that
there
is
differences
between
between
opinion
and
and
the
normal
notes.
In
terms
of
the
protocol
and
the
capabilities
themselves,
but
I
agree
that
that
we
need
to
once
you
identify
what
what
P
but
requires
I,
would
that
my
suggestion
this
is
is
to
is
to
make
that
available
on
peers
are
somehow
right.
Now
there
is
like
this.
There
would
like,
like
two
events,
receive
them
and
locked
loaded
and
saved
and
and
there's
a
specialization
of
the
saved
one
which
is
into
a
thinner
set
into
a
peer.