►
From YouTube: Filecoin Core Devs Biweekly #14
Description
Recording for: https://github.com/filecoin-project/tpm/issues/31
For more information on Filecoin
- visit the project website: https://filecoin.io/
- or follow Filecoin on Twitter: https://twitter.com/Filecoin
Get Filecoin community news and announcements in your inbox, monthly: http://eepurl.com/gbfn1n
A
That's
what
we're
going
to
do
all
right
good
morning,
good
afternoon.
Everyone
welcome
to
the
14th
falcon
core
dance,
bi-weekly
meeting,
hope
everyone's
having
a
nice
time,
those
of
us
in
the
northern
hemisphere
it's
getting
warmer
and
the
sun
is
shining
brightly
which
I'm
very
excited
about
cool.
A
So
today
we
have
our
usual
agenda
of
updates,
we'll
be
talking
about
the
status
of
fip
14
and
the
community
poll
that
we're
currently
running
in
order
to
essentially
decide
whether
to
whether
we
should
consider
it
accepted
by
the
community
or
not,
and
then
some
smaller
updates
and
q,
a
as
as
appropriate
yeah.
Let's
start
off
with
our
usual
updates,
maybe
we'll
hear
from
the
foreign
first.
Why
not.
B
Sure
so
I
mean
like
our
focus
has
been
very
much
on
preparing
for
the
audit,
which
is
happening
in
about
a
month,
we're
starting
on
april
26th
and
so
like
a
couple
things
that
we've
been
focusing
on
is
I'm
just
like,
I
guess,
fixing
sort
of
like
the
parts
of
the
code
that
that
need
improvement
prior
to
the
audit.
So
one
of
those
is
the
message
pool,
so
we've
been
refactoring
that
and
working
on
adding
optimal
message:
selection,
as
well
as
improving
our
syncing
performance.
C
B
A
whole,
I
don't
know
eric
if
you
want
to
jump
into
more
details,
if
they're
relevant
for
for
the
rest
of
the
group.
In
terms
of
like
the
actual
improvements
we've
been
working
on.
D
Yeah
sure,
so,
with
respect
to
like
tip
set
processing,
a
lot
of
the
time
spent
that
we
found
in
our
implementation
was
particularly
in
the
crown
actor
and
in
the
flushing
of
the
the
block
store
at
the
end
of
vm
execution.
D
So
we've
been
trying
to
see
exactly
kind
of
what
is
causing
like
these
slowdowns,
like
processing,
like
the
actual
messages,
doesn't
take
too
much
time.
It's
like
probably
like
a
few
yeah
like
3,
400,
milliseconds,
or
something
like
that,
but
processing
our
cron
jobs
takes
around
like
roughly
like
two
seconds
or
something
for
us
and
flushing.
D
The
buffered
boxer
also
takes
like
two
and
a
half
seconds
roughly,
but
we
have
a
solution
for
the
buffered
block
store
and
we
brought
it
down
from
two
seconds
all
the
way
down
to
like
40
milliseconds
like
just
this
morning,
with
a
little
bit
of
trade-off
with
so
the
trade-off
there
is
like
we
end
up,
writing
just
a
little
bit
more
data
to
our
block
store,
but
I
think
it's
a
good
trade-off
for
now,
but
yeah
with
respect
to
the
cron
actor.
D
It's
still
like
not
entirely
clear
why
it's
like
kind
of
slow
for
us,
because
I
from
my
understanding,
like
the
lotus
implementation,
is
able
to
process
like
a
tip
set
in
like
around
one
second
or
something
like
that,
obviously,
depending
on
the
hardware
but
yeah.
So
we're
still
in
the
process
of
trying
to
figure
that
one
out
and
also
with
respect
to
syncing.
D
We
are
just
making
a
little
bit
more
improvements
to
our
like
scheduling,
algorithms,
for
when
we're
in
gossip
sub
thinking,
so
that
you
know
to
to
like
kind
of
robustify
that
and
minimize
kind
of
some
of
the
overlapping
work
and
yeah.
So
that's
kind
of
an
update
on
our
syncing
stuff,
and
I,
I
guess
I'll
I'll,
give
her
update
on
some
of
the
other
stuff
too.
So
we
were
working
on
the
rpc
for
the
last,
the
last
couple
town
halls
or
what
are
these
called
court
dev
meetings?
D
It
feels
like
a
town
hall,
because
it's
like
everybody's
here
but
yeah,
so
we've
been
we
we
got
that
merged
in
and
we're
just.
You
know
we're
implementing
we're
reimplementing,
our
jwt
token
authentication
and
stuff
for
rpc,
endpoints
and
and
we're
also
starting
implementation
on
our
cli,
which
we
used
to
have
before
that's
and
it
stopped
working.
But
now
it
works
again
with
the
new
rpc
stuff.
D
So
we're
just
we're
just
you
know
in
the
process
of
like
figuring
out
the
most
useful
cli
commands
and
implementing
all
of
those
and
yeah.
That's,
I
think.
That's,
oh
and
we're
also
still
continuing
the
start.
Storage
market
integration
so
we're
we're
updating
our
payment,
channel
implementation
and
yeah.
That's
that's
moving
along
quite
smooth,
so
yeah,
that's
kind
of
our
progress.
A
Yeah
cool
that
that's
a
lot
of
work
and
yeah.
Your
audit
is
basically
a
month
away,
which
is
very
exciting
regarding
actor,
optimization
yeah.
So
this
is
something
that
the
lotus
team
and
like
the
has,
also
been
kind
of
interested
in
lately.
So
one
one
thing
that's
useful
here
is
that
the
crown
actor
is
one
of
the
rare
parts
of
spectaculars
that
you
don't
have
to
pay
too
close
attention
to.
A
You
know
matching
all
the
other
implementations
perfectly,
because
gas
doesn't
matter
right,
because
any
any
interesting
call
to
it
is
an
implicit
message
in
which
gas
is
encountered
so
kind
of
more
creative
about
the
box
solution.
So
long
as
you
end
up
at
the
same
state
or
possible,
I
think
you've
been
talking
to
zen
grand
right
about
this
already.
Is
that
the
case.
D
We
just
had
like
a
rough
conversation,
but
nothing
too
in
detail
yet,
but
yeah.
I
definitely
want
to
connect
with
white
a
little
bit
more
but
yeah.
That's
a
really
good
point
that
I
hadn't
really
considered
like
the
the
part
where
you
only
really
need
to
match
the
the
state
and
not
like
the
gas.
So
that's
a
good
point
because
yeah,
I
feel
like
a
lot
of
we.
D
A
A
Because
it's
something
that's
been
on
his
mind,
a
lot
and
he
has
a
bunch
of
ideas.
He
could
probably
share
something
like
sketchup.
A
Have
a
good
conversation
share,
some
sketches,
maybe
some
code
with
you
that
might
be
helpful.
D
Yeah
yeah,
I
I
just
think
we
just
need
to
do
a
little
bit
more
of
exploration
first,
because
I
mean
even
like,
if
we
even
without
optimizations,
I
feel
like
you
know,
our
crown
actor
should
be
much
much
faster,
regardless
and
then,
obviously
in
the
future,
we
can
all
collaborate
on
like
even
more
efficient
things,
but
yeah
like
I
feel
like
there's
just
like
there's
something
in
there.
That's
like
spooky
that
we're
just
trying
to
figure
out.
A
A
Wyatt
yeah
sounds
good
cool
great
all
right.
Let's
hear
from.
F
So
we've
been
focusing
mainly
on
stabilizing
our
solution
and
finding
possible
issues.
Recently.
We
I'm
not
sure
I
have
mentioned
on
the
previous
meetings
that
we
have
identified
a
pretty
big
memory
leak
on
the
part
of
elite
p2p.
F
It
was
successfully
fixed
at
the
beginning
of
the
week,
but
now
we
faced
another
issue,
probably
wisley
p2p
also,
which
causes
some
blocks
not
to
be
received
at
some
point.
So
we're
currently
figuring
it
out
taking
a
closer
look
on
what
is
happening
on
the
connectivity
side
and
we're
going
to
fix
it
somewhere
at
least
so
next
week,
probably
we
well.
It
is
worth
mentioning
that
we're
also
employing
quite
a
big
refactoring
in
terms
of
lead,
p2p,
multi-select
and
well.
F
Yamux
has
already
been
refactored,
so
we
expect
some
improvements
in
terms
of
both
performance
and
memory
usage,
but
I'm
not
sure.
When
exactly
will
we
be
able
to
deliver
these
improvements
on
the
p2p
site,
we
are
aiming
to
have
it
by
the
wall
as
as
we
planned
by
the
first
of
april,
but
now
with
we
have,
we
are
facing
some
more
problems
in
terms
of
stability.
Probably
the
schedule
might
sleep
a
little
bit.
D
I
I
have
some
questions
actually
with
respect
to
that,
because
you
guys
implemented
your
own
lift
d2p,
so
you
should
you're,
probably
really
familiar
with
it,
but
for
for
for
your
implementation
on
fujon
specifically,
do
you
do
any
of
like
the
gossip
sub
topic,
pram
tunings.
C
D
You
find
those
like
particularly
like
useful
so
like
right
now,
just
for
context,
we're
kind
of
using
gossip
sub
in
a
quite.
I
guess,
like
vanilla
way
like
we
don't
really
do
too
much
like
topic
peer
scoring
and
stuff
like
that,
and
I'm
wondering
if
those
are
like
super
super
important
and
if
they,
if
you
see
like
great,
like
memory,
improvements
and
like
olive,
and
I
o
improvements
from
tuning
those.
F
So
maybe
ruslan
will
come
cover
me
in
that,
but,
as
far
as
I
know,
we
well,
we
have
initially
implemented
some
scoring,
but
it
was
not
working
perfectly
so
we
disable
it
for
for
a
while.
I'm
not
sure
it's
enabled
now
so
about
the
fine
tune.
I'm
not
sure
what
what
fine
tuning
is
possible.
Exactly
maybe
ruslan
concur
means
that
otherwise
I
will
need
to
research
that.
A
G
Hello:
everyone
yeah
in
last
two
weeks,
work.
You
were
continually
talking
about
the
disability,
the
main
import
support
and
as
a
as
I
mentioned
before,
we
have
multiple
components:
working
in
paranormal,
for
example,
we
have
a
venous
minor
to
support
different
skill,
our
minors
and
for
mining
yeah.
That
is,
we
have
one
manner
working
for
multiple
many
actors
to
maine
and
yeah.
G
We're
still
working
on
that
and
we
have
a
independent
venus
wanted
component
to
support
different
different
security
policies
to
enhance
the
security
and
another
component
is
venus
messenger,
which
is
managing
the
pending
messages
and
also
message
selection
for
the
mining.
G
Well,
while
generating
a
block
yeah
we're
still
working
on
different
components,
integration
and
equator,
and
to
fix
some
bugs
in
each
component
yeah.
It
will
take
perhaps
a
few
sprints
to
complete
all
this
and
to
have
a
large
version.
G
We
hope
to
have
a
easier
version
to
have
all
the
components
work
together
in
two
weeks
and
and
then
to
yeah
polish
theme
in
the
next
and
the
next
sprint
yeah.
This
is
one
thing.
Another
thing
I
want
to
mention
here.
We
have
separated
some
component,
which
is
not
only
for
villas
and
actually,
for
example,
the
village
messenger
and
venus
one
it
they
can
actually
be
used
by
any
yeah
and
any
implementation,
because
we
have
those
components,
working
and
as
a
service.
G
So
I'm
thinking
if
we
should
separate
that
into
independent
projects.
For
example,
we
have
the
remote
wireless
security
water
as
an
independent
service,
and
we
have
the
messenger
to
manage
the
import
and
to
sign
on
demand
yeah
when
they're,
when
it
will
be
packaged
yeah
sending
out,
I
mean
yeah,
you
could
configure
some
policy
for
the
messenger
to
control
the
gas
limit
or
something
like
that
yeah.
G
I
think
we
can
discuss
this
later
in
yeah
and
how
could
a
civilization
and
to
have
them
independent
and
support
others?
Okay,
that's
all.
A
Yeah,
that
sounds
good.
Thank
you
yeah.
I
think
that's
a
really
good
idea.
I
I
like,
and
obviously
it
goes
beyond
like
even
the
implementations
right
like
I
think,
a
lot
of
kind
of
our
projects
and
ecosystem
partners
that
are
building
cool
stuff
could
potentially
benefit
from
having
yeah.
A
The
venus
wallet
in
particular
comes
to
mind
is
something
that
lots
of
people
might
want
to
use
yeah,
let's
have
a
conversation
in
slack
as
things
get
kind
of
more
polished,
because
I
think
there
might
be
there
might
be
good
demand
for
it
sure
yeah
from
the
lower
side
of
things
we've
had
a
busy
couple
of
weeks
kind
of
the
the
key
thing
that
happened
was
so
in
the
most
recent
network
version
upgrade.
A
We
introduced
optimistic
window
posts
and,
as
we
said,
kind
of
many
times
of
this
channel
and
other
channels,
we
weren't
expecting
a
lot
of
posts
to
be
disputed
because
we
kind
of
thought
that
you
know
someone
has
to
be
do
someone
has
to
be
trying
to
attack
the
system
or
doing
something
really
stupid
in
order
to
get
disputed,
and
then
the
number
of
posts
that
were
getting
disputed
was
higher
than
expected,
not
not
several
thousands
every
day,
but
you
know
a
few
every
day,
and
so
we
finally
kind
of
looked
into
that
and
did
find
the
lotus
bug
with
very
much
an
edge
case.
A
It
was
an
edge
case
with
how
lotus
how
the
lotus
miner
was
sourcing,
dram
randomness
for
its
window
post
when
there
was
a
null
round
at
the
target
epoch
for
the
dram
randomness.
It
was
basically
going
backwards
when
what
the
protocol
expects
and
what
actors
expects
is
that
it
waits
for
the
next
few
rounds,
all
right,
so
the
actual
d-rank
entry
which
will
show
up
in
the
next
episode
so
obviously
not
something
that
was
being
triggered
a
lot.
A
But
when
it
did
happen,
you
are
basically
guaranteed
to
submit
an
invalid
post
which
you
and
which
you
wouldn't
catch
locally,
before
submitting
easy
fix,
put
the
fix
out
there
in
a
in
a
in
a
lotus
patch
release
and
then
and
then
we
had
some
bug
fighting
from
other
stuff
that
was
kind
of
tangential
to
that.
So
that
was
kind
of
the
the
big
thing
that
happened,
but
the
problem
has
since
been
resolved.
A
So
we
have
the
latest
lotus
releases,
153
and
was
put
up
yesterday
and
so
far
things
look
good,
a
couple
of
other
quick
things
that
I
wanted
to
mention.
This
is
more
like
meta
stuff,
but
we've
been
putting
a
lot
of
effort
into
fine-tuning
our
release
process
in
response
to
the
fact
that
you
know
use.
A
I
think
we
have
a
trust
issue
with
with
the
community
right
now
where
people
feel
like
lotus
releases,
aren't
well
tested
and
they're,
probably
right
so
we're
trying
to
slow
down
our
pace.
A
You
know
put
a
lot
of
like
like
a
testing
checklist
in
place
so
that
we
know
we
know
what
we
should
be
doing
before
we
ship
a
release
and
if
we
do
have
like
a
critical
fix
that
we
want
to
introduce,
as
was
the
case
with
this
dram
post
thing,
then
that
goes
in
as
a
patch
release
with
just
that
change
and
doesn't
take
anything
else
with
it.
So
so
look
at
lots
of
work
going
into
that.
A
One
side
note
is
just
if
people
pay
attention
to
lotus
versioning
here
so
far,
we've
been
using
the
minor
version
bub
to
indicate
a
mandatory
release,
quote
unquote,
meaning
there's
a
network
bump
associated
with
it,
we're
going
to
disregard
that
practice
and
use
minor
versions
to
mean
minor
versions
and
use
patch
versions
to
mean
patch
versions.
So
if
you
see
160
or
170,
that
doesn't
necessarily
mean
that
it's
a
mandatory
release,
it
might
just
be
an
optional
release
or
yeah.
A
That's
going
one
note
there
other
thing
that
we've
been
spending
a
lot
of
time
on
which
is
basically
our
next
topic
on
the
agenda
is
the
tip
14
stuff.
So
we'll
put
a
pin
in
that
and
come
back
to
a
bit
yeah
and
I
think
I
think
that's
pretty
much
it.
A
lot
of
a
lot
of
our
focus
is
around
making
lotus
more
usable
as
I've
mentioned.
A
So
we
have
some
project
teams
that
are
working
on
yeah,
improving
how
messages
get
selected,
whether
messages
land
on
chain,
the
deal,
making
success
rate
and
kind
of
lotus
as
a
project
or
as
a
product,
as
opposed
to
like
the
network
side
of
things,
so
lots
of
improvements
happening
there
slowly,
but
not
not
too
exciting.
To
this
group,
I
think.
H
One
question
about
that
bug:
do
you
know
if
anybody's
steaks
got
slashed
because
of
that.
A
No
one's
steak
would
be
slapped
well,
a
steak
is
not
really
a
falcon
concept,
but
people
people
were
definitely
slashed
as
a
result
of
it,
because
if
you
do
submit
an
invalid
post,
you
get
you
get
slashed
and
the
numbers
are
not
small,
so
there
were
there
was
there
was
a
cost
to
to
that
bug
for
sure.
A
H
A
Yeah
decentralized
world:
what
can
you
do?
Our
bugs
are
everyone's
bugs,
unfortunately,
but.
D
Yeah,
I
have
a
question
so
have
you
guys
developed
that
whole
release
checklist
process?
Yet
because
I
would.
E
D
To
take
a
look
at
that-
and
you
know
try
to
use
that
for
us
when
you
know
I
mean
it'd-
be
useful
for
all
of
the
implementations,
for
you
know
when
we're
all
you
know
fully
in
production
and
stuff
and
yeah
100.
A
And
you
absolutely
can
so
long
as
you
promise
to
also
provide
feedback,
because
I'm
really
looking
back
here
yeah,
let
me
drop
the
link
in
the
chat
right
now.
It's
basically
it's
a
pr.
We
haven't
actually
landed
the
template
yet,
but
it
will
be
going
to
the
actual
code
base.
Yeah
that'd.
D
A
Awesome
thanks
pull
request
5826
for
the
benefit
of
anyone
during
the
recording,
yes,
feedback
very
welcome
and
obviously
feel
free
to
blatantly
plagiarize
it
for
your
own
purposes.
If
that's
helpful,
it's
basically
copied
from
the
ipfs
release
issue
template.
So
you
know,
there's
a
chain
of
plagiarism
happening
here,
cool
any
other
questions
about
a
lot
of
stuff
or
really
anything
else.
A
Cool
okay:
let's
move
on
to
flip
14,
which
I
think
is
the
next
item
on
the
agenda.
Yes,
I
think
hunter
already
dropped
the
the
link
in
the
chat
yeah.
So
this
is
14,
which
is
basically
the
do.
We
allow
folks
to
extend
b1
sectors.
A
We've
had
this
conversation
a
few
times
already,
so
from
a
security
perspective,
we
continue
to
feel
it's
okay,
it's
not
the
greatest
thing,
but
there's
probably
no
real
advantage
that
can
be
gained
no
real
or
no
real
cost
of
the
network
that
that
that
can
be
incurred
as
a
result,
a
result
of
allowing
these
to
be
extended
to
a
maximum
lifetime
of
540
days,
which
is
what
the
fit
proposes
anything
longer
where
you're
kind
of
uncomfortable
with,
and
that
that's
basically
where
we
were
the
last
time
we
spoke
about
it.
A
So
what
we've
done
since
then
is
just
dropped.
A
link
in
the
zoom
chat
to
phil
poll,
which
is
this
tool
that
we
built
a
while
ago,
that's
basically
for
polling
the
the
pipeline
community,
which
is
basically
exactly
what
we
wanted
to
do
here.
So
we
opened
this
poll
basically
saying
hey.
A
Do
we
want
do
we
want
this
clip
or
not,
and
support
can
be
used
by
pretty
much
anyone
on
the
network
like
if
you
hold
tokens,
if
you
are
a
storage,
client
and
so
on,
we
kept
this
poll
restricted
to
miners
because
looking
at
the
github
issue,
it
kind
of
felt,
like
miners,
also
super
divided
about
whether
this
poll
is
whether
this
clip
was
wanted
or
not.
So
basically,
yeah.
A
If
you
have
a
storage
miner
that
has
power,
you
can
vote
in
this
poll
and
right
now,
it's
winning
96
to
four
in
favor
of
adopting
the
set
there's
still
tons
of
stories
that
haven't
voted
yet,
though,
like
turnout
is
like,
maybe
16
or
so
right
now
so
yeah
so
we'll
see
we'll
see
what
the
results
of
the
poll
are.
The
poll
runs
until
mid-saturday
utc
and
then
based
on
that
we'll
kind
of
make
a
decision
like
this.
Isn't
a
you
know.
A
If
it's
above
50,
we
will
do
one
thing
or
the
other,
but
this
is
kind
of
like
so
based
on
this,
we
kind
of
we
have
a
sense
of
how
the
community
feels
and
in
this
case
yeah
if
the
community
is
strongly
in
favor
of
one
decision
or
the
other,
then
we'll
just
go
with
that,
because
we've
kind
of
talked
about
all
the
other
aspects
of
it.
Any
questions
about
that
or
thoughts
about
this
process
in
general.
Do
you
feel
it's
fair?
Do
we
feel
yeah?
A
D
Personally,
like
I
mean
I
don't
feel
too
strongly,
either
way
I
mean
I
I'm
leaning
towards
like
not
implementing
this
like
I
I
just
like.
I
don't
see
like
like.
I
think
most
of
these
v1
sectors
are
like
from
space
race,
and
it's
like,
I
feel
like
originally.
The
plan
was
to
erase
space
race
and
start
from
new
genesis
anyways,
but
somehow
we
like
fork
space
race
into
mainnet
and,
like
I
I
don't
know,
I
just
don't.
I
just
feel
like
you
know,
having
a
hard
fork.
D
Just
for
something
like
this
like
doesn't
seem
to
make
sense.
But
again,
I
don't
really
feel
too
strongly
either
way,
but
yeah.
That's
just
kind
of
my
thoughts
on
whether
it's
like
quote,
unquote,
fair
or
not.
A
A
The
thing
that
bugs
me
about
this
and
the
reason
I'm
kind
of
gently
in
favor
of
it
is
it
it
it
sucks
if
you
seal
the
sector
for
180
days
instead
of
540
days
and
as
a
result
of
that
thinking
that
you'd
be
able
to
extend
it
later
on,
and
as
a
result
of
that,
you
know
some
people
sealed
all
the
way
to
the
540
and
will
have
the
full
secure
lifetime,
whereas
other
people
are
like.
A
I
didn't
know
that
this
option
would
be
taken
away
from
me,
and
so
I
kind
of
feel
like
that
option
should
only
be
taken
away
if
we
feel
better
like
if
there's
a
proper
security
concern,
if
we're
like.
No,
this
is
unsafe,
which
is
not
where
we
are
right
now
and
so
that's
kind
of
where
I
come
to
it,
but
I
I
I
don't
necessarily
feel
strongly
either,
which
is
why
I
was
happy
to
like.
Let
me
you
know,
let
this
be
a
community
decision
for
the
most
part.
A
Yeah,
no,
I
think
that
I
think
that's
a
very
good
flag
and
like
this
is
what's
tricky
right,
like
at
the
end
of
the
day
on
some
level
like
the
implementers
are,
who
ultimately
will
be
you
know,
adopt,
will
be
like
implementing
a
fip
or
not,
and
so
there
on
some
levels,
kind
of
like
the
decision
is
made
here,
but
at
the
same
time,
like
especially
with
a
flip
like
this,
the
the
change
in
question
is
two
lines
and
any
anyone
in
the
world
could
make
it
and
if,
if
that
gained
enough
traction,
then
that's
what
the
file
point
network
is
doing
like
there's,
there's
no
one
really
steering
the
ship
here,
which
is
as
intended.
A
Of
course
I
do.
I
do.
I
do
agree
with
that.
Concern,
though
I
think
I
think
the
important
thing
to
do.
There
will
be
that
like.
If
something
doesn't
like
you
know,
we
can.
We
can
have
fips
written,
obviously
and
the
more
flips
the
better
and
the
more
community
participation
the
better.
But
if
something
doesn't
pass
the
security
check,
then
it
can't
get
any
further
like
if,
if
we
feel
like
and
that
that's
something
that
we
can
kind
of
try
to
be
objective
about,
like
is
will
this
is.
A
Are
there
network
security
concerns
here
and
if
that's
a
yes,
then
then
we
then
it
doesn't
get
much
further
than
that,
but
agreed
like.
We
are
we're
definitely
setting
president
here,
because
this
is
the
first
time
we're
doing
any
of
this
stuff.
D
Yeah-
and
I
mean
like
maybe
the
fip
itself
doesn't
necessarily
have
a
security
concern,
but
doing
a
hard
fork
is
a
security
risk
like
in
itself
right
yeah
but
yeah,
I'm
just
I'm
curious
to
I
mean
it's
obvious
like
what
the
miners
want
already
like,
even
if
there's
only
16
of
the
people
voting,
but
I'm
very
curious
to
hear
from
the
other
implementations
on
like
their
opinions
on
this.
G
In
my
opinion,
I
would
think
in
this
time,
especially
the
process
is
a
very
good
thing
for
the
communities
and
for
the
ecosystem
is
because
before
we
have
yeah
almost
everything
and
yeah,
I
made
a
decision
for
everything
and
were
made
by
the
pao
yeah.
But
right
now
we
are
have
a
kind
of
process.
We
use
this
process
to
have
the
community
members
to
discuss
all
of
this,
and
we
have
the
process
you
know
for
the
new
child
and
our
governance,
which
is
very
good.
I
think.
G
Okay,
this
is
one
thing
and,
and
another
thing
is
that
we
still
depends
on
the
professional
team
to
make
the
decision
this
kind
of
your
kindness,
which
is
good
yeah.
I
would
say
this
too,
and
this
is
the
first
time
actually
we
we
handle
this
kind
of
vape
and
there
are
lots
of
arguments
actually
to
you
know
to
have
one
process
to
handle
this
so
yeah.
I
would
like
to
see
how
it's
you
know
move
on.
G
How
could
it
be
just
successful
or
yeah
and
or
a
failure
and
yeah
see
this,
and
we
we
come
back
to
some
postmortem
and
to
perfect?
This
is
a
process,
that's
my
opinion,
yeah
personally
yeah
for
this
vape,
particularly
I
okay
as
one
of
yeah,
the
community
member
and
also
one
of
the
miners.
I
would
think
that
there's
fairies,
usually
I
strongly
support
yeah,
that's.
This
is
one
thing
I
want
to
mention
too.
A
Yeah
yeah,
I
think,
there's
two
two
important
threads
they're,
almost
like
just
restating
what
you
said
number
one
yeah,
I'm
definitely
happy
with
how
much
of
this
did
not
come
from
any
of
us
really
or
did
not
come
from
pl,
at
least
where
you
know
the
flip
was
written
by
the
community.
The
implementation
came
from
came
from
outside
of
us
as
well
so
like.
That
is
definitely
good
because
that's
something
we
really
want
just
like
push
a
lot
of
this
decision
making
away
from
us.
So
so
that's
been
good
to
see.
A
Obviously
this
is
something
the
mining
community
is
heavily
invested
in,
but
if
that
generates
momentum
for
other
for
stuff,
like
this
to
happen
with
other
in
other
domains,
that'll
be
really
good.
Also,
yeah
definitely
echo
the
point
about
yeah:
let's
do
a
postmortem
here,
let's
get
some
learnings,
let's
get
feedback
from
everyone.
A
This
group,
the
community
kind
of
everyone
affected
by
this
already
a
lot
of
small
minors,
have
been
pointing
out
that
that
they
feel
it's
unfair
because
you
know,
if
you
only
have
a
few
terabyte
minor,
then
you're
you're
dropping
the
pond
or
drop
in
the
ocean.
In
the
in
the
current
poll-
but
you
know
that's
on
some
level-
that's
kind
of
how
blockchains
work,
but
at
the
same
time
is
there
stuff.
We
can
do
to
allay
that.
A
One
thing
we
can
do
is
when
assessing
the
results
of
the
poll.
We
can
also
look
at
the
number
of
ballots
cast
on
either
side
instead
of
just
the
the
overall
power.
So
at
least
we
get
a
sense
of
you
know
how
smaller
miners
feel.
So
there's
definitely
some
some
some
good
learnings
that
we'll
draw
here
on
some
level.
It's
good
that
at
the
end
of
the
day,
this
is
a
fairly
small
decision
like
it's
contentious
for
sure.
A
But
it's
not
you
know
it's
not
the
biggest
decision
we've
ever
made,
so
it's
good
to
be
kind
of
almost
getting
like
a
test
run
with
with
something
small
and
not
too
consequential
before
there
will
be
scarier
bigger,
harder
decisions
to
make
in
the
future.
I'm
sure.
D
Yeah
yeah
do
we
have
any
people
from
the
actors
implementation
here,
because
I
have
like
one
question
with
respect
to
this
upgrade
in
particular,
or
maybe
you
can
like
anybody
like
it
was
a
question
like
I
was
just
wondering
if
this
like
requires,
like
a
new
major
release
of
the
actors-
crate.
No,
it
doesn't
it
doesn't
it's
okay,
yeah,
because
yeah
that
would
be
kind
of
annoying
yeah.
A
The
the
entire
upgrade
here
should
be
in
total
in
on
our
side,
it'll
be
like
10
lines
of
code,
it'll
be
the
actual
pr
and
then
adding
network
version
11
where,
if
yeah,
if
you're
past
network
version
11
allow
extending
these
sectors
with
this
restriction
is
basically
all
that
needs
to
happen,
and
then
we're
doing.
D
D
On
the
fifth,
if
you
got
like
I,
so
I
from
what
you're
telling
me
it's
like
you
guys
on
your
side,
have
implemented
this
already
preemptively.
If
you
could
just
link
to
those
pr's
where
you've
implemented
this
on
lotus
and
on
spec
actors,
like
maybe
I'll
like
you
know,
just
go
ahead
and
like
make
a
draft
on
that
on
our
side
as
well.
That
sounds
good.
A
Yeah
we
haven't,
we
actually
haven't
done
anything
on
our
side,
but
we
did
have
the
community
pr
coming
from
stephen,
which
has
been
reviewed.
So
that's
specs
actors
1384,
dropping
it
into
chat
now
yeah.
So
that's,
basically
all
the
work
that
needs
to
happen
on
the
actor
side
lotus.
Will
the
lotus
side
of
things
will
be
use.
A
This
new
version
of
actors
and
network
version
11
exists
at
this
height
so
should
be,
should
be
nice
and
simple
yeah,
so
that
does
bring
up
the
sorry
any
other
questions
about
this
process.
Before
we
start
to
talk
about
kind
of
the
timeline
that
we're
looking
at.
E
Sorry,
I
just
want
to
point
out:
I
just
linked
a
discussion
in
the
song
chat,
so
community
discussion
114,
is
where
we're
going
to
collect
feedbacks
on
like
how
we
do
this,
like
pulling
like
community
governance,
pulling
a
mechanism
like
in
the
future
like
if
you
have
any
thoughts,
leave
a
comment
there
and
I
also
want
to
point
out
like
I
want
to
put
that
on
record.
Yes,
this
is
not
a
voting
tool,
but
a
polling
tool.
E
It's
like
this
group
of
people,
you
know
co-implementers
at
faulcone
foundation,
should
review
the
results
from
different
perspective
after
the
pool
ends.
Instead
of
we
are
just
saying:
hey,
we
take
storage
minus
power
and
make
the
decision
and
also
like
in
the
future.
We
definitely
want
to
open
other
categories
like
stakeholders
to
vote.
For
example,
you
know
core
devs
token
holder
storage
class,
which
is
also
a
huge
part
of
the
network.
A
Thanks,
jennifer
yeah,
okay,
so,
let's
so
in
terms
of
timeline,
so
we
did.
If
you
look
at
the
fips
issue,
we
got
this
like
report
done
or
block
science
gave
us
this
report
about
when
these
sectors
will
start
expiring
and
basically
they've
already
started
expiring,
but
like
in
the
in
the
few
like,
like
tens
every
day,
and
that
number
starts
to
escalate
around
mid-april.
A
So
we
were
thinking
of
an
april
12th
network
upgrade
with
the
lotus
release
for
that
going
out,
basically,
in
the
first
few
days
of
april,
probably
and
giving
people
about
seven
to
ten
days
to
upgrade
yeah.
Does
that
say
because
the
change
is
so
small?
You
know
the
integration.
The
implementation
work
is
basically
non-existent
here,
so
it's
mostly
about
testing
it
in
a
devnet
making
sure
that
it
does
what
we
think
it
does.
A
So
you
can
actually
extend
these
sectors,
then
put
it
on
calibration
net
and
then
put
out
the
release
and
tell
everyone
to
put
it
on
mainnet,
like
it's
very
much
the
exercise
of
it
as
opposed
to
the
actual
implementation
work
here.
Does
april
12th
for
the
network
upgrade
itself
sound
good
to
people.
A
We
should
share
this
timeline.
We
haven't
actually
posted
it
yet
because
we
don't
actually
know
if
this
will
be
happening
or
not
yet.
E
A
Is
all
like
tentative,
but
that's
kind
of
what
we've
been
thinking.
G
Yeah,
I
think
we're
better
to
stick,
and
the
plan
is
because,
as
we
yeah
and
according
to
the
data
we
gathered
was
the
first
wave
of
the
exploration
is,
I
think,
yeah
from
april
13,
so
yeah,
because
we
are
doing
this
for
yeah
to
make
it
more
meaningful
yeah
that
date
will
be
yeah
good
data.
I
would
think
yeah.
A
It's
a
little
tight
but
yeah.
I
definitely
I
feel
strongly
that
we,
if
we
are
doing
this,
it
should
happen
before
the
first
rape
of
expiration
starts,
if
not
we're
only
making
it
even
more
unfair,
where
you
know
some
percentage
of
people
who
will
have
their
sector
saved
and
others
won't.
But
that's
just
my
opinion.
E
And
sorry,
I
I
just
want
to
like
ask
the
other
implementers:
what's
the
best
way
like
in
the
future,
for,
like
mandatory
network
upgrade,
what's
the
best
way
for
us
as
a
group
to
decide
the
timeline
so,
like
I
mean
like
so
like
before,
it's
like,
we
will
propose
a
timeline.
We
will
ask
opinion
like
for
like
within
this
group
and
see
how
everyone
feels
you
know
with
all
this
actor
integrations
community
concerns,
like
all
that
are
we
do
we
prefer
to
do
that
async
or
like
how
shall
we
start
the
conversation?
D
Yeah
I
mean
how
I
kind
of
like
what
what
I
kind
of
envision
future
network
upgrades
to
look
like
is
obviously
with
a
regular
cadence
and
like
a
like
a
more
formal
process.
So
it's
like
I
mean
I
envision,
like
most
of
these
network
upgrades
to
be
based
on
like
fips
that
are,
you
know,
submitted,
and
then
you
know
how,
like
some
x
amount
of
time
before
our
scheduled.
D
You
know
upgrades
like
we
will
come
together
and
kind
of
discuss
what
fips
should
go
into
this
next
upgrade,
and
then
we
kind
of
do
those
do
do
it
like
that.
That
I
mean
that's
kind
of
like
the
ideal
situation,
I
don't
know
how,
like
I
think
that
could
that
should
work,
but
yeah.
F
Yeah
I'm
gonna
agree
with
eric,
though
I
I
also
think
that
the
the
exact
I
think
like
the
way
you
communicate
the
changes
might
be
also
useful
to
establish,
so
it
might
be
you
via
github
or
the
discussion
slack
or
the
most
effective.
I
think
is
these
bi-weekly
meetings,
where
we
like
discuss
future
plans
for
at
least
like
several
weeks.
A
Yeah
we
talked
about
this
back
in
meeting
12.
I
remember
correctly
about
yeah
how
we
want
to
do
this
kind
of
thing
moving
forward
and
we
basically
hit
those
same
points
which
is
most
of
the
work
should
be
happening
in
fib,
so
everyone
already
knows
what's
going
on,
and
the
question
is
mostly
which
fips
are
going
into
this
release
and
which
aren't?
A
There
was
also
the
flag
that
a
few
of
you
all
raised,
which
was
we
also
if
we
do
have
smaller
tweaks
to
specs
actors,
then
a
detailed
change,
log
and
explanation
of
exactly
what.
If
then,
essentially,
if
anything
happened,
not
through
a
fifth,
then
clear
explanation
of
exactly
what
was
happening
there
is
is
very
helpful,
because,
if
not
kind
of
decoding
that
is
a
nightmare
and
then
yeah,
the
missing
piece
was
like
agreeing
upon
the
launch
epoch
together.
A
Probably
what
one
thing
that
would
be
useful
is
yeah
just
kind
of
like
discussing
when
we
think
the
next
network
upgrade
will
be
in
every
one
of
these
meetings
just
being
like
this
is
when
we
think
it
is.
Are
we
still
on
track
for
that?
Do
we
want
to
push
it
forward
or
backward?
So
I
guess
on
that
note
yeah
right
now.
Obviously
there
may
be
an
upgrade
depending
on
the
outcome
of
this
poll
on
april
12th,
but
that'll
be
a
very
small
one.
A
Apart
from
that,
we
were
thinking
v4
actors-
that
includes
proof,
act.
The
proof
aggregation,
flip,
flip
13
in
june-
is
that
workable
for
for
everyone
here
or
are
what
are
initial
thoughts
about
that.
D
I
think
june
makes
some
sense,
especially
if
you
know
we
can
have
like
a
release
candidate
out
for
spec
actors
and
all
of
that
stuff
before
then,
and
if
we
have
like
a
good
idea
of
exactly
which
fips
are
going
in,
so
we
can
kind
of
like
prepare
for
that
and
get
like
this
interop
net
and
all
that
stuff
back
up
and
running.
D
I
think
those
are
going
to
be
very
crucial
because
I
think
by
the
by
the
time
june
comes
like
we
should
be
finish
our
audit
and
we
should
be
on
main
net,
so
yeah
fingers
crossed
like
we're
still
syncing
mainnet
right
now.
It's
still
working
so
hopefully,
hopefully
in
june,
like
nothing
blows
up
but
yeah
like
so
yeah,
like
I
think
you
know,
june
sounds
okay.
A
A
A
Yeah
yeah
happy
to
announce
that
wyatt
is
kind
of
like
fine
like
pushing
forward
on
that.
So
after
a
lot
of
after
a
lot
of
lost
momentum,
we're
hoping
to
to
make
that
a
priority
again.
So
that's
good.
A
Cool
sounds
good,
yeah,
obviously
we'll
iterate
on
this,
we'll
we'll
see,
I'm
sure
there'll
be
stuff
that
we
didn't
put
down
on
this
list.
That
we'll
discover
later
would
have
been
helpful
and
we'll
we'll
include
them
next
time,
but
yeah.
Let's
do
a
check
in
kind
of
every
every
time
we
have
these
conversations
being
like
yeah
we're
still
looking
at
june
or
june's
not
happening
or
worst
case.
A
It
actually
has
to
be
made
for
some
reason
or
the
other,
but
let's
not
allow
those
worst
cases
to
happen,
because
we've
we've
done
enough.
We've
done
enough
of
rush
timelines.
I
think
we
can,
let's,
let's
slow
the
beep
down.
A
Cool
yeah,
the
last
thing
I
wanted
to
talk
about
very
quickly
is
fip
12..
Let
me
drop
a
link,
real,
quick
thanks
hunter
for
having
them
on
the
issue.
You
beat
me
to
it.
I
took
12.
This
is
the
very
small
one
that
we
talked
about
a
few
meetings
ago.
A
Basically,
we
want
people
to
be
able
to
reuse,
verified
client
addresses
to
get
top-ups
instead
of
having
to
you
know,
either
fully
exhaust
their
their
data
cap
or
just
like
abandon
one
address
and
start
and
use
the
second
address.
If
they
want
to
get
some
more
data
cap
yeah,
I
I
don't
think
anyone
had
any
issues
with
it,
so
it
was
sitting
in
last
call
for
a
while
we're
moving
into
removing
it
to
accept
it.
Unless
someone
suddenly
discovered
a
major
objection
to
it,.
A
I'm
hoping
to
get
it
into
the
next
network
upgrade
just
because
it's
it
should
be
very
simple,
like
the
next
network
upgrade
the
next
plan
regularly
scheduled
network
upgrade
not
anything
that
happens
for
as
a
result
of
the
football.
So
the
june
time
yeah
in
my
in
my
head,
that's
the
next
real
network
upgrade,
and
this
is
just
a
a
one-off
thing.
A
I
I'd
also
would
also
like
to
get
fip
11
in
which
was
the
one
that
removes
the
dutch
auction
for
consensus,
fault
reporting
again,
that's
about
deleting
10
30
lines
of
code,
so
it
should
be,
should
be
fairly
easy
to
do
as
well,
but
we'll
we'll
cross
those
boats
when
we
come
to
them
or
whatever
the
phrase
is
yeah.
That's
basically
everything
on
my
timeline.
Any
questions
any
stuff
that
folks
want
to
talk.
A
A
Yeah
jennifer
flags
flip
eight,
which
we
can
talk
about.
It's
probably
not
you,
it's
probably
not
getting
into
the
june
network,
upgrade
because
it's
going
to
be
a
bit
of
work
as
well.
So
maybe
we
hold
off
on
that
because,
if
not
we'll
talk
about
it
now
and
then
revisit
it
like
several
months
later,
when
we
actually
get
around
to
implementation,
but
we
can,
if
people
want
to.
I
don't
have
a
lot
of
context
here,
but
nikola
is
here
who
can
maybe
give
us
a
two
minute
version.
I
Hello,
let
me
just
give
you
a
two
minute
version
of
the
things
that
we're
working
on
in
general
ipp
is
doing.
Is
this?
We
need
to
find
a
better
name.
I
think
we
we
do
have.
We
do
actually
have
a
better
name.
I
think
it's
a
snark
pack
or
something
like
this.
It's
a
way
to
package
snacks
in
a
single
proof.
I
So
maybe
we
will
stop
using
ipp.
We
started
calling
it
snack
pack
and
the
idea
is
that
you
can
put
multiple
proofs
into
one
proof,
and
so,
instead
of
paying
for
gas
1000
times
for
1000
proofs,
you
only
pay
as
close
as
it
gets
to
one
proof.
So
it's
massive
gas
saving
for
miners.
We
added
a
small
issue,
was
an
economic
issue
which
is
we
don't
want
small
miners
to?
They
will
save,
of
course,
but
a
small
miner
could
save
less
than
much
miners.
I
The
difference
was
between
2x
small
miner
and
the
large
miner
would
be
able
to
save
up
to
20x
and
because
of
this
and
other
analysis
that
that
we
did,
we
decided
that
we
don't
want
to
allow
the
full
potential
of
snack
pack
to
everyone,
and
this
means
that
you're
not
going
to
be
able
to
to
pack
thousands
of
proofs
we're
going
to
limit
this
to
200
proofs
and
so
in
a
way
we're
limiting
the
power
of
the
technology
to
make
sure
that
it
is
fair
but
at
the
same
time
we're
also
extending
the
submission
window
between
a
pre-commit
and
approve
commits
from
one
day
to
six
days.
I
This
means
that
a
minor
can
now
accumulate.
Many
proof
commits
and
then
submit
a
single.
A
single
proof
commits
this
make
sure,
because
we've
seen
that,
according
to
the
stats,
the
average
small
miner
reaches,
maybe
200
proof
commits
in
in
six
days,
and
they
will
definitely
not
be
able
to
generate
wonder
proof
commits
in
a
single
day,
and
we
think
that
this
is.
I
This
is
already
like
a
huge
saving
by
itself
and
also
due
to
some
gas
things
that
I'm
not
really
understanding,
because
I
didn't
look
into
it
looks
also
that
200
proves
it's
some
sort
of
plateau
improvement,
because
even
if
you
can
aggregate
more
than
200
proofs
you
still,
the
the
cost
of
the
cost
of
gas
for
state
reads
starts
to
become
really
high
yeah,
so
snack
pack
is
happening,
we
have
the
ffp
is
being
fixed,
has
been
fixed,
and
so
you
should
see
everything
that
we
are
discussing
into
into
the
fip.
I
Maybe
the
name
is
still
iqp
we're
going
to
change
it
to
snapback.
It's
also
friendlier,
and
the
second
thing
is,
but
we're
probably
going
to
postpone
this
change
is
a
pretty
commit
one.
So
the
smartpak
saves
massively
proof
commits
messages
on
chain.
You
will
basically
go
from
the
amount
that
we
have
today
to
a
potential
of
maybe
10x
less.
I
This
is
like
a
maximum
idea
and
we
can
do
something
similar
for
pre-commits
and
we
do
have
this
fap,
but
we,
we
probably
won't
be
able
to
ship
this
ffp
at
the
same
time
as
the
snack
pack.
It
may
be
also
risky
to
integrate
two
different
changes
of
this
kind,
but
this,
like
heads
up
multiple
implementations
we
may
have.
We
may
have
a
pre-commit
batching,
maybe
not
in
these
updates
in
the
next
one.
A
That's
good,
can
we
say
definitely
not
in
the
june
upgrade
yeah.
Definitely
not
sorry
sounds
good
yeah,
so
we'll
we'll
move
along
that
as
the
momentum
builds
for
it,
but
right
now,
basically,
we
only
have
attention
for
13.
A
Oh
yes,
sir
nicole,
I
didn't
ask
you
to
to
share
your
updates.
I
also
did
not
ask
the
falcon
foundation
to
share
their
updates,
so
if
you
all
would
like
to
jump
in,
please
do
so
sorry
about
that.
J
A
I
would
say
that
a
lot
of
the
phil
pole
stuff
was
driven
by
the
foundation,
which
was
great,
because
that
is
the
right
place
where
for
stuff
like
this
to
be
driven.
So
it's
nice.
I
think
a
lot
of
this
just
feels
like
we're
taking
steps
in
the
right
direction
and
you
know
moving
towards
much
more
decentralization,
which
is
exactly
what
we
want.
A
Yay
cool
any
other
questions
or
comments.
H
I
just
wanted
to
introduce
jorge
to
the
team.
You
know
the
implementers
team
he's
joined.
You
know
the
first
project
and
yeah
really
excited
to
have
a
mom.
K
K
But
and
I've
been
kind
of
working
in
startups
and
in
the
blockchain
ecosystem
for
a
while
now
and
been
watching
filecoin
from
afar
ever
since
juan
kind
of
started
releasing
like
the
specs.
So
I'm
happy
to
join
the
effort
and
yeah
participate.
A
All
right.
I
think
we
can
wrap
up
there
if
no
one
has
anything
else.
I'd
like
to
worry
about
cool
bye.
Everyone
thanks
for
joining.