►
From YouTube: Governance and Risk Meeting #138
Description
## Introduction
* @LongForWisdom: Hosting, Introduction, Agenda.
## Weekly Updates
* Governance - @Gov-Facilitators + Contributors
* Protocol Engineering - @Smart-Contracts
* Oracles - @Oracles
* Risk - @Risk
* Real World Finance - @SebVentures
* Operational Support - @operational-support
* MIPs - @MIP-Editors
## Procedural
Monthly MIPs Poll Review - @Gov-Facilitators
## Resources
Website: https://makerdao.com
Twitter: https://twitter.com/makerdao
Forum: https://forum.makerdao.com/
Chat: https://chat.makerdao.com/home
Email: info@makerdao.com
Governance and Community Events Calendar:
https://calendar.google.com/calendar/u/0/r?cid=bWFrZXJkYW8uY29tXzNlZmhtMmdoaXBrc2VnbDAwOWt0bmlvbWRrQGdyb3VwLmNhbGVuZGFyLmdvb2dsZS5jb20
A
My
name
is
long
for
wisdom.
I'm
the
governance
facilitator
at
make
it
dao.
Currently
I'm
joined
by
a
bunch
of
mega
people,
both
both
those
heavily
involved
and
those
not
so
much
as
usual.
We've
got
a
agenda
to
get
through
we're
gonna
have
some
sort
of
general
updates
from
the
usual
suspects,
we'll
kind
of
review,
the
monthly
governance
cycle
or
where
we're
currently
out
in
the
month
we
go
in
the
cycle.
A
A
B
B
Our
latest
executive
did
pass
with
some
critical
bug,
fixes
that
will
likely
be
addressed
later.
In
this
call,
we
still
have
seven
collateral,
green
light
polls
running
for
the
next
10
days,
so
vote.
If
you
haven't
there-
and
we
did
have
two
polls
that
did
conclude
this
week
before
the
call
both
of
them
passing,
that
would
be
the
wi-fi,
a
liquidations,
2.0
parameters
and
the
mips
monthly
cycle.
B
Likewise,
good
progress
on
the
mips
portal
we're
looking
to
get
some
additional
functionality
there
with
filtering
options.
So
if
you
haven't
checked
that
out
in
a
while,
progress
is
coming
on
pretty
nicely
and
finally
just
wanted
to
give
update
on
source
cred,
as
well
as
a
shout
out
to
matt
from
success
there
for
giving
a
really
good
forum
post
on
asking
about
increasing
the
distributions.
Basically,
we
we
work
pretty
closely
with
the
source
cred
team,
on
setting
parameters
and
kind
of
having
some
forthcoming
changes
there.
B
A
C
Thank
you
long.
I
hope
you
rested
your
son
arrested,
as
someone
pointed
out,
so
my
name
is
eli,
I'm
doing
the
forum
in
advance
for
you
guys
this
week
and
first
up
a
three
point
summary:
what
do
I
need
to
know
as
an
mkr
holder
this
week?
Well,
you
probably
want
to
know
that
the
mkr
price
hits
an
all-time
high
today
with
4
878,
the
system
surplus
buffer
is
steadily
steadily
filling
up
and
now
sitting
at
26.7
million
and
the
total
die
supply
is
also
growing.
C
Currently
at
3.6
billion
on
to
announcements,
the
liquidations
2.0
upgrade
is
now
voted
in
feel
free
to
check
out
the
linked
blog
post
to
see
what
this
means
for
maker
dao
and
the
community
and
another
post
kurt
and
the
smart
contracts
team
go
into
detail.
Regarding
the
reasons
for
the
recent
executive
vote
delay
onto
active
discussions,
we
have
quite
a
few
first
one
by
subventures,
with
the
first
real
world
asset
operational,
which
is
new
silver.
The
rwa
team
discusses
strategy
for
2021.
C
another
one
by
premarch,
who
presents
an
analysis
of
possible
scenarios
resulting
from
different
liquidations,
2.0,
auction
durations
and,
lastly,
won
by
al
progreso,
who
starts
this
discussion
on
various
emergency
shutdown
attack
vectors
when
it
comes
to
single
requests.
We
have
two
ongoing,
so
please
go
to
the
forums
and
vote
the
first
one
by
alexis
who
is
looking
to
adjust
the
pack's,
stablecoin's
debt
ceiling
and
secondly,
one
by
monash
supply,
who
is
seeking
consensus
on
reducing
the
liquidation
ratio
on
the
uni
v2
die.
Usdca
fault!
D
Yeah
thanks
a
lot,
so
I'm
gonna
try
something
a
little
different
today.
I
feel
like
for
this
call
the
the
like
big
rattling
off
of
our
entire
task
list
as
though
it's
like
a
stand-up
is
getting
unwieldy.
So
I
think
I'm
just
going
to
talk
about
sort
of
what
we've
delivered
and
the
sort
of
big
items
that
we're
working
on
coming
up
but
know
that
there's,
like
all
that
sort
of
smaller
minutiae
reviews
and
all
of
that
stuff
is
sitting
in
there
as
well
so
yeah.
D
First,
I
wanted
to
thank
everybody
that
that
did
vote
on
the
executive
and
that
it
finally
passed.
We
were
getting
a
little
bit
worried
there,
but
that
that
was
that's,
probably
like
12
months
worth
of
work
to
finally
ship
liquidations,
2-0
tons
of
cross-team
coordination
just
in
an
insane
lift.
There
are
commits
in
there
from
team
members
that
that
left
long
ago
and
and
so
we're
we're
thankful
that
it's
it's
finally
going
out.
D
Also
the
the
the
two
additional
bug
fixes
in
the
esm.
In
the
end,
that
was
were
mentioned
that
kerp
posted
were
in
that
fix
as
well.
I'm
you
know
more
length
about
those
if
anyone
wants
to
discuss
those
later
but
yeah.
All
of
that
I
don't
know,
I'm
just
I'm
very
grateful
that
it
went
out
and
thanks
everybody.
It's
a
lot
of
work.
So
so
in
the
I
guess,
what
we're
calling
like
the
death
march.
D
We
we've
gotta
now
ship
liquidations2o
on
this,
pretty
tight
cadence
from
in
covan
to
mainnet,
for
probably
the
next
like
month
to
month
and
a
half,
and
so
in
that
vein
we
are.
We've
currently
put
a
wi-fi
on
covan,
which
is
the
second
collateral
and,
and
we
are
beginning
to
construct
the
main
net
spell
for
it
now,
in
addition
to
wi-fi
going
out
for
the
main
net,
spell,
there's
gonna
be
a
bunch
of
other
changes
as
well.
D
I
think
mostly
because
we
have
a
large
backlog,
so
we're
gonna
see
a
ton
of
work
from
the
rates
group
and
some
debt
ceiling
changes
and
instant
access
modules
and
whatnot.
So
all
of
that's
going
into
that
spell-
and
so
that's
that's
a
decent
amount
of
work
and
then,
hopefully,
on
friday,
we'll
begin
and
deploy
to
kovan
the
all
of
the
f
collateral
types
in
wbtc
in
preparation
for
mainnet
next
week.
D
A
Great
thanks
chris
yeah,
just
to
briefly
comment
on
the
on
the
25
thing.
We
have
kind
of
make
very
clear
in
in
the
signal
requested
in
the
polls
that
it's
very
much
like
a
like
an
approximate
target
rather
than
their
guarantee.
So
hopefully
everyone
has
has
clear
expectations
around
that,
but
it's
yeah.
It's
always
good
to
get
a
reminder.
A
E
F
D
Thanks
yeah,
I
did
actually
want
to
bring
up
one
more
thing
which
I
you
know
maybe
long.
This
is
more
in
your
camp,
but
we
have
one
of
those
really
densely
compact
executive
runs
coming
up,
so
we've
got
the
friday
executive
and
then
the
monday
executive.
So
really
this
wi-fi
all
these
rate
changes
and
everything
needs
to
be
voted
in
over
the
weekend.
E
E
Yeah,
maybe
we
should
mention
that
we
are
probably
gonna
have
to
put
a
either
a
time
limiter
or
some
kind
of
limiter
on
the
the
spell
for
this
friday.
Just
so,
there's
not
a
race
condition
with
the
existing
dog.
That's
going
out
with
the
office
hours
on
it,
so
that'll
get
passed
sometime
like
monday
morning.
We
just
need
to
make
sure
that
the
the
spell
that
we
deploy
this
friday
doesn't
get
deployed
or
it
doesn't
get
executed
before
the
other
ones.
E
A
G
They
just
released
a
a
new
new
feature
that
they've
had
in
the
works
for
quite
a
while
right,
as
this
furl
combo
style
thing
right,
where
you
can
like
compose
transactions,
so
they're
dealing
with
some
some
bugs
in
production
that
they're
kind
of
scrambling
to
fix,
but
definitely
think
we
can.
We
can
have
them
up
and
running
early
next
week
and
hopefully
include
it
in
in
next
week's
executive.
G
So
that's
exciting.
I
definitely
think
we
should
probably
be
looking
to
onboard
like
another
fee
like
another
wave
of
like
light
feeds
anyway.
I
I
kind
of
think
that,
as
the
protocol
right
scales
right,
there's
more
die
supply,
there's
more
collateral
right.
You
want
to
increase
right,
the
the
amount
of
kind
of
signatories
on
on
this
multisig
right
and
and
just
increase.
H
Well,
I
forgot
exactly
what
the
light
feed
is.
I
mean
I
use
d5
saver
for
leverage
and
like
boosts
and
repays
you
at
different
levels,
but
it's
like
a
two-hour
delay
when
the
osm
kicks
in
so
is
the
likely
gonna
somehow
help
them
execute
those
things
like
at
a
market
price,
or
is
this
just
like
unrelated
to.
G
F
Hey
nick,
I
wanted
to
go
back
a
few
months
ago,
when
you
mentioned
you
brought
up
the
the
raising
the
oracle
sensitivity
for
some
of
the
collaterals,
and
I
brought
this
up
to
pre-motion
about
link.
I
know
you
had
mentioned
it
back,
then
that
they
were
one
of
the
candidates
and
he
mentioned
that
they
did
raise
it
to,
or
you
guys
raised
it
to
four
percent.
G
So
just
to
kind
of
validate
a
lot
of
your
assumptions.
Yeah
we
did
raise
the
price
sensitivity
to
four
percent
right,
which
means
that
the
price
has
to
change
by
four
percent
before
we'll
even
update
the
medianizer
right,
and
we
did
that
because
the
right
just
doing
doing
a
one
percent
sensitivity
right
was
just
way
way
way
too
expensive
and
yeah
and
that's
worked
out.
That's
worked
out
extremely
well
right
for
things
like
eth
and
btc
right.
We
have
one
percent
and
those
vaults
are
rather
large
right.
G
F
A
F
A
F
G
G
We
can
just
basically
we
can
try
to
be
a
bit
more
aggressive
on
on
the
gas
cost
pricing
right,
but
it's
it's!
It
just
comes
down
back
to
this.
Like
cost
function
right,
so
the
frequency
of
updates
versus
you
know
how
much
in
gas
costs
you're
willing.
You
know
how
aggressive
you're
willing
to
be
on
gas
cost
to
get
it.
You
know
right
on
the
dot
versus
how
much
money
are
you
willing
to
save
if
you'll
go,
you
know,
plus
three
minutes
plus
five
minutes.
G
K
That
might
seem
a
bit
short
to
some
people,
so
we
made
an
analysis
to
show
how
potentially
delayed
bids
could
impact
maker
or
world
owners,
and
we
found
out
that
even
if
auctions
were
to
be
were
to
last
up
to
hour
and
half
in
some
cases,
even
two
hours.
So
if
we
would
have
you
know
more
than
one
hour,
delay
maker
would
still
be
safe,
so
no
no
losses
for
maker.
K
I
also
want
to
know
that
there's
there's
one
other
reason
why
you
know
makes
sense
to
propose
same
set
of
parameters
for
for
it
on
red
bitcoin,
as
we
did
for
link
and
wifi
is
because
we
may
not
want
to
create
some
kind
of
disorder
between
keepers
when
they
set
up
their
their
software
to
beat
at
least
in
this
initial
testing
phase.
So
in
testing
phase,
probably
better
to
have
four
parameters
unified
and
then
we
start
with
them.
K
K
And
then
I
think
it
is
already
mentioned.
We
had
a
debate
among
commended
detectors
about
increasing
the
surplus
buffer
with
the
alert
module
we
might
potentially
hit
it
in
about
eight
days,
maybe
even
even
sooner,
if
we
have
some
regulations-
and
my
suggestion
was
to
enable
it
already
this
week
so
tomorrow
because
of
the
the
issue
that
actually
chris
mentioned.
So
if
we
have
a
large
degradation
event
in
front
of
us,
you
know
we
could
collect
million
million
of
these
or
even
more,
and
then
I
think,
there's
about
four
hundred
thousand
of
fees.
K
We
die
because
it
seems
there's
no
other
good
way
to
to
make
hilton
packs
currently
and
maker
doesn't
charge
any
stability
film
packs
because
of
the
abandoned
faults
issue
that
we
know,
and
I
think
we
might
propose-
probably
the
next
rate
proposal-
sorry
ram
parameter
proposal
meeting
to
to
set
debt
ceiling
of
this
world
type
to
zero
because
it
doesn't
really
add
any
benefit.
Any
benefits
to
maker
currently
and
yeah.
That's
pretty
much
it
from
from
our
side.
This.
A
K
I
mean
with
liquidations
2.0
it's
much
less
seriously.
I
mean
we
still
need
to
see
how
they
perform,
but
I'm
much
more
optimistic
and
there's
one
other
thing
that
maybe
is
relevant
here.
To
mention
that
you
know
most
of
the
risks
or
large
majority
is
related
to
liquidity
and
because
heater
price
went
higher
and
there's
more
farming,
there's
more
dying
on
the
centralized
venues.
You
know
this
risk
is
getting
lower,
so
you
know,
even
though
we
are
seeing
increased
exposure,
the
liquidity
metrics
are
much
better,
so
it's
kind
of
correlated.
A
Okay,
cool-
and
so
I
guess
like
a
princess
to
a
relationship
which
is
maybe
more
smart
contracts
but
like
given
that,
like
we're
not
kind
of
guaranteed
to
see
liquidations
for
like
or
wi-fi
before,
I
guess
the
point
where
we
vote
on
ether
and
wbtc
for
the
relationship.
Are
we
gonna
wait
until
we
kind
of
confirm
that
things
are
working
with
lincoln
wifi
before
pushing
allocations
to
2.03.
E
I
I
would
love
that
honestly,
we're
kind
of
right
right
now
working
on
this
governance
schedule.
That's
not
entirely
everyone
design.
So
maybe
maybe
that's
a
question
that
we
asked
the
broader
audience.
Can
we
can
we
wait?
I
think,
there's
a
lot
of
concern
that
there's
going
to
be
some
large
market
drawdown
here
in
the
next
month
or
two
so.
L
E
M
L
We've
just
we've
discussed
doing
some
mainnet
tests
with
the
integrations
team
just
to
test
their
keepers
and
the
system
and
the
uis
that
we'll
have
in
place
going
coming
down
the
line.
So
I
think
that's
a
good
idea
for
us
to
do.
I
can
I
can
work
with
integrations
and
communicate
to
you
guys
what
our
plan
for
that
is.
M
Yeah
sure
so
the
biggest
news
is
obviously
new
silver.
We
have
our
first
investment
in
real
world
assets
such
here
thousand
died.
It
worked.
Let's
hope
it
worked
in
the
other
way
as
well.
I
will
start
that
soon.
I
want
to
thank
the
smart
contract
team
that
has
worked
a
lot
on
that,
as
well
as
a
centrifuge
team.
M
They
spent
a
lot
of
time
together
to
to
craft
the
spell
and
all
the
meep
22
and
the
making
sure
that
newtonian
one
is
working
fine.
So
that's
good.
We
had
a
nice
piece
on
coin
desk
as
well
with
that,
and
there
was
some
some
pr
activities.
So
that's
always
good,
so
people
can
see
that
maker
is
always
innovating,
so
we
did
innovate
with
the
uni
swap
lps
earlier
this
year
now
with
real
world
assets.
M
A
M
So
yeah,
so
I'm
also
so
it
was
delayed
a
bit
this
week,
but
we
are
having
a
law
firm,
which
is
called
arnold
porter.
They
have
partners
which
is
in
new
york
that
has
worked
on
a
lot
of
asset-backed
securities
and
all
those
criteria,
legends
trust.
They
know
that
stuff.
So
they
will
help
us
on
this
one,
but
we
still
need
to
have
everything,
because
if
I'm
correct
that
we
don't
have
the
trust
agreement,
for
instance,
last
time
I
checked,
but
there
is
maybe
too
much
activity
on
the
form.
M
J
While
I
encourage
the
review,
I
do
encourage
the
community
to
appreciate
the
notion
that
all
of
the
transactions
come
with
a
legal
opinion
to
ensure
enforceability.
So
the
salient
points
related
to
the
business
terms
should
really
really
be
the
focus.
The
legal
opinions
are
what
the
community
and
I,
as
well
as
the
trust
companies,
are
relying
upon.
F
F
F
J
J
Yeah,
but
in
general,
the
way
this
just
has
a
big
picture
and
I'll
be
exceptionally
brief
on
this.
The
way
this
big
picture
works
is
that
we
trust
the
trust
we
trust
the
trustee.
That's
the
net,
that's
the
notion
of
it,
and
there
is
a
credit
agreement
that
would
be
executed
between
lendco
and
the
trust
and
transactions
that
fall
within
the
box
within
the
scope
of
the
credit
agreement
are
things
that
could
be
lent
against
from
the
trust
to
the
lend
co
if
it
doesn't
fit
within
the
box.
The
trustee
won't
extend
credit.
F
J
Well,
I
I
won't
speak
for
centrifuge
and
I'll
speak
only
for
6s.
On
this
notion,
I
would
say
on
all
scenarios
where,
in
the
v1
v1
version
for
all
of
us,
everybody
will
evolve.
The
structure
it'll
probably
take
three
total
iterations
before
we
get
to
a
place
where
all
structures
are
harmonious
and
we
know
exactly
how
we're
working
we're
going
to
get
to
a
place.
O
Yeah,
I'm
just
dropping
our
weekly
update.
I
won't
go
too
in
detail.
I
just
want
to
remind
people
of
the
up
and
coming
calls
that
we
have
the
know.
Your
mip
direct
deposit
module
is
coming
up
next
week,
pretty
exciting
and
then
for
collateral
onboarding
we
have
mco2,
which
is
with
the
carbon
footprint
that'll,
be
next
wednesday
as
well,
and
then
last
week
we
had
a
few
calls
from
the
launchpod
series,
which
was
governance,
communications
and
the
sustainable
ecosystem.
Scaling
both
are
have
links
to
the
recording.
O
I
wanted
to
just
bring
out
one
interesting
topic
of
conversation
that
seth
brought
up,
which
is
around
intentional
organization,
and
what
we
can
do
to
communicate
better
in
the
dao.
I
think
we're
growing
all
the
time
and
communication
is
definitely
one
of
the
most
important,
but
also
challenging
aspects
of
of
any
community,
so
feel
free.
To
put
your
ideas,
if
you
disagree,
that's
great
too
part
of
it
is
just
that
we're
trying
to
figure
out
something
that
works
for
our
community
and
how
we
might
communicate
better.
O
O
A
I
Okay,
okay,
I
thought
my
mic
had
a
problem
but
yeah.
Basically,
yesterday
we
had
a
presentation
with
the
with
the
community
where
we
presented
a
bit
what
we
are
trying
to
achieve
and
how
and
we
are
looking
forward
to
to
feedback
and
hopefully
to
get
everything
set
up
for
the
next
for
the
next
round.
N
N
N
N
So
if
that
executive
doesn't
pass
after
the
four
day
limit
expires,
it
is.
It
does
mean
that
the
proposals
will
effectively
be
kind
of
rejected
from
that
one's
govern
cycle.
So
once
it
goes
up,
don't
forget
to
vote
and
then
on
thursday,
we'll
have
long.
Do
the
governance
cycle
review,
highlighting
everything?
That's
happened
over
the
course
of
the
month
and
possibly
discuss
the
upcoming
may
government
cycle,
because
there
are
quite
a
lot
of
eligible
proposals
for
next
month.
N
In
terms
of
some
other
updates,
we
have
the
community
greenlight
polls,
there's
seven
of
them.
They
went
live
on
on
monday
and
they'll
be
running
until
may
3rd,
so
also
don't
forget
to
vote
for
those
as
of
now.
There
are
26
proposals
in
in
request
for
comments,
and
a
lot
of
them
have
actually
been
actively
improved
on
by
the
authors,
so
those
changes
have
been
incorporated
on
the
forums
and
and
on
github,
and
most
of
these
proposals
are
eligible
for
next
month's
govern
cycle.
N
There
were
no
new
proposals
submitted
this
past
week,
but
there
was
one
new
proposal
idea
in
the
forums
for
for
basically
redone
renominating,
the
like
mkr
token
paid
in
the
adoption
and
increase
like
the
overall
backstop
potential,
there's
a
decent
amount
of
activity
in
that.
So
if
you
are
interested
in
checking
it
out,
it
is
on
the
forums.
I
can
also
drop
a
link
to
it,
but
yeah.
That
concludes
my
summary
today
and
yeah
I'll
pass
it
off
to
long
to
do
that.
Governance
poll
review.
A
A
There's
also
a
declaration
content
proposal
for
eurodye
and
a
number
of
amendments
to
various
mips,
including
four
mid-nine
and
m0.
If
zero
members
are
pretty
minor,
the
mid-four
amendments
are
a
little
bit
more
involved,
but
also
not
groundbreaking,
and
then
the
minor
moments
are
just
adding
setting
it.
So
we
have
green
light
poles
twice
a
month
rather
than
once
a
month.
A
N
A
H
I
mean
there's,
there's
a
pretty
detailed
forum
post
up,
so
you
know
we
can
certainly
go
into
detail
here
too.
People
want
to
talk
about
it.
Maybe
I'll
give
this
super
high
level
overview
in
case
anybody's,
not
caught
up
on
the
news
so
shipped
with
the
liquidations.
2.0
spell
that
just
got
voted
end
were
some
bug,
fixes
to
the
emergency,
shutdown,
module
or
esm,
and
also
the
end
module,
which
is
the
smart
contract
that
orchestrates
the
whole
global
settlement
process
after
an
emergency
shutdown.
H
So
those
were
two
things
that
had
to
be
replaced
anyway,
with
liquidations
2.0
and
in
the
process
of
scrutinizing
them.
We
found
some
issues.
First
of
all,
with
the
emergency
shutdown
module,
it
didn't
revoke
the
access
of
governance
to
certain
aspects
of
the
system,
which
would
mean
that
if
it
were
triggered
and
I'm
sort
of
assuming
people
know
what
the
emergency
shutdown
module
is,
I
guess
maybe
I
should
clarify
that
too.
H
So
that
was
the
fix
we
did
there,
which
was
to
you
know,
kind
of
completely
shut
governance
out.
If
the
esm
gets
triggered,
which
you
know
is,
is
largely
an
improvement.
The
like
one
edge
case
where
it's
not
an
improvement
is
if
there's
some
sort
of
bug
in
the
emergency
shutdown
process
itself,
that
governance
could
theoretically
intervene
and
fix,
which
turns
out
was
exactly
the
kind
of
bug
we
found
at
the
end.
So
we
had
to
fix
it
at
the
same
time.
H
So
the
end
had
a
bug
where,
if
there
was
a
certain
amount
of
collateral
kind
of
reserved
to
cover
die
and
that
exceeded
a
of
a
particular
value,
you'd
get
an
overflow
and
the
end
process
would
get
stuck
like.
I
said.
Governance
could
have
stepped
in
to
fix
that.
But
since
we
were
changing
the
esm
to
completely
shut
governance
out
after
a
shutdown,
it
was
very
important
to
fix
that
bug
in
the
end
as
well.
So
so
that
was
addressed
with
a
small
tweak
to
the
math.
H
A
H
Yeah,
so
to
answer
the
question
of
how
were
these
caught
in
the
first
place
and
why
weren't
weren't
they
caught
sooner
and
there
were
you
know
there
will
be
full
post-mortems
at
some
point
in
the
near
future,
we're
a
little
bit
over
subscribed
at
the
moment.
So
I
don't
want
to
promise
exactly
when
they'll
come,
but
you
know
coming
soon
tm.
H
But
how
were
these
caught?
The
esm
bug
was
just
sort
of
caught
through.
I
I
forget
exactly
the
discussion
that
led
to
it,
but
basically
it
was.
It
was
some
discussion
we
were
having
about
how
it
worked
as
we
were
looking
at
it
to
redeploy
it,
and
we
figured
out
that.
Oh
hey,
here's,
here's
an
issue,
I'm
so
you
know
like
it
would
have
been
great
if
that
could
have
been
caught
a
lot
sooner
and
that
kind
of
comes
down
to
there.
H
Just
wasn't
like
a
good
enough
specification
for
the
esm
originally
and
it
would
have
been
well
it
was.
It
was
also
around
the
time
that,
like
the
system
was
kind
of
first
being
deployed,
and
so
the
governance
system
wasn't
even
necessarily
well
defined
or
finalized.
That
was
still
a
time
when
we
were
like
well,
maybe
there
will
be
like
several
pauses
and
maybe
there
will
be
in
the
future.
It'll
keep
evolving,
but
you
know
so
the
sort
of
lesson
from
that
one
is.
H
You
need
better
specification
upfront
for
the
for
the
requirements
of
these
things
and
then,
of
course,
we
can,
you
know,
add
test
coverage
for
that
now,
which
we
have
then
for
the
the
end
bug.
You
know,
why
wasn't
that
one
caught
sooner?
How
was
it
caught
now?
You
know
it
was
caught
now
because
we
were
writing
tests
to
do
the
emergency
shutdown
process
with
the
link
a
executive
applied,
and
we,
you
know,
by
doing
those
tests,
we
said:
oh
hey,
they're,
they're
getting
stuck
and
they're,
not
working.
You
know.
H
Why
is
this
reverting
and
that's
so
that's
how
we
found
it.
So
the
question
comes:
how
could
you
find
that
sooner?
You
know
we
have
a
few
different
things
we
want
to
do.
One
is
you
know,
kind
of
continuous
integration
tests
that
are
running
all
the
time
against
main
net
state
or
not
all
the
time,
but
periodically
right.
So
if
we
had
had
those
before,
then
we
could
have
kind
of
caught
this
bug.
H
H
You
know,
formal
verification
things
where
you
say
here
are
the
expected
magnitudes
for
values
in
the
system
and
then
run
some
automated
checking
tool.
That
says,
okay
do
any
of
all
any
of
these
intermediate
calculations.
You
know
overflow
and
revert
if
the
magnitudes
are
in
those
expected
ranges
right.
So
that's
that's
another
thing
we
want
to
do.
A
Awesome
yeah.
Thank
you
very
much
for
that
summery
car,
that's
great
yeah!
I
I
think
I
kind
of
read
the
post
sort
of
talking
about
the
sort
of
having
continuous
integration
tests
for
like
for
like
shutdown,
and
I
guess
for
other
stuff
as
well,
and
that
sounded
like
kind
of
really
really
good
idea
to
me.
That's
that's
cool.
I
kind
of
like
the
like
the
idea
of
just
having
like
a
sandbox
somewhere
running
like
constantly
shutting
down
the
protocol
to.
H
Make
sure
everything
works
yeah,
you
know
running
against.
You
know
the
current
state
and
values,
one
main
net,
because
you
know
you
you
might
say
well,
wasn't
emergency
shutdown
tested
before
well?
Well,
it
was
tested,
it
was
tested
in
unit
tests.
It
was
tested
on
covan,
it
was
even
tested
on
mainnet.
H
You
know,
values
that
were
too
small
to
trigger
the
bug,
and
so
that
was
why
so
yeah.
So
that's
why
you
know
continuous
integration
test
is
really
good,
so
you
can
see
it
as
soon
as
it
happens,
but
it's
it's
still
not
ideal,
because
you
know
you're
you're
having
to
find
that
it's
happened.
You
know
on
mainnet
or
something
so
so,
while
I
think
that
those
are
great
and
we'll
do
them,
I
don't
think
they're
a
complete
solution.
You
know,
I
think
you
need
like
this
better.
H
You
know
magnitude
checking
earlier
in
the
dev
process,
where
you
don't
just
write
some
unit
tests
with
any
old
values.
You
you
write
some
tests
or
do
some
verification
where
you
check
a
whole
range
of
values,
and
you
say:
are
there
any
problems
if
any
of
these
parameters
or
quantities
are
anywhere
in
these
specified
ranges.
D
Yeah
I'll
call
it
in
the
sidebar
that
I
didn't
mention
during
the
segment,
but
we
have
a
full
response
to
the
mip
45
audits
that
was
posted
in
the
forums
today,
also
primarily
authored
by
kurt.
So
we
took
a
little
while
to
collect
all
of
our
thoughts
and
then
kurt
did
a
really
good
job
of
sort
of
describing
what
each
issue
in
the
audit
was
and
then
why
that
particular
issue
was
either
resolved,
fixed
addressed
in
a
certain
way
or
wouldn't
be
resolved,
and
so
it's
a
pretty
good
read
through.
H
A
A
H
Yeah,
we
didn't
just
want
to
publish
the
audit
reports
and
have
people
like
you
know
asking
questions
of
well.
Was
this
issue
addressed?
Was
it
not
addressed,
or
you
know
potentially
even
like?
What
does
this
issue
even
mean?
You
know
because
they
refer
to,
like
very
you
know,
detailed
things
about
the
system.
A
A
F
Yeah,
I
could
speak
to
that
so
in
this
scenario,
where
there's
sort
of
a
malicious
delegate
who
has
a
large
amount
of
power,
they
have
to
maintain
that
voting
power
for
two
days
to
secure
the
hat.
So
in
that
event,
delicate
delegators
could
be
made
aware
of
this
and
they
can
pull
their
maker
out
in
time
to
reverse
the
vote.
So
I
don't
think
this
is
a
terribly
likely
scenario
but
something
to
be
concerned
about
for
sure.
A
H
You
know
obviously
long
so
that
it's
not
too
inconvenient,
but
so
that
you
know
if
a
bunch
of
people
delegate
to
one
delegate
and
then
they
you
know
either
start
misbehaving
or
you
know,
just
like
those
people
sort
of
wander
off
and
completely
forget
things
just
to
make
sure.
There's
not
you
know
perpetual
maximum
voting
power
there.
There
were
some
kind
of
expiries
put
in.
D
Yeah,
it's
a
real
problem.
I
and
I'm.
I
know
that
we
tried
to
address
it
in
dss
gov,
but
the
type
of
delegation
we're
going
to
have
is
sort
of
the
minimum
viable
implementation
of
delegation
right
so
we're
we're
kind
of
like
hacking
up
a
vote
proxy
in
order
to
allow
for
delegation,
and
I
don't
think
that
we
had
any
kind
of
like
vote
decay
in
there.
D
Sam
might,
I
know
you
were
the
one
that
like
touched
most
recently,
there's
no
vote
decay,
but
I
don't
think
we
really
need
it
because
it
will
need
to
we're
redoing.
The
delegation
when
dss
gov
comes
out,
so
the
decay
is
effectively
when
we
either
upgrade
the
chief
or
switch
to
dss
gov,
which
I
think
is
sufficient
for.
F
D
D
Yeah,
if
anyone
else
subjects
to
that,
you
should
definitely
speak
up,
because
it's
it
really.
I
mean
I,
you
know
I've
seen
this
manifest
in
the
real
world
it,
but
the
I
think
the
worry
is
so.
Sam
takes
one
angle
of
it.
You
know
like
okay,
well,
if
someone's
acting
maliciously,
then
the
people
that
are
delegating
can
pull
out
and
redelegate
and
stop
them.
D
The
the
real
danger
is
the
the
mkr
voters
that
are
delegating
are
apathetic
or
die
or
lose
their
keys
or
whatever,
and
then
that
that
person
with
the
delegation
power
is
sort
of
entrenched
and
irremovable,
and
so
that's
the
that's.
The
like
real
failure
scenario
of
delegating
without
decaying
those
delegates.
So.
D
It's
more
just
that
you
know
at
a
certain
threshold
like
if
I've
delegated
to
alice
right,
then
at
some
point,
like
my
delegation.
After
let's
say
six
months
or
some
threshold
of
time
either
I
don't
think
you'd
lose
it
over
time.
So
maybe
like
decaying,
isn't
the
right
term.
But
let's
say
at
a
particular
cliff
my
delegation's
no
longer
effective
against
alice.
D
I
have
to
re-up
my
delegation
periodically,
which
gives
me
at
least
some
sort
of
minimal
participation
in
the
protocol
and
doesn't
allow
like
early
delegates
to
basically
monopolize
a
bunch
of
like
locked
or
lost
mkr
or
or
like
totally
apathetic
voters.
Although
I
know
that's
kind
of
why
we're
adding
delegation
is
because
and
care
holders
apparently
want
to
be
a
little
more
apathetic.
But.
A
E
I
want
to
mention
that
I
I
think
I
can
speak
for
everybody
on
the
protocol
engineering
team.
We
have
no
intention
of
being
delegates
for
this.
We're
writing
the
code.
We're
publishing
the
code.
I
think
it's
too
too
much
power
in
our
hands
if
we
were
also
delegates
and
and
executing
the
code
as
well.
So
this
is
entirely
in
the
hands
of
the
community.
F
D
It's
permissionless
right,
so
I
mean
I
mean
someone
on
the
let's
say
the
protocol
engineering
team
could
put
their
name
forward.
I
just
think
it
would
be.
I
think
it
would
be
a
really
bad
idea
for
mkr
holders
to
delegate
to
them,
even
though
it
seems
like
they
know
the
protocol.
It's
just
that's
way
too
much
power.
E
H
That
but
I'm
not
even
saying
it's
a
problem,
I'm
just
saying
that
it
was
just
suggested
that
yeah,
I'm
just
saying
that
you
can't
avoid
that
that
sort
of
scenario
I
mean
you
also
can't
avoid
collusion
scenarios
where
somebody
comes
up
and
says:
hey,
I'm
a
completely
different
person
delegate
to
me
and
they
have
a
real
identity.
But
you
know
in
actuality
they're,
just
taking
like
you
know,
direction
from
somebody
who's.
So
there's
there's
like
a
byzantine
element
here
to
it
that
you
can
never
totally
exclude
yeah,
but.
A
So,
just
to
speak
a
little
bit
for
that,
if
you
don't
mind,
because
we're
kind
of
planning
to
have
gov
alpha
do
like
some
level
of
like
verification
of
delegates
that
are
willing
right.
So
I
mean
you
know,
obviously
we're
not
going
to
force
anyone
to
doesn't
want
to,
but
there
will
be
a
point
where
we
sort
of
start
saying
like
okay
yeah,
you
know
someone
from
alpha
has
spoken
to
this
to
this
person
and
it's
kind
of
confirmed
that
that
they
don't
sound
like
anyone
else.
A
Who's
in,
like
he's
kind
of
in
you,
know
the
air
at
the
sphere
right,
you
know
so
like
if
you
set
up
like
a
like
a
maybe
someone
as
an
example.
If
someone
said
it's
like,
I
said
something:
it's
someone
we
know
and
sets
up
a
delegate
account
and
then
governor
comes
to
their
list
and
says:
hey.
Do
you
mind
just
like
having
a
chat?
A
You
know,
then
we
don't
like
give
them
that
check,
mark
or
whatever
right
so
and
that
doesn't
kind
of
really
deal
with
the
issues
of
collusion.
You
brought
up
right
because
you
know:
there's
always
the
risk
if,
if
governor
is
kind
of
attacked
or
or
it's
malicious
or
whatever,
then
then
that
kind
of
goes
through,
but
it's
maybe
slightly
safer
because
it
kind
of
requires
to
kind
of
two
parties
to
kind
of
go
rogue
rather
than
just
one.
H
Yeah
that
sounds
like
a
good
place
to
start.
I
imagine
the
you
know
how
we
vet
delegates
or
whatever
will
just
evolve
over
time,
and
you
know
who
knows
what
we'll
end
up
with
sort
of
you
know.
Maybe
the
dow
ends
up.
You
know
having
way
better,
like
private
investigational
capabilities
than
like
a
government
someday.
You
know
who
knows.
H
B
L
So
it
would
become
a
full-time
job.
Doing
the
one
hand
have
a
vision
as
to
what
the
priorities
should
be,
that
the
the
dao
should
be
working
on
and
then,
on
the
other
hand,
figure
out
which
work
that
is
happening
where
the
capital
needs
to
be
allocated
which
work
needs
to
happen
in
the
dao
and
therefore
needs
to
be
funded
in
the
dao
to
to
make
those
objectives
happen,
and
then,
of
course,
also
check
up
on
the
quality
that
these.
L
The
the
main
teams
would
be
delivering
on
and
the
the
last
element
that
was
mixed
in
there
is
the
the
the
need
that
we
will
soon
see.
I
think
for
for
budget
limit,
because
right
now
it's
like
a
buffet
where
we're
adding
tasty
core
units
to
to
the
budget,
and
we
assume
that
we'll
never
see
the
limit
of
it.
But
sooner
or
later
we
will
get
there.
And
then
the
question
will
be
how
how
do
we
prioritize
between
different
core
unit
works
and
budgets?.
L
That
seems
like
a
good
model
that
that
improves
a
bit
on
the
on
the
core
unit
model
that
we
have
today
and
at
the
same
time,
is
very
compatible
with
it.
So
you
can
like
gradually
evolve
into
that
and
putting
the
allegation
in
place
is
actually
is
the
first
step
to
get
there
so
yeah.
I'm
definitely
excited
to
further
work
on
this
and
create
more
documentation
around
what
this
could
look
like
in
the
future.
L
But
for
now
it's
it's
pretty
high
level
and
long
term.
So
I
would,
I
wouldn't
want
to
influence
the
immediate
discussion
around
delegates
that
is
needed
right
now
too
much
with
with
a
long-term
vision,
but
we
definitely
yeah
we're
working
as
fast
as
possible
to
get
something
presentable.
That
would
include
a
bit
more
detail.
A
Future
yeah,
thank
you
for
that
yeah.
I
still
need
to
catch
that
recording
mr
lastly,
unfortunately,
oh
this
earlier
this
week,
yeah,
I
think
I
think
you're,
probably
right
right.
We're
gonna
have
to
move
to
some
something
like
that
in
the
future.
At
some
point,
just
due
to
scaling
issues
yeah,
I'm
hoping
it's
going
to
be
a
little
bit
more
more.