►
From YouTube: 2023-05-30 Rook Community Meeting
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
All
right
recording
has
started,
and
this
is
the
May
30th
2023
Rook
community
meeting
I
think
the
last
recording
has
still
not
been
uploaded.
So
I
will
follow
up
today
to
remember
to
upload
both
this
today's
recording
and
last
last
times
recording
as
well
sorry
for
that
delay.
So,
let's.
A
A
Been
a
combination
yeah
exactly
all
right,
so
we
don't
have
anything
planned
for
1.10,
so
I
will
go
ahead
and
Skip
on
past
that
and
let's
just
review
real
quick,
the
FI,
the
last
1.11
patch
release,
which
was
the
dot
six
release.
The
memory
reserves
me
correctly.
This
was
this
release
was
going
to
be
done
right
after
the
community
meeting
a
lot
from
two
weeks
ago.
So
I
don't
think
we
talked
about
this
in
the
last
community
meeting.
A
So
let's
just
look
at
the
any
highlights
here
that
you
want
to
call
out
real
quick
from
this
last
release.
B
Well,
let's
see
the
monitoring
it's
interesting
skipping
the
service
monitor
if
we're
not
enabled
that's
kind
of
and
there's
there's
been,
a
follow-up
fix
there
related
to
monitoring
which
yeah,
because
of
how
some
things
worked,
where
we
changed
the
version
supported
for
the
self-exporter
and
now
it's
effectively
disabled
for
the
latest
set
versions.
There's
well
I!
Guess
it's
not
related
to
this
release,
but
a
follow-up
release
for
the
next,
the
next
one
we
have.
B
We
have
an
issue
related
to
monitoring
and
exporter
that
we'll
get
to,
but
it
was
about
halfway
down
what
kind
of
caused
this
regression
is
the
default
active
manager
label,
if
only
one
manager
is
running
so
related
to
that
one,
and
also
the
exporter,
version
change
I'm,
trying
to
remember
now
from
the
long
weekend
what
it
was,
but
basically
it's
a
couple
issues:
we've
we've
had
some
fixes
for
and
we'll
get
into
the
release
today
or
tomorrow.
Anyway.
B
A
Ready
then
so
we
want
to
so
yeah.
That's
gonna
be
essentially
follow
the
same
pattern
right
of
a
the
next
release
about
seven
packs
release
for
1.11
being
done
not
too
long
after
today's
community
meeting
as
well.
Then
so,
let's
look
at
the
board
and
see
if
we
can
get
the
context
in
on
on
those
particular
fixes
that
we
expect.
B
Yeah,
so
the
Avon
has
a
PR
in
progress
which
I
need
to
put
it
to
the
blocking
release,
column
and
a
chance
to
look
at
the
board.
Yet
today,
but
yeah
we
need
his
fix
in
so
we
can
get
the
exporter
cleaned
up
if
which.
B
A
Moved
it
there,
it
is
all
right
there
we
go:
okay,
okay,
I
believe
you
I,
believe
you
all
right.
So
yeah
I
want
to
catch
us
up
on
on
on
this
PR
and
is
blocking
release.
So
we're
saying
that
that
until
this
is
reviewed
and
merged,
approved
et
cetera,
that
will
the
release
1.11.7
will
not
will
not
be
completed.
B
B
It's
just
still
in
progress,
oh
well,
and
this
is
basically
if
so
a
couple
releases
ago,
we
enabled
the
exporter
demon
for
the
latest
Steph
version
1726,
but
then
the
next
release
we
disabled
it
because
there
were
a
couple
issues
that
wasn't
quite
ready
and
then
now
the
exporter
stays
behind
and
we
need
to
clean
up
some
of
it,
for
the
set
version
doesn't
support
anyway.
So
it's
a
cleanup
ER
for
that
situation.
B
A
Yeah
got
it
okay,
so
then
we
expect
the
fear
feedback
here
to
be
incorporated
in
hopefully
sometime
today,
and
maybe
we
can
get
that
get
that
all
wrapped
up.
Is
there
a
desire
at
the
end,
to
also
get
this
PR
in
as
well?
It's
it's
currently
in
in
progress
for
review,
but
it's
not
blocking
the
release.
Everyone
will
consider
it
that
way.
Right.
B
Right
I
think
that's
my
list
of
to
do
to
review
today.
I,
don't
think
you'd
block
through
the
release,
but
certainly
if
it's
in
then
we'll
include
it
cool
sounds
good
to
me.
That's.
A
Been
in
review
for
over
a
month
now,
I
think
okay,
so
today
isn't
necessarily
going
to
be
the
magic
day
for
that
one.
Thank
you
got
it,
okay,
anything
else
that
you
want
to
call
out.
That's
already
done
that
will
be
included
in
the
release
as
well.
C
A
All
right
sounds
good
all
right,
so
that
is
so.
The
plan,
then,
is
to
hopefully
get
out
1.11.7.
Today
we
took
a
look
at
the
board
there.
Anybody
else
want
to
raise
anything
important
for
1.11
in
the
upcoming
patch
release
that
we
expect
to
go
out
today.
C
A
All
right
so
then
we
can
go
ahead
and
take
a
look
at
the
upcoming
1.12
board.
I
think
last
time
we
talked
about
reviews,
yeah
I
think
it
was
last
time
about
the
you
know
the
themes
for
1.12-
and
you
know
the
roadmap-
is
up
to
date,
Etc,
but
then
I
don't
know
if
we've
taken
the
follow-up
to
get
the
board
completely
Upstate
and
scrubbed.
A
So
this
is
the
1.12
board
here
and
the
release
is
about
I,
don't
know
sixish
weeks
away,
I
suppose
a
month
and
a
half
away
so
still
plenty
of
time
left
in
in
1.12
life
cycle.
Here
and
let's
see
what
is
important
on
this
one
I
know:
there's
been
some
progress
around
1507.
That
would
be
highly
impactful
for
a
long
long
time
of
investment
in
that
one.
Is
that
looking
like
it's
going
to,
you
know
be
included
in
in
1.12
time
frame
too
Travis?
A
B
I
expect
that
one
and
then
nice,
actually
amazing,
I,
don't
know
if
it'd
be
emerged
in
the
next
couple
weeks,
but
it's
large
more
of
a
design
exercise
than
an
implementation
exercise.
Just
because
trying
to
make
sure
we
could
cover
all
the
corner
cases,
but
at
the
end
of
the
day,
I
don't
think
there
will
be
a
lot
of
code.
So
it's
more
relying
on
what
happens
with
the
new
out
of
service,
node
out
of
service
label
that
kubernetes
has
or
the
taint
and
then
how
we
react
to
that.
B
We
need
to
you,
know:
fence
the
node,
which
I,
don't
believe
is
a
huge
change.
I
think
Siobhan's
getting
ready
to
open
up
here
soon,
though,
for
that
beyond
the
design
peer
yeah
other
than
that
the
Cozy
drivers
yeah
the
biggest
feature,
probably
that
we'd
really
love
to
see
and
jiffin
is
actively
working
on.
Getting
that
so
I'd
say:
that's
yeah,
there's
still
a
lot
to
do
with
the
Kazi
driver.
B
As
yes,
I,
don't
know
what
else
it
might
be
too,
but
those
are
the
main
ones.
On
my
mind,.
A
Got
it
and
how
about
Blaine
Alexander
others,
folks
for
things
that
are
important
for
you
in
in
1.12
that
are
still,
you
know,
either
on
the
board
already,
and
you
know
we're
going
to
get
an
investment
in
or
things
that
maybe
still
need
to
be
added
to
the
board
that
we
expect
or
hope
to
have
in
1.12..
C
Nothing
nothing
specifically
one
of
the
things
that
I'm
constantly
kind
of
like
looking
at
is.
C
I
think
for
the
4.12
like
or
412,
the
1.12
timeline,
I
think,
where
we're
still
just
gonna
be
seeing
some
CSI
improvements
for
NFL
rather
than
Rook,
specifically
but
they're
yeah.
Looking
looking
forward,
there
are
several
of
us
in
the
like
step
and
like
our
community
looking
into
like
high
availability,
like
actually
multiple
active
servers.
A
Well,
it's
definitely
a
good
theme
to
be
investing
in
so
I
know,
that'll
be
useful
for
sure
yep.
Let
us
know
if
you
have
any
like
any
tangible.
You
know
concrete,
like
findings
or
or
plans
to
address
certain.
Like
particular
issues
come
up.
We
can,
you
know,
make
sure
that
they're
added
onto
the
board
and
tracked
here
too.
A
Sounds
good
how
about
other
folks
for
1.12
anything
else
that
folks
want
to
bring
up
here
while
we're
still
kind
of
halfway
through
the
Milestone
or
so.
E
C
E
Yeah
I
just
need
to
get
care
Bros
rolled
out
first
and
then
I'll
be
willing
to
be
getting
picked.
Something
else
it's
in
here
in
the
card
is
the
ldap
integration,
so
I
haven't
done
anything
in
the
last
month.
After
two
meetings
ago,
I
said
I
was
going
to
look
at
it.
I
rebased
it
on
Master,
just
to
see
if
CI
would
pass
and
I
did
not
understand
what
had
bit
rotted
in
the
the
testing
and
so
I
need
to
figure
that
out
before
I
start
working
on
Blaine.
E
E
Yeah,
well,
if
you
can,
you
know
help
me
figure
out
why
the
CI
is
not
working.
That
would
be
great,
but
I
guess
I'm
hoping
to
get
to
it.
Yeah.
Let.
B
E
I
mean
there
are
certain:
there
are
certain
canaries
that
randomly
fail
with
no
reason
and
I
retrigued
it
two
or
three
times,
and
it
failed
in
the
same
way.
I
think
there
may
be
some
timeout
changes
like.
We
need
to
wait
longer
for
resources
to
settle
on
on
the
the
mini
Cube,
but
I.
Don't
know
why
it
used
to
work
reasonably
reliably
on
that
branch,
and
now
it's
just
it
took
fail.
B
C
This
might
be
a
good
good
place
to
pull
in
the
teammate.
C
I
I
don't
want
the
pr
number
is
it's
effectively.
Just
I
was
doing
some
work
related
to
stuff
related
to
multis
and
who
was
having
to
debug
their
tests
that
don't
have
any
log
collection
and
I
was
like.
Maybe
I
can
just
like
put
teammate
in
a
like
a
little
Fedora
container
and
just
like
run
it
and
then
have
teammate
run
into
kubernetes
cluster
with
like
permissions
to
like
edit
anything
it
wants
to,
and
it
worked.
C
It's
it's
really
helpful
and
nice
I
did
I
did
actually
I
I've
been
trying
to
set
up
a
an
integration
text
for
the
multi-salidation
tool
that
I
wrote
and
I
did
notice
there.
There
were
a
couple
like
minor
script
issues,
so
I
can
I
can
comment
Josh
on
what
what
we
can
add
to
that
PR
to
like
debug
during
execution,
yeah.
E
C
C
E
C
E
C
B
Yeah
after
we
get
a
little
more
familiarity
with
this
new
approach,
I
think
we'd
even
want
to
disable
the
tests
running
at
the
end
of
the
teammate
running
at
the
end,
because,
like
you
said
Josh,
it
does
slow
it
down.
You
don't
know
it,
you
don't
even
realize
it's
failing,
and
so
let's
just
fail
immediately
and
if
you
need
to
debug
it.
Let's
enable
this
and
just
rerun
the
tests.
Yeah.
C
Oh
so
I
mean
one
of
the
big
things
that
I
did
last
week.
C
Was
that
maltes,
the
multi
c2e
test
that
I
was
talking
about
so
in
the
files
there
there's
like
a
in
the
GitHub
action,
there's
a
start,
teammate
action
and
it
is
on
the
conditional
ifrunner.debug
which,
if
I
mean
most
people,
I,
guess
like
I
I,
don't
think
Josh
would
but
like
Travis
and
I
have
been
Alexander
and
a
few
others
of
us
have
abilities
to
like
restart
tests
with
like,
like
a
debug
checkbox,
and
that
would
automatically
start
that,
like
during
runtime
teammate
parallel
session,
and
then
I
also
made
it
except
a
like
debug
CI
label,
which
is
also
something
that,
like
we
could
add
on
to
PR,
to
get
that
extra
debug
info
I
think
it's
probably
good
that
only
I
mean,
like
only
you
know,
maintainers
and
reviewers
can
can
actually
allow
that
so
that
we
don't
have
people
abusing
email,
for
you
know
crypto,
mining
or
whatever,
but
that
has
that
is
something
I
experimented
with
a
little
bit,
and
is
that,
like
current,
you
know,
however,
like
eight
line
version
of
it's
working
well,.
A
D
A
A
Cool
all
right,
so
then
we
kind
of
popped
from
1.12
and
then
started
the
community
topics
session
there.
So
I
think
we
covered
everything
we
wanted
to
around
1.12
and
next
steps
for
that
and
then
we've
now
covered.
You
know
using
teammate
as
a
debugging
aid
for
for
CI
failures,
so
I
guess
they
can
move
on
down
then
to
the
next
agenda
item
as
well,
then
Travis.
B
I
guess
back
on
the
one
1.12
just
thinking
about
tentative
dates
for
the
release
around
July
18th
I
think
is
the
timeline
I'm
thinking
for
for
that
I
guess.
I
should
update
that
in
the
agenda,
but
possible
they
want
to
do
it
earlier.
Even
like
July
5th
depends.
I've
got
some
vacation
planned
in
July,
so
I
just
want
to
plan
around
that,
but
I'll
get
back
on.
We
could
say
July
18th
for
now,
potentially
yeah
adjust
that
I.
C
And
that
that's
after
the
Seth
brief
release
that.
B
Before
that's
true,
actually
that's
right,
so
the
but
yeah
earlier
July,
maybe
after
it
was
when
I
was
thinking
more
because
of
the
separate
Reef
release,
the
reef
is
being
released
in
June.
That's
all
the
ways
I've
heard
just
June,
and
so
maybe
we
should
Target
more
like
July
5th,
so
that
that's
closer
to
when
Reef
would
be
yeah
so
yeah.
Actually
Jerry.
Can
you
change
that
to
July
5th,
because
that
that'd
be
more
in
line
with
this
here
with
Reef
yeah
yeah?
B
C
I
think,
even
even
if,
if
release
even
if
REITs
release
slips
a
little
bit,
we
can
at
least
say
that
we
have
like.
C
We
should
support
released
versions
of
reef
when
they
are
released.
E
C
B
All
right,
yep,
let's
go
down
there
now,
so
the
we
have
had
a
little
bug
bashing
in
the
last
couple
of
weeks
on
the
crew
plug-in
the
building
from
implementation
instead
of
using
a
label
on
the
blocking
issues,
as
we
should
put
together
a
board
and
follow
the
same
pattern
we
do
for
for
the
work
project
but
yeah.
Basically,
we've
got
two.
These
two
tags
marked
with
blockers.
So
we
need
to
figure
out
why
the
OSD
Purge
is
hanging
and
I
think
has
a
PR
on
the
namespace
logs
or
no
blame.
B
A
B
Yeah,
the
goal
of
this
change
has
been
to
try
and
get
backward
compatibility.
You
know
same
functionality,
basically
for
the
chameleon
options,
I
think
I'd,
say
background
compatibility,
isn't
particularly
necessary
or
exactly
the
goal,
but
it's
still
expected,
though
that's
what
we
are
expecting.
Okay
and
if
there's
a
regression
or
maybe
not
exact
backward
compatibility,
I
I,
don't
think
it's
really
a
concern
honestly,
because
these
are
more
of
manual
commands,
not
something
people
are
probably
scripting.
A
A
Okay,
anything
else
on
that
Travis.
A
All
right,
Alex
update
us
with
the
newest
from
last
week
with
getting
the
orbit
Tool
up
and
running
yep.
D
Yeah
so
I
had
to
call
with
them
and
they're
going
to
get
back
to
me,
and
hopefully
a
few
days
now.
D
A
Yeah
and
then
Alex
I
saw
an
email
that
the
trial
is
about
to
expire.
Now.
Is
that
is
that
a
concern,
while
you
know
you're
kind
of
weight,
still
waiting
for
them
to
get
back
to
us
or
that
just
kind
of
drop
us
down
to
the
standard?
You
know
free,
because
I
have
like
a
starter
tier
that
you
know
you
don't
have
additional
functionality
in.
A
D
I
kind
of
went
like
I
kind
of
just
talked
about
like
where
sometimes
people
ask
questions
and
I
was
able
to
integrate
stack
Overflow,
for
example,
as
well
into
orbit
now.
So
we
basically
get
data
from
GitHub
slack
and
stack
Overflow
into
it
to
see.
What's
going
on
who's
using
it
and
yeah,
we
can't
plug
into
it
anymore,
but
well.
I'm
I,
don't
want
to
go
into
the
accessing
an
API
anymore
right,
yeah.
D
A
The
trial
I
feel
like
it
was
like
yesterday,
like
within
the
last
40,
at
least
within
the
last
48
hours.
I
thought.
A
Yeah
your
orbit
trial
ends
tomorrow,
and
that
was
today
at
3
P.M
European
time
so
like
6
a.m.
Pacific
time
this
morning.
A
D
A
Okay,
well,
that's
great
great
progress
with
that
Alex.
That
sounds
really
cool.
D
Jazz
at
one
point,
for
example,
right
now
that
it
would,
for
example,
show
us
what
I'm
seeing
right
now
here,
for
example,
is
at
Praveen
merged
the
first
PR
on
Rook,
for
example
like
that's.
What
orbit
to
some
degree
is
also
about
like
that.
We
gain
some
more
insights
there,
so
yeah
and
when
anyone
from
the
maintenance
who
isn't
added
to
it.
Yet,
please
reach
out
to
me
yeah.
D
A
And
then
Alex
new
architecture
diagram
is
in
progress
as
well.
It's
already
have
a
PR
open
for
it.
Yep.
D
Where
should
I
start
with
this?
This
is
the
typical
I
guess.
Example
of
yeah.
We
have
a,
we
have
a
diagram,
but
we
can't
edit
it,
but
now
we
can
edit
it
because
the
source
code
for
the
diagram
is
checked
into
the
repository.
You
might
not
necessarily
want
to
click
on
the
draw
IO
file.
It's
XML
magic.
Let's
just
keep
it
at
that.
Okay,.
D
A
Oh
here,
this
guy
I.
D
Is
kind
of
like
the
higher
level
architecture
that
I
would
currently
propose
here,
but
we
you
see
that
you
have
retry
it.
You
have
read
write
many
years,
ufs
free.
If
you
have
the
operator,
you
have
to
see
site
driven
provisioners
and
that
you
have,
in
this
case
free
nodes
as
an
example
with
some
components.
D
It's
a
bit
detailed
on
that
side,
but
I
think
it's
you
know
like
it
gives
people
deep
enough,
Insight
there's
at
least
one
pod
running
in
the
end
for
you
to
be
able
to
consume
your
storage,
I
guess
if
that
makes
any
sense
and
of
course,
the
other
diagrams
that
I'm
currently
working
on
there,
it's
going
to
be
a
more
detailed
kind
of
like
further
storage.
What
components
are
in
play
to
for
you
to
create
a
PVC
and
get
your
storage
mounted
to
your
pod,
but
this
is
currently.
A
And
I
I
assume
that
you
know
the
one
that
gets
committed
to
the
you
know
the
Rook
docs
and
code
base.
Whatever
would
it
be
called
Rook
architecture.
A
Yep
sounds
good.
This
is
yeah
this.
This
looks:
yeah,
okay,
I,
maybe
talk
more
about
it.
On
the
on
the
pra,
it
looks
like
we're
trying
to
balance
like
a
you
know,
component
interaction,
you
know
boundary
type
of
thing
and
then
also
like
a
more
simplified
architecture
and
stack
as
well.
So
maybe
this
this
finds
a
good
balance
between
that,
but
I
wouldn't
think
about
that.
More.
D
That's
that's
currently
the
idea
job,
basically
to
kind
of,
like
the
reason,
also
why
it's
so
layout
is
that
it's
I
guess
giving
the
like
I'm
still
fighting
with
myself
a
bit
there
or
I
think
even
a
little
bit
with
Travis
and.
A
D
There,
if
it
might
make
sense,
for
example,
to
even
at
like
a
small
kubernetes
or
like
a
box
around
everything,
I
guess
you
know
for
like
you
need
kubernetes
for
this.
So
to
say,
like
everything
is
in
puberty,
is
like
it's
well,
so
I
think,
maybe
just
bringing
that
up
as
an
idea
should
we
add,
like
a
frame
around
at
least
the
lower
part
of
the
high
level
chart
of
like
this
is
what
will
be
incubernetes,
or
is
that
too
far.
B
D
Up
on
the
pier
yeah
sure
it
has
a
broader
main
point
is
really
just
that
we
have
the
source
code
of
the
diagrams
and
can
then
easily
upgrade
and
update
them,
as,
for
example,
in
one
of
the
earlier
ones
already
as
well.
I
think
it's
already
started
mentioning
cozy,
which
isn't
quite
there
yet.
So
that's
why
it's
currently
just
a
free
API,
but
when
cozy
is
there
that
we
will
just
someone
needs
to
plop
it
into
dry,
oh
and
edit.
It
comment
it
and
we're
done
again
so
yeah.
B
D
The
robot
CSI
is
also
quite
low
level,
but
it's
like
it.
The
reason
why
I
would
like
to
include
maybe
even
cozy
there,
at
least
as
like
a
mansion
in
at
least
one
box
is
that
did
you
saw
all
things
that,
like
you
know
like?
If
we
stack
everything
on
top
we're
basically
having
kubernetes,
we
have
the
Rook
operator,
then
we
have
technically
Seth
that
is
running.
Then
we
have
CSI
technically
cozy
in
the
operator
or
somewhere
else
and
like
it
kind
of
all,
you
know
Stacks
together,
so
I.
C
D
C
C
D
D
I
feel
that
we
should
mention
it
at
least
a
10
as
one
part,
because
in
the
end
it's
one
primary
component,
you're,
probably
going
to
use
if
you
use
Rook
with
well
at
least
one
of
the
storages.
Well,
if
you
don't
use
rgw
okay,
we
can
argue
about
it,
but
like
it's.
C
B
C
B
C
I
yeah
and
I
think
calling
out
CSI
and,
like
especially
I,
think
also
cozy
will
probably
drive
like
people
to
check
it
out.
If
we
have
it
in
that
diagram,
and
that
is
ultimately
like
what
we
want
and
what
we
need
and
I
I
was
finally
able
to
to
sync
up
with
the
Cozy
folks
on
the
stand-ups
that
are
now
bi-weekly
and
yeah.
C
They're
they're
kind
of
just
been
an
information
gathering
phase
that
now
they're
one
team
users,
they're
wanting
user
feedback,
yeah
I
think
we're
in
a
good
position
to
like
be
able
to
help
provide
a
lot
of
that
and
I
mean
we.
We
already
have
have
been
by
nature
of
having
experience
with
the
Proto
cozy
live
bucket,
provisioner
stuff.
A
Yep
that
makes
sense
to
me
as
well
all
right.
So
then
it
sounds
like
maybe
a
couple
more
comments
to
dropped
in
on
that
PR.
Now
that
Alex
just
yesterday,
provided
that
updated
diagram
they're
the
most
recent
version
we're
just
looking
at
so
thanks
for
that
effort.
Alex
appreciate
that
that
was
everything
that
was
in
the
agenda
document
for
today.
Anything
else
that
folks
wanted
to
bring
up
before
before
we
adjourn
for
the
week.
D
There's
an
action
items
bill
I,
guess
setups
like
an
extended
any
right
under
Jared.
No.
A
No,
it
was
not
something
that
I
prioritized
in
the
last
two
weeks,
while
I
was
catching
up
from
my
from
my
multi-week
off
time.
Since
you
know,
we
I
think
there's
something
that
should
be
done
before
the
next
community
meeting,
but
was
not
done
since
last
time.
Since
we,
you
know
kind
of
understood
that
it's
not
super
super
pressing,
because
we
have
some
breathing
room
with
the
current
invite
system.
D
D
B
D
Yes,
I
opened
the
document
and
I
think
I
even
clicked
on
at
the
right
point.
But
I
didn't
write
anything
sorry,
good
point:
it
was
about
the
adoptos
MD
file,
I
guess.
D
Just
voice
it
as
like
a
concern,
or
just
as
like
a
what
can
we
do
about
it
is
so
the
adoptos
and
default
well
contains
company
names
and
also
links.
D
To
company
websites,
the
problem
is:
there's
I,
guess
for
like
that,
there's
quite
some
more
and
more,
which
are
not
working,
at
least
if
you
are
not
working
last
time,
I
went
through
quite
like
a
we
clicked
through
them.
D
So
I
guess
the
question
about
us
like
how
should
we
handle
this?
Should
we
reach
out
or
try
to
reach
out
to
them,
should
be
just
remove
them.
B
B
Yeah
exactly
I
think
if
links
are
broken,
we
could
just
remove
the
link
and
if,
if
people
want
to
add
themselves
like,
we
had
a
PR
just
opened
a
few
days
ago,
where
somebody
wanted
to
add
themselves
to
list
great.
That's
fine
it'd
be
nice
to
do
a
survey
again,
sometime
soon
to
say:
hey,
let's
update
this,
who
wants
to
be
on
it
and
just
see
what
response
we
get?
I
think
that
would
be
nice
too,
but
to
fix
links
in
the
short
term.
A
I
that
that
sounds
fine
with
me,
Travis,
similar
sentiments
here,
is
that
you
know
broken
links.
We
can
yeah
just
turn
them
not
into
links.
C
A
You
know
if
they're
like
do
a
little
research
to
see
how
okay
did
they
change
websites
or
domains
or
something
like
that
and
try
to
update
it
appropriately.
That's
that's
totally
reasonable.
I
think
a
model
that
I've
seen
work
pretty
well
for
adopters
files.
You
know
which
we
didn't
really
need
to
do,
because
we
did
this
for
graduation
and
essentially
never
touched
it
again,
but
a
model
I've
seen
work
really
well
and
I
think
the
first
place
I
saw
was
in
the
backstage
project.
A
But
basically
you
turn
this
into
a
self-service
effort.
Where
you
know
you,
you
have
explicit
language
and
wording
around
encouraging
people
to
open,
PRS
and
add
themselves,
and
you
know
talk
about
the
benefits
to
the
project
that
that
provides
Etc.
So
that's
probably
a
good
idea
as
well
too,
to
go
ahead
and
turn
this
into
a
self-service
type
of
experience.
A
There's
a
we
did.
We
I
talk
to
the
backstage
folks,
and
then
you
know
took
that
idea
from
them
when
they
told
me
about
it,
so
that
that's
where
the
credit
goes
to
is
at
least
as
far
as
I
know
is
the
backstage
folks,
and
then
we
did
that
in
the
crosswind
project
as
well,
where
we
give
three
options
where
you
know
you
could
say
Hey,
you
can
open
up
a
PR
here.
A
Then
we
also
like
created
a
you
know:
Google
form
survey,
sort
of
thing
like
if
you're
not
comfortable,
using
gits
or
opening
a
PR
or
whatever
you
can
just
you
know,
go
to
the
form
and
fill
it
in
and
then
you
know
we
get
that
data
on
the
back
end
from
the
Google
forms
result
and
then
we'll
just
do
a
PR
ourselves.
Maintainers
can
do
that
and
then
also
you
can
like
reach
out
to
the
steering
committee.
A
A
Let
me
just
drop
a
link
to
it,
because
I
took
what
backstage
did
and
then
kind
of
took
a
little
bit
further.
So
you
see
here
like
there's
three
different
options
there.
So
I'll
just
drop
this
into
the
agenda.
Doc
too,.