►
From YouTube: OKD Working Group Meeting 07-18-2023
Description
The OKD Working Group's purpose is to discuss, give guidance to, and enable collaboration on current development efforts for OKD, Kubernetes, and related CNCF projects. The OKD Working Group includes the discussion of shared community goals for OKD 4 and beyond. Additionally, the Working Group produces supporting materials and best practices for end-users and provides guidance and coordination for CNCF projects working within the SIG's scope.
https://okd.io
A
B
Folks,
welcome
to
the
okd
working
group
meeting
for
July
18th
of
the
Year
2023
and
let's
take
a
few
seconds
to
look
over
the
agenda.
Let
us
know
if
there's
anything
that
we've
missed
anything
that
needs
to
be
added
or
moved
around,
so
we'll
take
about
30
seconds
to
to
look
it
over.
B
All
right:
well,
then,
let's
jump
right
into
it.
Let's
start
with
the
okay
to
release
in
cicd
updates
with
looks
like
we've
got
Luigi
and
yeah
Luigi
go
ahead,
yeah.
C
Cool,
so
really
not
much
change.
C
From
last
week,
I'm
working
with
Christian
to
reorganize
the
okd
project
core
is
pipeline
so
that
we
can
separate
the
tasks
and
pipelines
into
its
sort
of
basically
own
repo,
so
that
we
can
submit
the
actual
costs
and
pipelines
directly
to
to
the
techton
hub,
and
then
they
can
make
use
of
like,
for
example,
The
Matrix,
The
Matrix
task
that
can
use
as
as
a
separate,
almost
task
and
plugin,
and
then
yeah
we'll
be
in
a
position
to
do
a
release
at
the
end
of
the
Sprint.
C
So
it
should
be
this
week
sometime
for
4,
13,
stable
and
4
14.
Okay,
the
S
costs
we
call
internally
is
up
next,
so
yeah.
C
D
B
Questions
or
feedback
on
what
Luigi
was
thrown
down
there.
D
Hey
folks,
okay,
yep,
okay,
great
so
well,
I
don't
have
much
update.
It's
a
slow
moving
right
now
we're
starting
to
get
into
the
summer,
so
there's
not
much
happening
besides
the
items
from
the
last
time,
which
was
200,
I
think
so.
Hypervis
support
coming
real
soon
we're
pushing
the
change
in
federal
arm
right
now,
so
this
should
land
in
February,
39,
etc,
etc.
D
E
B
B
The
only
thing
I
would
say:
Timothy
is
you'll
notice.
Later
in
the
agenda.
We
have
a
link
to
merch
request
of
a
blog
post
for
okd
building
like
building
okd,
basically
on
fcos
and
S
costs,
and
so,
if
you
could
review
that
for
accuracy
for
anything
related
to
F
costs
that
you
think
you
know
needs
to
be
fleshed
out,
or
you
know
that
you
think
would
be
beneficial
for
this
blog
post.
Please
feel
free
to
comment.
We
want
any
input
and
all
input
all
right.
B
Let's
see
Moving
On
Now
to
S
cost
stuff
and
I.
Don't
know
that
we
have
anyone
here
to
speak
to
that
unless
Timothy,
unless
you
have
anything
to
provide
in
terms
of
s-cost
stuff.
B
So
then
we
will
jump
forward
by
the
way.
I
should
mention
that
it
was
Christian
when
I
approached
Christian
about
doing
a
video
segment,
and
he
mentioned
that,
in
fact,
he
was
going
to
be
doing
some
internal
demonstration
and
that
he
would
record
a
video.
He
is
the
person
who
hooked
us
up
with
the
the
person
who
wrote
the
blog
post
for
us,
and
so
that's
very
much
appreciated.
We
appreciate
Christian
setting
that
up
all
right,
Community
Development
updates
with
Brian.
F
Okay,
so
so
last
week
we
had
a
review
of
the
the
planning
board,
not
much
of
change,
so
we
had
quite
a
few
interesting
discussions.
F
F
I
was
trying
to
think
do
we
need
to
get
with
Timothy
on
any
of
that
discussion.
I
think
there
was
a
discussion.
Do
we
should
we
split
out
things
like
the
ignition
file
and
generic
pieces
of
core
Os
from
say
the
the
Fedora
or
the
Centos
streams
core
OS?
F
Would
it
be
worth
bringing
out
some
of
this
sort
of
core
OS
features
like
the
ignition
file
to
try
and
reduce
them
to
like
15
minutes
so
15
minutes
on
the
specifics
of
certain
areas,
we're
capturing
that
in
the
in
the
ticket
item?
F
So
we
just
need
to
sort
of
come
to
an
agreement
on
that
and
then
start
getting
those
videos
sort
of
and
created.
So
that
was
one
of
the
discussions
that
we
had
around
that
video.
We
also
had
sort
of
a
general
discussion
about
getting
started
and
we
do
need
to
do
some
work
on
the
website
to
actually
create
that
getting
started
section
and
the
front
page.
Get
that
finished.
That's
down
to
me.
F
I
just
need
to
find
some
time
to
get
to
get
that
going
to
provide
people
with
a
faster
or
a
simpler
way
to
actually
understand
what
is
okb
and
how
they
can
get
going
with
it.
F
I
don't
realize
we
need
to
take
better
notes
in
the
hack
MD,
rather
than
just
going
through
the
board,
where
we
want
to
do
the
board
Lookout,
because
we
didn't
actually
catch
or
anything
in
the
meeting
minutes.
Last
week
we
also
had
a
discussion
around
how
we're
going
to
sync
up
the
operators,
and
one
of
the
challenges
is
that,
with
ocp
all
of
the
release
information
is
sort
of
behind
the
red
hat
firewall
so
working
out
when
they
actually
do
a
a
a
release.
F
None
of
that's
written
to
the
GitHub
repo,
it's
all
within
the
build
system
within
red
hat,
so
there
is
going
to
be
some
challenges
to
try
and
keep
keep
okd
operators
in
line
with
The
ocp
Operators.
Unless
we
get
access
to
some
of
that
internal
metadata,
so
I
think
what
would
you
have
a
conversation,
Luigi
myself
and
whoever
write
is
needed
actually
see
if
there's
a
way
around
that
yeah.
D
F
B
So
we
did
also
have
volunteer
for
producing
videos
and
Vlogs
that
are
user
oriented,
customer
oriented,
and
so
that's
very
helpful
as
well.
I
think
is
to
round
that
stuff
up,
because
a
lot
of
our
material
in
that
area
is
old.
I
think
you
know
a
couple
of
years
old.
E
F
Other
than
that
I
think
that
really
covered
it
up
for
last
week.
It's
discussions.
B
B
I
think
that
for
the
for
the
video
series,
that's
going
to
be
our
Focus,
maybe
next
week
is
starting
to
script
out
the
video
series
and
then
maybe,
if
we
could
get
Christian
I
know
you
you
jumped
on
since
we
started
the
meeting.
Are
you
able
to
were
you
able
to
record
any
video
yet
of
the
presentation
or
did
that
get
distilled
into
the
blog
post?
Or
do
you
can
you?
Can
you
still
commit
to
doing
some
video
or
where
do
you
stand
on
that.
D
Yeah,
why
well
Christian
fine
is
me,
is
Mike
there's
how
much
time
should
each
section
be?
Does
the
title,
but
no
matter
how
much
time
each
part
of
the
video
should
be
so.
B
I
I
think
what
we
settled
on
is
like
so
ignition
for
15,
then
F
cost
specifically
for
15.
Does
that
sound
about
right
for
videos.
B
D
I
have
15
minutes,
maybe
it's
a
little
bit
long,
but
I
don't
know
okay,
we
can.
We
can
start
with
that.
Well,.
D
B
D
Okay
so
so.
B
I
would
say
like
15
on
ignition
itself
and
then
15
on,
like
the
concepts
of
core
OS
and
whatnot,
and
then
that
will
then
allow
Christian
or
whomever
to
do
the
video
about
sort
of
the
build
part
and
on
top
of
core
OS,
and
that
gives
us
sort
of
the
layers
right
there.
You
know
of
everything,
and
then
we
did
have
someone
volunteer,
actually
our
our
sort
of
long
time,
Community
member
Brendan
Shepard.
B
It
actually
is
working
on
a
video
for
MCO
in
anyway
and
volunteer
to
share
that
with
us,
and
so
there
was
an
exchange
between
Brendan
and
I.
Earlier
this
morning,
where
he's
like
I'm
a
little
behind
on
my
opinion,
communities,
but
I
watched
the
latest
meeting
recording
on
YouTube
and
saw
you're
looking
for
a
video
on
MCO
I
actually
have
recorded
one
already
and
need
to
edit
it
were
there
any
NCO
specific
tops
topics
you
wanted
covered,
though,
and
I
just
said.
B
In
short,
we're
looking
for
a
video
explaining
how
MCO
factors
into
the
overall
installation
process
and
configuration
of
an
okd
cluster.
The
goal
of
the
video
series
is
to
provide
some
foundational
material
for
people
who
wish
to
contribute
to
okd
and
or
help
provide
community
support,
and
he
acknowledged
that.
So
that's
great,
so
we've
got
MCO
covered
now
and
Brendan
I,
don't
know
if
anyone's
ever
seen,
Brendan
Shepard's
videos
but
they're
really
good.
B
B
B
F
I
I
think
they
do
want
to
be
sort
of
Fairly
Standalone,
but
what
we're
trying
to
achieve
is
the
collection
of
videos
gives
you
the
understanding
of
how
that
that
bootstrap
works
and
how
an
okd
cluster
goes
from
nothing
to
a
running
cluster.
So
this
should
help.
You
understand
the
process,
because
then
the
plan
is
that
our
Nico's
doing
the
troubleshooting
Essentials.
So
this
is
trying
to
give
you
the
grounding.
F
So
when
you
get
into
troubleshooting,
you
understand
what
the
main
pieces
are,
what
part
they
play
in
the
process
and
then
you
just
start
looking
at
the
debugging
information,
you
get
an
understanding
of
what
each
piece
is
doing
in
the
process.
So,
yes,
they
do,
they
can
be
Standalone,
but
together
they
should
provide
a
sort
of
coherent
body
of
knowledge.
B
I
would
say
you
could
do
butane
separate
because
to
me,
that's
something
that
that's
a
topic
on
its
own,
that
you
know
we
can
point
people
to
when
there's
specific
issues
with
butane,
stuff
or
or
a
need
for
understanding,
then
we
can
just
point
them
specifically
at
that
video
and
they
don't
have
to
like
navigate
through
other
stuff.
Does
that
feel
about
right
Timothy?
Does
that
feel
right
to
you.
B
And
thank
you
for
doing
that.
It's
much
appreciated.
I
think
this
is
going
to
be
really
helpful.
All
right
anything
else
on
the
video
series
stuff,
yeah.
A
Go
ahead,
Jamie,
yeah,
I
have
a
question:
I
was
hoping
to
see
Christian
here
for
making
a
video
on
the
agent
based
installation.
A
Currently,
we
need
to
land
a
little
bit
of
work
in
the
installer
right
now.
I've
made
the
pr
that
was
a
Poca
I,
put
a
shirt.
The
link
and
I
added
shuttle
with
Christian
I
understood.
He
was
about
to
open
a
new
PR
based
on
on
this
work,
essentially
to
embed
the
ask
custom
metadata
into
the
installer
and
to
activate
the
escos
and
bad
being
into
the
machine.
Or's
images
with
that
stuff.
A
We
could
have
a
working
thing
and
doing
the
the
video
I
think
it
could
be
a
prerequisite
for
doing
any
I
could
do
the
video
even
right
now
using
a
workaround,
because
we
currently
landed
the
fix,
we're
using
an
environmental
variable
that
you
can
set
a
base
image
of
a
ride
and
make
it
work.
But
it's
not
the
idiot
thing
or
the.
A
How
can
I
say
that
they
recommended
approach?
Maybe
for
making
the
video
could
be
fine,
but
if
we
could,
the
land
is
work
before
I
think
it
would
be
better.
So
I
didn't
see
anything
new,
so
let
me
know
if
we
can
help
somehow
or.
B
I
I'd
be
inclined
to
wait,
I
think
because
I
mean
we've
got
time
on
that
particular
topic,
and
it
seems
to
me
that
we
probably
don't
want
to
educate
people
on
a
workaround
and
have
that
floating
around
so
I,
don't
know
Brian.
What
do
you
think.
F
D
F
Yeah
I
I
think
we
do
want
to
try
and
get
this
prioritized,
because
that
that
then
solves
a
number
of
issues
around
okb.
So
and
then
we
can
make
the
content
around
that,
because
I'm
waiting
on
this
for
the
SNL
documentation
as
well.
B
So
I
think
that's
the
answer,
yeah
and
thank
you
for
asking
and
thank
you
for
doing
that.
Work.
That's
awesome!
Thank
you
all
right.
Anything
else
for
oh
we've
got
Christian
Christian
is
here
hey.
G
I
can
hear
you
now
awesome
hello.
Everybody
I
had
to
reboot
my
my
laptop
and
sorry
for
being
late
meeting
before
ran
over
a
bit.
It's.
G
So
we
have
set
up
our
infrastructure
on
the
Massachusetts
open
Cloud.
We
are
running
the
builds
there
successfully.
It's
it's
looking
really
good.
We
are
going
to
cut
another
release
this
week
and
for
both
the
next
and
the
stable
stream.
G
So
it's
it's
all
looking
good
at
the
moment
in
terms
of
our
pipeline.
There's
some
changes
coming
so
right.
Now
we
have
everything
in
the
okd
Colorado
West,
Pipeline
repository
right
and
we'll
I've
created
a
new
repo,
already
the
techtone
catalog
repo
in
the
OPD
project
and
we're
going
to
move
all
the
tasks
and
pipeline
definitions
for
both
the
the
core
OS
pipeline,
as
well
as
the
release
pipeline
over
there
and
then
we'll
just
use
the
OK
decoros
pipeline
Repository
as
our
git
Ops
repository
essentially
referencing.
The
tasks
and
pipeline
definitions.
E
G
The
other
repo,
and
that
way
we
can
also
make
the
the
tecton
catalog
repo
accessible
to
all
other
repositories,
making
it
possible
to
reference
the
tasks
and
pipeline
definitions
in
there
and
to
do
things
like
the
pipelines
as
code,
which
is
another
thing
we
will
add
pipelines
as
code,
is
something
that
the
pipelines
operator
provides,
and
essentially
it's
it's
a
it's
having
the
definitions
for
your
CI
CD
pipelines
in
the
repository
that
you're
actually
testing.
G
So
you
have
a
DOT
tacton
directory
in
the
root
of
your
repo
and
in
there
you
you're,
going
to
be
able
to
put
custom
pipelines
referencing
the
tasks
from
the
other
repo
that
we
will
maintain
inversion
and
also
publish
to
techton
HUB
to
make
it
yeah
to
distribute
it.
So
really
that,
in
terms
of
changes,
we'll
have
this
new
techton
catalog,
where
all
the
tasks
and
pipelines
will
live
and
then
we'll
have
one
GitHub
throughpo,
which
is
the
OPD
coreos
pipeline.
For
now
for
managing
our
build
Farm
through
githubs
and.
G
Set
up
githubs
on
on
the
build
Farm
we
have
now
it's
not
running
in
enforcing
mode
yet,
so
it's
not
actually
applying
the
changes,
but
we
are
going
to
do
that,
probably
at
the
beginning
of
August,
when
our
SRE
guy
is
back
from
PTO.
In
the
meantime,
we
don't
want
to
mess
with
the
setup
too
much,
but
we've
set
it
all
up.
G
It's
running
Argo
in
the
background,
actually
on
yet
another
cluster,
so
we
don't
run
Argo
on
the
cluster
that
we
on
the
build
Farm
cluster
later
we
might
change
that
up
a
little
bit,
but
for
now
we
will
yeah
we'll
we'll
use
the
the
formal
operate
first
folks
to
help
us
with
the
setup
and
and
they
have
infrastructure
still
in
place
to
host
Argo,
so
we're
using
that
instance
at
the
moment,
so
I'm
very
excited.
G
There
is
also
another
PR
I'm
just
going
to
mention
on
the
okd
chorus
pipeline
repository
that
introduces
a
split
between
the
minimal
s-course
base.
That
is
not
that
does
not
include
any
openshift
and
okd
dependencies,
and
then
we
will
use
that
as
a
common
base
for
both
the
next
and
stable
streams.
We
will
just
layer,
we
found
out.
E
G
On
top
for
stable
and
defaulted
14
ones,
on
top
for
the
next
Stream
So
on
a
shared
base,
since
we
currently
do
two
composers
a
day
and
99
of
the
packages
are
the
same,
so
the
the
idea
here
is
to
to
use
a
common
base,
and
actually
that
is
something
the
the
core
OST
is
also
interested
in.
So
we
are
kind
of
pioneering
that
idea
and
then
we'll
we'll
feed
it
back
to
the
car
OS
team
and
see
what
how
how
they
can
use
it.
G
B
B
B
Is
this
in
lieu
of
a
video
or
will
you
still
be
able
to
do
a
video
or
what?
What
are
your
thoughts.
G
So
there's
going
to
be
a
presentation
I'm
going
to
hold
tomorrow
internally,
which
I
will
do
again
and
record
I,
guess
or
I'll
at
least
have
slides
that
will
it'll
it'll
be
similar
to
the
blog
post,
it'll
it'll
kind
of
be
a.
This
is
what
we
have
in
terms
of
our
tecton
pipelines.
This
is
how
we
do
it
and
the
Eternal
discussion
this
is.
G
This
will
be
just
to
Showcase
what
we
do
and
okay
there's
other
teams
that
that
do
things
differently
and
we're
really
bringing
together
a
lot
of
different
viewpoints
tomorrow.
So
that
is
going
to
be
the
okd
representation.
So
I'll
still
have
that
yeah
bye
by
next
week.
B
Okay,
cool.
Thank
you
so
much
any
other
questions
or
comments
on
that
area.
F
F
Christine
I
was
going
to
say:
did
you
want
to
comment
on
the
pull
request
around
the
agent
installer
for
to
add
SEO
I?
Think
there's
a
question
for
you
when
you're
having
difficulty
joining.
F
A
Yeah
so
Christian
I
was
asking
since
I
volunteered
for
making
a
video
about
agent
based
installation
on
okd.
I
was
mentioning
that
we
need
the
first
to
land
the
missing
work
for
embedding
the
metadata
for
s-cos
into
the
installer
and
then
generating
also
the
machinos
image
so
wandering.
A
G
Yes,
I'm,
so
we
we
still
don't
so
to
give
some
context
the
metadata
required
here.
We
only
recently
started
generating
and
we
haven't
had
a
release
with
that
metadata
created.
So
we
are
now
generating
that
metadata
on
our
daily
builds.
With
the
next
release,
we'll
we'll
cut
this
week,
we
will
have
a
set
of
metadata
that
is
canonical
that
that
is
part
of
a
release
and
that
we
can
then
use
in
the
install
because
it's
not
going
to
be
coverage
collected
and
all
so
that
is
coming
soon.
G
Close
your
PR
and
reference
that
as
a
follow-up,
yeah
I
have
not
forgotten
about
it.
Yes,
that
is
that's
the
thing
next.
B
Excellent
and
if
folks
could
put
in
the
comments
of
the
video
ticket
any
updates
that
they
have
over
the
coming
weeks,
that
would
be
good,
so
just
go
ahead
and
comment
on
the
Lincoln.
That's
in
the
meeting
notes
the
the
ticket
for
the
foundations,
video
series.
B
B
All
right,
it
looks
like
we're
good
thanks.
So
much
for
your
time,
folks
and
Community
Development
meeting
next
week,
we'll
focus
on
sort
of
fleshing
out
the
last
few
details
of
the
video
series
and
how
to
present
that
I
think
to
the
public,
and
you
know,
if
there's
anything
you
need
reach
out
to
myself
or
Brian
and
we'll
make
sure
that
we
yeah
go
ahead.
G
Yeah
I
probably
missed
it.
I
just
wanted
an
update
on
the
operator
building
effort.
Since
we
now
have
a
built
Farm,
you
could
deploy
any
any
kind
of
pipeline,
tecton
wise
on
there
to
build
the
operators
and
just
to
get
an
invitation.
If
you
have
something
that's
ready
to
be
deployed
and
auto
triggered,
let's
say
a
weekly
build
or
a
daily
build
of
an
operator
of
your
choice.
We
could
deploy
it
on
our
build
farm
and
have
it
you
know,
build
build
automatically
managed
through
githubs
on
our
build
Farm,
it's
just
it.
F
Yeah
I
mean
just
just
when
I
am
we
haven't
got
that
far
yet
and
we
do
need
to
sort
of
have
a
sit
down
conversation.
Probably
you
myself,
Luigi
and
any
other
people
from
red
hat
and
the
issue
is
that:
how
do
we
trigger
operator
bills?
How
do
we
or
do
we
need
to
stay
in
line
with
ocp,
so
version
X
of
an
operator
on
ocp
matches
version
X,
because
none
of
that
is
tagged
in
git
repos.
F
It's
all
done
behind
the
red
hat
firewall,
so
we're
not
even
sure
if
it's
going
to
be
possible
for
them
to
stay
in
line,
in
which
case
how
do
we
know
what
good
commits
are
that
we
should
be
building
operators
on.
So
there's
a
that
front
piece
that
until
we
actually
know
what
we
need
to
build,
we
can't
really
don't
actually
do
any
building,
so
that
bit
is
yet
to
be
resolved.
G
Okay,
so
my
recommendation
would
be
not
to
try
to
try
that
to
get
the
same
comment,
because
it's
going
to
be
hard
to
find
and
really
that
there
is
different
branches
for
for
those
operators
on
the
repo,
and
one
of
them
is
the
development
branch.
And
then
you
have
release
branches
and
everything
that
goes
into
a
release.
Branch
has
gone
through
and
also
the
death
Branch,
even
though
there
might
be
other
things
at
play
in
on
the
release
branches.
G
The
confidence
we
can
have
that
things
will
work
is
pretty
high,
so
my
recommendation
would
be
to
just
say:
look
we're
going
to
do
a
weekly
or
a
daily,
build.
Let's
say,
weekly
weekly
build
of
the
operator
from
the
latest
comment
on
that
release,
branch
and
that'll.
That's
going
to
be
the
one
for
stable
for
the
stable
stream
on
s-cos
and
then
MDF
cross,
okd
stream
as
well,
and
then
the
other
build.
We
can
do
another
build
from
from
the
master
Branch
or
the
development
branch
and
that'll,
essentially
yeah.
G
F
G
So
I
again,
my
recommendation
would
be
to
have
two
streams
here.
We
we
essentially
have
two
stable
streams
right.
We
have
the
stable
stream
on
top
of
f
cross
and
the
one
on
top
of
s-cross.
They
are
really
the
same
cluster
code,
the
only
difference
there
is
the
operating
system.
We
don't
usually
have
any
operating
operating
system,
specific
things
in
the
operator,
so
it
really
the
operator
doesn't
know
it
doesn't
care
most
of
the
times
that
there
may
be
some
special
operators,
like
the
CMV.
E
D
G
Long
term
is
to
also
introduce
a
second
f-course
stream
to
have
the
same
thing,
one
next
and
one
stable
stream
for
both
operator
for
both
operating
systems.
Right
now,
the
F
cross,
one
we
have
is
stable
right.
It's
not
it's,
not
the
next
string,
long.
G
Like
to
to
test
everything
and
expand
that
Matrix
a
little
bit
right
now,
we're
really
only
focusing
on
s-course,
but
once
we
have
that
in
a
very
good
place,
we'll
try
to
bring
that
same,
builds
a
built
process
to
the
f-cos
build
as
well.
F
cost
is
a
bit
special
because
F
cross
currently
already
is
a
derivation
built
from
from
fcos
all
right.
The
okd
on
F
cross
just
has
a
Docker
file
for
building
the
the
operating
system
and
with
the
split
in
in
the
image
in
escos
that
I
mentioned
earlier.
G
We'll
have
something
similar
with
s-cos,
so
we'll
again
work
it
out
in
s-cross
and
then
we'll
try
to
bring
that
over
to
f-cross
too,
but
yeah.
My
recommendation
would
be
to
just
have
two
streams:
one
one
built
from
the
stable
Branch
one
built
from
the
development
branch
and
everything
there
should
already
have
a
good
amount
of
testing,
and
then
obviously
we
can
see-
and
that's
another
thing
we
have
on
the
backlog.
There's
the
OC
I,
always.
D
G
Up
the
there's,
a
tool
for
running
self-certification
for
openshift
clusters,
which
is
essentially
a
tool
that
allows
you
to
run
an
end-to-end
test
and
a
yeah
all
the
tests
yourself
on
any
cluster,
and
we
can
use
that
in
techton
and
Luigi
has
already
some
done.
Some
work
to
to
get
that
working
with
tech.
Talk
we'll.
E
G
Able
to
use
that
to
run
and
try
and
tests
ourselves
eventually
ourselves
eventually
for
anything
we
want,
and
obviously,
if
there's
an
operator,
we
think
this
needs
this
additional
testing.
We
want
to
do
an
end-to-end
test
ourselves,
we'll
be
able
to
do.
D
F
G
It's
not
too
far
out
so
I've
I
I
talk
to
and
now
I
forgot.
His
name
I
talked
to
the
in
an
AWS
employee
at
Defcon
in
Bruno
recently,
and
he
promised
me
some
arm
worker
notes
that
will
we
haven't.
We
haven't
attempted
to
to
look
into
that,
but
we'll
we'll
onboard
those
worker
nodes
as
remote
workers
to
our
build
farm,
and
when
that,
when
that
happens,
we'll
be
able
to
do
multi-arch
builds.
G
Is
currently
on
on
PTO,
but
once
he's
back
I
hope
we
can
drive
that
forward
again.
So
that
is
definitely
a
thing.
We
we
very
very
much
want
to
get
done,
and
so.
D
G
To
answer
your
question:
keep
multi-arch
in
mind
like
it.
Doesn't
it's
not
going
to
be
built
multiage
from
from
the
off
the
bat,
but
soon
Niche
will
will
have
it
so
keep
it
in
mind,
and
if
we
see
anything
obvious
that
won't
work,
we'll
probably
you
know,
add
a
to-do
to
it
or
make
sure
it
just
would
work
theoretically.
F
B
Thank
you.
It
sounds
like
what
we
want
to
do,
then,
is
maybe
dedicate
maybe
half
of
the
next
community
meeting
to
actually
fleshing
out
where
we
want
to
start
with
this
and
get
something
going.
Yeah.
F
Because
what
I'm
trying
to
do
is
work
out
the
end
to
end
for
an
operator.
What's
the
end-to-end
sort
of
pipeline
need
to
look
like,
and
then
we
can
start
implementing
the
individual
tasks.
B
Yeah
excellent:
we
did
get
a
question
on
a
different
topic:
Neil's
mic
isn't
working,
but
I
want
to
flesh
this
out
for
people
that
might
watch
the
video
that
might
be
interested
in
this
one's
for
Timothy
or
whomever.
Since
apparently
things
crash,
whenever
I
try
to
use
my
mic
I'm
going
to
ask
this
in
chat,
can
we
switch
to
run
and
c
groups
V2
for
okd
on
both
s
costs
and
F
costs,
Timothy
and
Christian.
D
And
Sue
yeah
I
think
but
yeah
it's
siren.
It's
in
413
I,
think
it's
available
and
but
not
enabled
by
default
and
I.
Don't
know
the
plans
for
14
minutes.
Yeah
I
think
it's
a.
G
G
Also,
one
of
the
reasons
a
couple
of
folks
have
hit
problems,
and
at
least
it's
there's
a
lot
of
chatter.
On
on
our
issue,
tracker
regarding
that
c
groups,
V2
thing
so
I
I
do
think
it'd
be
nice
to
align.
G
Although
I
wouldn't
want
to
switch
over
okd
4.13
to
c
groups
V2
on
s-course,
if
the
r
cross
one
isn't
it
doesn't,
doesn't
do
it
so
I
think
for
s-cos,
we'll
do
it
when
ocp
does
it,
which
is
I,
think
it's
going
to
be
in
Fallout
14,
even
though
it
might
still
be
yeah,
it's
it's
either
photo
14
or
4.15
I
hope
it's
photo
14
personally,
but
I'm.
Maybe
Timothy
has
found
out
when,
when
that
default,
the
switch
of
defaults
is
going
to
happen
and.
D
Yeah
I
linked
I
could
I
should
paste
that
into
the
notes.
You're
right,
the.
D
F
Because
I'm
guessing
anybody
that
wants
a
more
stable,
they
may
want
to
stay
with,
what's
known
but
as
Neil
said,
the
more
people
that
we
can
actually
get
using
this
new
feature,
the
more
information
we
have,
which
you
can
then
feed
back
so
I
think
it
would
be
good
if
we
have
something
really
clear
because
out
of
the
bat
I
wouldn't
know
how
to
turn
it
on,
I'd
have
to
go
and
do
a
lot
of
digging
to
find
out
how
to
turn
it
on
so
yeah.
We've
got
a
link
there.
G
Doesn't
it
doesn't
do
that,
so
it's
it's
only
F
cross
I,
at
least
as
far
as
I
know
and
I'm,
not
sure
what
the
default
in
in
all
of
openshift
is
going
to
be
in
Fallout
4T.
So
we
will.
We
don't
want
to
diverge
I
think,
but
it's
it's
possible
to
relatively
easily
switch
that
even
during
install
it's,
it's
just
one
kernel
argument
that
you
have
to
change.
Essentially,
so
really,
if
the
ocp
PMS
want
data,
they
should
set
up
clusters
with
c
groups
V2
they
can.
G
They
can
do
that
with
okd
or
with
ocp
I
I.
Don't
really
see
why
we
should
change
the
default
for
everybody
to
something
that
might
I
mean
I'm,
pretty
sure
sequence.
Youtube
is
good
enough
already,
but
and
we
do
it
in
fcos,
but
in
s-cos
we
are
really
strictly
aligning
with
the
product.
That's
just
it
wouldn't
make
sense
to
diverge
here.
I
think.
E
B
B
All
right,
I
think
we're
at
the
end,
feel
free
to
reach
out
folks
have
any
questions
or
comments
and
Community
Development
meeting
next
week,
but
then
we'll
focus
on
the
video
series
and
working
out
the
last
few
details
of
oh
Christian.
Do
you
have
something
else,
yeah
go
ahead.
G
Yeah
just
very
quickly
for
Neil
Neil.
Would
you
I
think
this
is
important,
so
would
you
mind
creating
an
issue
on
the
okd
board
in
jira,
so
we
can
include
that
in
our
refinement
and
work
on
it
internally.
B
All
right
thanks
folks,
so
much
and
we
will
focus
on
video
and
operator
stuff
at
the
next
meeting
at
the
next
working
Community
Development
meeting,
and
thank
you
for
your
time.
We'll
talk
to
you
again
soon.
Bye.