►
From YouTube: CentOS Automotive SIG meeting - 2021-09-16
Description
Meeting of the CentOS Automotive SIG on Sept 16, 2021
A
So
this
is
the
second
meeting:
we've
got,
I'm
hoping
that
everybody
was
able
to
finally
make
it
in
this
time.
I
apologize
for
the
the
telecommunications
problems
from
last
time
because
they
were
they
were
terrible
and
we
will
not
be
using
that
particular
solution
again
we'll
be
going
ahead
with
this
one,
but
let's
go
ahead
and
dive
in,
because
I
want
to
make
sure
that
we
cover
everything.
A
I
put
together
a
brief
agenda.
I
re
the
first
four
bullets
are
things
from
the
previous
agenda
to
make
sure
that
everybody
gets
them
and
we
can
buzz
through
them
pretty
quickly,
but
I
do
want
to
make
sure
these
things
are
covered,
so
diving
in
let's
talk
about.
Why
do
we
have
an
automotive
stick
in
centos
and
that's
a
really
good
question.
A
The
primary
reason
is
the
advancement
of
distro
based
automotive
linux.
There
are
been
a
lot
of
of
attempts
at
automotive
linux.
There
have
been
a
lot
of
great
successes
in
the
case
of
agl.
There
have
been
a
lot
of
people
building
parts
of
automotive
linux
systems
in
linux
and
in
the
linux
ecosystem,
and
what
we
would
like
to
do
is
try
to
get
just
get
started
using
a
distro
as
the
base,
which
is
somewhat
different
from
the
way
that
other
people
have
approached
it.
A
We
want
the
centos
sig
to
be
a
focal
point
for
working
with
other
upstreams
we'd
like
to
bring
in
as
much
hardware
as
possible
and
would
like
to
be
in
contact
with
other
related
projects.
I
know
that
there
are
some
new
projects
starting
up
and
some
existing
projects
out
there,
that
that
are
really
useful,
and
so
we
want
to
be
in
contact
with
them
and
bring
them
in
either
as
upstream
or
adjacent
projects,
or
in
some
cases
is
standards
bodies.
A
This
sig
will
be
a
public
development
point
for
proofs
of
concept,
we'd
be
putting
together
a
variant
of
centos,
sometimes
also
called
a
spin,
which
will
be
essentially
like
a
reference
platform.
I
would
not
expect
somebody
to
take
it
and
dump
it
into
a
product,
but
it
will
certainly
be
the
place
where
we
can.
We
can
look
at
it
and
and
play
with
things
and
and
test
out
solutions.
A
We
want
to
be
make
sure
that
we
are
fully
transparent,
that
the
reason
that
red
hat
wants
this
stick
to
exist
because
we
intend
to
use
it
as
an
upstream
for
our
eventual
automotive
product.
We
do
not
have
an
automatic
product
at
the
moment,
but
it
is
coming
and-
and
this
stig
is
the
place
where
we
want
to
do
the
work,
because
it's
it's
part
of
red
hat's
principles
and
guiding
values
to
do
this
kind
of
work
in
public,
using
the
open
source,
best
practices.
A
A
There
will
be
things
that
happen
in
an
automotive
context
that
aren't
really
appropriate
for
for
rel
and
so
the
place
that
we
would
do.
Those
is
either
in
those
upstream
projects
or
here
within
the
centos
sig,
and
given
that
one
of
the
principles
that
red
hat
operates
behind,
that
is
an
open
source.
Best
practice
is
to
move
upstream
first
and
to
go
as
upstream
as
possible
to
do
the
work.
A
We
will
be
doing
a
lot
of
work,
I
think
in
other
upstreams
and
then
bringing
that
work
into
the
centos
cig,
at
least
that's
the
the
theory
given
that
we're
only
four
weeks
into
it.
I
I
don't
have
an
example
to
show
it
yet,
but
I
think
that
that
is
exactly
what
we
will
be
doing,
and
red
hat
has
a
long
history
of
of
working.
In
this
way,
I
could
also
see
some
work
being
done
in
fedora.
A
In
fact,
I
would
like
to
see
that
very
much,
but
but
dentos
is
a
stable
enough
base
and
the
centos
stream
process,
I
think,
is
going
to
be
very
useful
for
us.
I'm
hopeful
that
we
can.
We
can
do
all
of
that.
Any
questions
so
far.
If
you
have
any
questions,
feel
free
to
just
mention
them.
I
don't.
I
cannot
see
the
chat
at
the
moment.
So
if,
if
anybody's
like
making
cool
snarky
comments
on
this,
though.
A
The
main
thing
that
we
want
to
just
kind
of
get
our
feet
wet
with
right
now
is
working
on
this
centos
variant.
This
is
a
mistake
on
this
slide.
I
don't
think
it'll
be
based
on
rail,
for
edge
it'll,
be
based
on
centos
and
we'll
have
a
regular
release
schedule.
I
hope
hopefully
we'll
get
the
ian,
and
possibly
pierre
can
talk
about
that
in
a
few
minutes,
so
some
specific
goals
through
the
end
of
the
year
2021.
A
I
wish
this
wouldn't
build.
I'm
just
going
ahead
and
fill
this
out,
so
we,
the
the
main
thing
that
we
want
to
do
here,
is
to
build
a
foundational
community
and
that
is
to
establish
working
relationships.
Among
all
of
you,
everybody
who
shows
up
and
to
bring
people
into
the
conversation
who
we
think
would
be
useful,
who
would
benefit
from
it
and
who,
for
whom
the?
Who
would
who
would
who
would
contribute
to
this
particular
conversation?
A
A
A
As
far
as
the
code
of
conduct
goes,
we
are
subject
to
the
red
hat,
sorry
to
the
centos
code
of
conduct
and
and
that
will
guide
our
behavior
and
the
way
that
we
that
we
treat
each
other
in
these
meetings.
You
can
take
a
look
at
it.
It
is
online
on
the
centos
site.
It
is
intended
to
be
both
welcoming
and
to
create
a
safe
place
for
everybody,
and
if
there
is
anything
that
happens
in
this
cig
or
elsewhere
in
centos
that
you
feel
is
unsafe,
please
feel
free
to
come
to
me
privately.
A
A
We
will
need
to
vote
for
some
administrative
roles.
The
idea
when
I
made
this
slide
a
month
ago,
was
that
we
would
select
and
vote
for
them
in
the
last
meeting.
Everybody
kind
of
took
one
step
back
and
left
me
standing
in
line
by
myself,
so
I
think
that
I
will
be
probably
functioning
as
the
the
person
who
puts
slides
together
and
stumbles
through
the
initial
discussion
in
every
meeting
for
the
next
few
months,
at
least,
and
that's
a
role
that
I'm
happy
to
take.
A
So
another
specific
goal
is
to
develop
plans
for
this
variant
and
to
create
the
build
and
test
infrastructure
and
we'll
hear
in
a
few
minutes
exactly
how
that's
going
and
then
to
do
a
presentation
at
automotive
linux
summit.
This
is
a
note
that
I
believe
the
the
call
for
participation
for
automotive
linux
summit
ends
on
sunday.
B
C
A
A
I
don't
know
if
anybody
from
who
is
on
the
program
committee
is
on
the
call,
but
if
you,
if
you
are,
if
you
have
any
desire
to
say
something,
let
me
know
I'd
be
happy
to
give
you
the
floor
for
a
few
minutes.
I
I
think
it's
it's
a
great
conference.
I
have
listened
to
many
many
of
the
the
talks
that
have
gone
on
at
als.
D
D
It's
it's
a
mix,
so
it
really
does
tend
to
be
a
mix
we
do.
We
do
tend
to
have
people
doing
deep,
dives
into
things
that
they've
done
either
with
agl
or
with
other
linux
bits
to
get
into
cars,
and
then
we
we
tend
to
have
some
higher
higher
level.
You
know
presentations
around,
what's
been
going
on
in
automotive
grade
linux
and
other
automotive
projects
around
the
around
the
world.
So
it's
a
mix.
A
Cool
yeah,
it
seems
to
me
to
be
the
the
tone-
seems
to
be
very
much
very
similar
to
embedded
linux
conference
in
that
regard.
Where
you
get
some,
you
know
a
mix
of
high
level
stuff
and
really
deep
down
in
the
weeds,
low
level,
stuff
and
and
yeah.
It's
a
great
it's
a
great
conference.
I
enjoy
it
very
much.
I
enjoyed
it
even
before
I
was
working
in
an
automotive
software.
D
D
D
A
Okay,
excellent:
I
know
that
I
I
will
be
there
virtually.
I
I'm
in
canada
and
travel
is
not
allowed
from
my
country
to
back
to
the
us
at
the
moment
so,
but
so
I
will
be
attending
virtually,
but
I
do
plan
to
be
there.
Do
you
know
if
the
buffs
are
going
to
be
kind
of
simulcast
or
webcast,
or
if
they're
going
to
be.
A
D
A
F
A
Sig,
if
anybody
has
something
that
they
would
like
to
add
to
this
list,
whether
it
involves
stickers
or
not,
let
me
know,
and
we
can
make
sure
to
get
that
up.
Also
all
of
these
things
exist
on
the
wiki,
the
content
for
these
slides
and
the
notes
from
the
last
meeting.
The
notes
for
this
meeting
will
be
on
the
on
the
centos
wiki.
That
is
so
definitely
go
and
take
a
look
and
add
things
you
can,
since
it's
a
wiki,
you
can
change
things
if
you
wish.
A
So
one
of
the
things
that
we
were
going
to
be
doing
was
talking
about
the
decision
making
process
I
right
now.
We
don't
really
have
a
whole
lot
to
make
decisions
about
the
most.
The
thing
that
we're
mostly
making
decisions
about
is
how
we're
all
going
to
work
together
and
the
main
idea
is
that
let's
decide
this
together,
this
is
not
something
where
one
person
or
one
company
or
one
group
is
going
to
be
handing
down
a
decision
making
process.
A
A
We
do
have
some
community
standards
that
are
are
already,
some
of
which
are
already
set
up
and
some
of
which
we
need
to
set
up.
We
do
have
mailing
list,
there's
a
link
to
the
main
list
mailing
list
on
the
wiki
page.
There
is
an
irc
channel
that
has
been
quiet
for
several
weeks.
I'd
be
glad
to.
I
would
be
very
happy
to
see
conversation
start
happening
there.
If
they
aren't
going
to
happen
there,
there's
no
requirement
that
we
have
to
use
that.
A
That's
always
fun.
There
is
this
periodic
meeting.
This
particular
formal
meeting
happens
about
once
a
month.
What
I
would
like
to
start
on
the
alternating
two-week
cycle
in
this
same
time,
slot
or
possibly
in
a
different
time
slot,
because
we've
been
asked
to
to
mix
up
the
times
a
little
bit
so
to
enable
other
people
to
join,
is
to
set
up
an
office
hours.
That's
more
informal
that
maybe
is
just
we'll
just
open
up
a
call.
A
We
won't
have
slides
and
we
will
just
open
it
up
to
general
conversation
for
an
hour,
so
so
we'll
be
I'll,
be
setting
that
up
after
this
meeting
and
we'll
send
a
note
out
to
the
mailing
list
on
this
group
and
I'd
still
need
to
make
sure,
I
still
need
to
figure
out
how
to
get
things
into
the
centos
calendar
seems
to
involve
incantations
and
sacrificing
animals
and
things,
but
we'll
figure
it
out
code
of
conduct.
We
already
discussed.
A
A
A
A
G
G
G
G
Try
with
a
video
to
see
that
helps,
so
we
don't
really
know
what
the
future
of
the
the
packer
platform
and
software
will.
The
second
aspect
to
this
is
that
there
are
so
already
being
work
done
to
build
a
ci
pipeline,
something
that
basically
every
time
there
is
a
change
in
center
streams
or
every
time
there
is
a
change
to
the
automotive
ci,
the
automotive
sig
project.
G
It's
it's
going
to
build
all
the
images
to
see
if
they
still
work,
so
there
has
already
been
work
being
done
there
and
that
work
was
not
based
on
using
the
pager
project.
So
if
we
were
to
choose
that
one,
we
would
have
to
basically
do
a
little
bit
of
work
to
adjust
the
work.
The
second
platform
is
github.
That
is
the
first
platform
where
things
were
worked
upon
using
github
actions,
but
using
getting
options
means
it's
you
you're
very
much
tied
to
the
to
the
platform.
G
F
G
G
A
G
Annoying
that
is
something
that
the
center's
project
is
looking
to
fix,
but
that
is
going
to
require
a
few
months,
probably
discussions
and
figures
things
out
between
centers
right
out,
and
so
one
thing
we
could
do
today
is
actually
it's
grab
space
under
the
red
hat
name
space
of
gitlab,
so
that
will
be
something
like
club,
dot,
com,
slash
credits,
slash,
automotive,
slash,
automative,
stick
a
little
bit.
You
know
we
have
twice
automotive
in
there,
but
at
least,
but
it
does
give
us
the
flexibility
of
having
other
repos
under
that
space.
G
If
we
want
there
is
pros
and
cons
to
this
approach
as
well.
The
pros
is
that
that's
something
we
can
quickly
set
up
today
and
start
publishing
our
work
and
documentations
there.
The
cons
is
that
only
people
with
reddit
accounts
will
be
able
to
have
commit
access
to
the
project.
So
anybody
in
the
community
that
wants
to
contribute
will
have
to
go
via
merch
that
can
work
with
accounts,
but
that
is
not
very
community
friendly
as
a
wall.
No.
G
I
can
so
I
can
conclude
with
my
recommendation
and
then
I
let
anybody
voice
their
opinion
on
this
is.
I
would
recommend
that
we
do
gitlab.com
automotive
relatives,
automotive
sig
now,
because
this
is
the
place
where
it
gets
us
something
out
in
the
open
way
that
we
can
recommend
as
this
that
we
can
present,
as
this
is
what
the
sig
is
working
on
in
the
meantime.
In
the
same
time,
we
we
mirror
that
project
under
the
center
slam
space.
G
G
B
Okay,
not
good
to
me,
and
it
also
sounds
like
it's
a
good
problem
to
solve
more
generally
for
the
idea
of
centos
cigs
like
getting
this
sorted
out,
I
can't
imagine
any
other
group
of
this
type.
That's
going
to
want
to
encourage
more
open
collaboration,
we'll
want
to
have
an
open
account
system,
essentially.
C
C
G
D
G
G
And
there
is
one
other
thing:
is
I've
started
to
work
on
documentation
for
the
for
basically
what
the
sig
has
been
up
to
until
now,
I
have
a
proof
of
concept,
or
example,
of
what
this
documentation
could
look
like,
and
I've
shared
the
link
in
the
chat
here,
my
home
server.
It
may
be
slow
to
answer,
but
it
is
also
steady
casting
and
it's
still
working
progress,
not
all
the.
B
And
just
to
add,
like
just
in
case
the
broader
context
of
this
wasn't
clear.
What
we're
hoping
to
demonstrate
here
is:
we've
essentially
been
creating
hypothetical
software
loads
of
what
we
think
an
automotive
load
derived
from
rel
content
might
look
like,
and
we
want
to
make
this
visible
in
the
sig.
We
want
this
to
be
one
of
the
first
things
that
we
show.
B
So
it's
not
there's
nothing
earth-shattering,
but
it's
a
way
for
us
to
start
to
understand
what
sub-components
of
rel
are
likely
to
be
relevant
for
automotive
and
also
to
show
how
some
of
the
tooling
that
we've
already
created
for
edge
can
be
used
to
create
these
images.
So
that's
the
that's.
The
broader
goal
of
this
initial,
offering
up
through
the
sig.
C
G
C
Yeah
right
now,
the
convention
is
that
we
create
a
kernel
package
called
kernel,
rt
kernel,
dash
rt,
so
usually
the
place
that
things
freak
out
is
somebody
presumed
that
the
kernel
package
is
going
to
be
named
colonel
and
we
need
you
to
make
sure
that
any
tooling
is,
you
know,
adapted
to
deal
with
something.
That's
not
just
kernel
but
kernel
dash,
rt,
or
we
have
been
toying
with
the
idea
of
calling
this
package
kernel
dash
auto
to
distinguish.
F
F
C
You
know,
let's,
let's
thrash
it
out,
if
there's
some
better
name
or
they're
happy
to
accommodate
there,
but
I'm
just
trying
to
figure
out
how
to
get
the
right.
Now.
If
you
look
up
on
the
centos
stream,
none
three
there's
a
there
are
two
branches.
There's
main
and
main
rt
you
might
expect.
Maine
is
the
is
the
main
centos.
C
C
C
That's
just
the
way
it's
been
staged,
but
that
means
that
we're
going
to
churn
quite
a
bit
in
the
rail
and
centos
world,
because
eventually
we
want
to
run.
We
want
to
be
able
to
build
all
this
stuff
out
of
the
same
branch.
We
don't
need.
We
don't
really
want
to
have
a
main
rt
branch.
We
just
want
to
have
a
main
and
be
able
to
say,
use
these
configs
and
make
me
a
rail
kernel
or
make
me
an
lrt
kernel,
or
maybe
an
automotive
kernel.
C
So,
but
that's
going
to
be
into
next
year
before
we
can
hope
to
resolve
all
this,
and
so
I
I
hope
I
hope
people
are
not
disappointed
in
us
that
we
don't
have
a
better
plan
to
to
roll
this
stuff
out
and
make
it
available.
But
unfortunately
it's
it's
we're
really
churning
and
we're
just
gonna
have
to
kind
of
roll
with
it
and
adapt.
C
A
A
H
H
H
C
H
C
H
C
C
A
F
F
E
C
E
A
A
C
I
just
kind
of
wanted
to
throw
that
out.
Just
so
everybody
knows
where,
where
I'm
coming
from,
and
originally
we
thought
oh
well,
you
know
we'll
build
a
file
system,
you
know
like
x4
or
xfs
or
whatever.
We
pick
we'll
build
that
in
the
image.
So
we
don't
have
to.
But
it's
it's
a
it's
a
stupid
fraction,
whereas
we
got
800
milliseconds
just
being
taken
out
by
device
initialization
in
the
firmware
before
we
even
get
to
loading
linux.
C
C
If
we
come
to
conclusion
and
say
well,
we
need
these
six
modules
built
into
the
rather
than
as
modules
yeah.
That's
easy,
so
that
so
I
guess
right
now
the
the
the
general
tenant
is.
Everything
is
a
module
for
the
moment
and
we'll
come
up
with
a
list
of
the
specific
modules
we
care
about.
Building
and
posting
wiki.
B
B
I
You're,
first,
okay,
now
yeah
with
respect
the
you
know,
we're
talking
about
areas
of
focus
right,
so
I
was
thinking
you
know.
We
know
that.
I
think
that
from
an
automotive
point
of
view,
it
would
be
you
know
quite
important,
for
instance
to
for
for
the
centers
automotive
distribution,
support
the
autosar
vendors
right
and
to
allow
them
special.
You
know
to
experiment
and
collaborate
with
the
auto
consortium.
I
Now
I
don't
know
how
to
do
that
practically
because
you
know
it's
I
mean
centos.
Automotive
is
like
it's
an
open
project,
whereas
the
autosar
consortium,
it's
not,
however,
the
autosar
specification
and
the
explanatory
document
of
the
architecture.
These
are
public
right
so
and
effectively
they
are,
you
know
they
can
be
evaluated
with
respect
automotive,
distro
of
centers.
On
the
flip
side,
you
know
yeah.
I
think
it
would
be
good.
I
You
know
if,
if
we
were
able
to
to
to
to
build
a
collaboration
really
with
the
with
the
with
the
consortium,
so
that,
for
example,
when
they
deliver,
you
know
the
the
the
demo
framework
you
know
we
may
be
able
you
know
to
to
use.
You
know
santos,
os
automotive
distro,
also
as
a
sort
of
one
of
the
test
distribution.
Ideally
now
I
don't
know
if
it
is
possible,
like
politically
you
know,
and
with
respect
to
the
all
you
know
constraints.
So
I'm
not.
I
C
You
know
gaba,
it
looked
to
me
like
a
ton
of
the
auto
stars,
auto
auto
star,
auto
sar,
and
I
keep
wanting
to
say.
Auto
zar
is
in
russian
old
ruler,
because
that's
the
way
it
sounds
like
whenever
dimitri
says
it.
So
whenever
you
hear
adesara
to
me,
you're
just
thinking
of
the
ancient
russian
ruler,
but
it
looked
to
me
like
the
autosar
stuff
came
from,
was
basically
drawn
from
a
lot
of
the
ieee.
The
posix
specification.
C
A
A
C
A
I
I
I
mean
my
idea
was,
like
you
know,
to
to
provide
a
sort
of
to
start
a
collaboration
to
to
to
optimize
the
features
that
may
be
required
by
the
how
to
start
vendors-
and
you
know
to
you,
know
to
even
you
know
to
maybe
to
work
together.
You
know
to
improve
these
reference
demos
so
that
they
can
be
like
maybe
a
stronger
baseline
right,
yeah.
F
A
C
A
E
C
F
J
Yeah,
I'm
sorry
I'm
in
a
noisy
place.
I
just
come
out
of
visiting
the
new
simulation
facility
of
bmw
in
munich,
which
was
really
cool
and
they
have
14
a
new
simulator
in
place.
So
if
there
are
two
level
in
autosar,
you
have
autos
are
classic,
and
I
think
we
can
forget
about
this
one,
because
this
is
for
more.
J
You
know
specification
and
you
have
to
to
have
formal
proof
of
implementation.
So
this
is
not
going
to
fly
with
linux,
and
then
you
have
autozar
classics
that
rely
on
posix
interface
and
for
this
there
is
one
specification
that
is
named
ara,
which
is
the
one
that
is
doing
the
gateway
in
between
the
classic
world
and
the
project
world.
J
I
I
C
F
Yeah,
I
understand
that
you
fundamentally
initially
like
what
we
want
is
is
to
get
the
allosaur
vendors
to
run
on
on
this,
but
long
term.
It's
kind
of
would
be
very
nice
to
have
additional
you
can.
If
this
is
fundamentally
what
you're
going
to
use
to
run
a
car
and
we're
not
allowed
to
ship
it.
It's.
B
B
C
B
B
B
B
B
Guys,
but
we
could
add
them
to
people
or
we,
but
here
we
I'm-
and
this
is
this-
is
speaking
more
towards
the
relationship
between
the
sig
and
the
offering.
But
the
reason
we
have
stream
and
the
goal
of
stream
is
to
have
a
state,
an
area
that's
very
close
to
what's
actually
going
to
become
released,
but
have
a
way
for
people
to
see
it
and
also
collaborate
on
it,
and
we
would
want
to
do
the
same
thing
here,
whereas
epil
doesn't
that
epo
charter
is
a
little
different.
B
So
that's
the
that's
my
point
is
we
would
there
will
be
things
that
the
sig
will
be
the
home
for
that
we
want
to.
We
want
to
be
closely
tied
to
the
types
of
workflows
that
we
have
in
centos
stream
that
will
be
unique
to
automotive
there'll,
be
an
automotive,
optimized
kernel
and
there'll
be
packages
associated
with
it
as
well.
So
that's
just
me
attempting
to
high
level
that
won't
just
be
a
kernel.
H
I
was
thinking
that
what
we
would
do
is
we
would
end
in
a
multi-stage
setup.
We
would
do
stuff
in
the
sig
first
get
it
proven
and
then
work
on
moving
it
up
into
fedora
and
then
into
apple
for
the
stuff
that
really
matters.
There
are
certain
things
that
really
matter.
There
are
certain
things
that
you
know
we
could
put
a
bunch
of
time
into
some
little
part
and
find
out
it
own.
H
Well,
yes,
we
are
upstream
for
automotive,
but
there
is
stream
upstream
yeah
upstream
upstream
from
the
stream
stream
yeah
I
mean
I
mean
I
as
a
kernel
developer.
You
send
something
to
substream
person,
okay,
yeah
and
then
they
did
to
lioness
or
they
send
it
up
to
the
next
layer,
and
then
they
send
it
to
lioness
part
of
that
fish.
Going
upstream
bless.
You.
A
C
B
B
C
Guess
what
I'm
saying
is,
could
we
could
we
flesh
out
the
centos
some
of
the
centos
big
wiki
pages,
with
a
start
of
these
documents,
because
we've
been
looking
for
a
place
to
actually
put
this
stuff
and
collaborate
for
a
while?
Why
not
just
use
the
sig
wiki
and
say:
okay,
look!
Here's
what
we're
going
to
do
with
the
kernel
and
everybody
can
argue
and
express
their
opinions
and
we
can
all
come
to.
You
know
consensus
and
come
up
with
a
good
config
for
a
kernel
in
automotive.
B
My
my
my
preference,
what
I've
been
pushing
for
is:
let's,
let's,
let's
show
that
in
the
form
of
a
working
artifact
as
the
reference
point
and
then
discuss
it.
On
top
of
that,
much
like
you
could
do
that
with
the.
I
think
you,
you
have
a
few
it's
easier
for
you
to
put
together
a
handful
of
guiding
principles,
which
I
think
you
have
done
a
good
job
and
it's.
B
Document
and
then
the
reason
the
other
document
looks
ugly
is
is
basically
trying
to
like
in
have
a
kernel
config,
as
in
in
document
discussion
forum.
When
really
right,
I
don't
know
at
some
point,
it
becomes
better
to
just
have
it
be
the
config
and
the
kernels
that
that's
compiled,
and
you
can
comment
in
that
context
precisely
I.
A
Yeah
that'd
be
interesting,
so
we
are
exercising,
and
I
know
I
know
that
I
have
a
hard
stop
as
well,
as
I'm
sure,
a
lot
of
other
people
do.
Let's
continue
this
discussion
on
the
mailing
list.
Let's
get
some
of
these
wiki
pages
rolling
and
I
will
also
set
up
and
send
out
a
meeting
invitation
for
two
weeks
from
now
to
have
an
open
house
or
office
hours
or
whatever.
We
want
to
call
it
so
great
and
we
will
continue
the
discussion
about
the
infrastructure
as
well
on
the
mailing
list.