►
From YouTube: 20210527 SIG Arch Code Org
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
Hi
everybody
today
is
may
27th,
and
this
is
the
meeting
for
the
kids
code
organization,
we're
going
to
keep
today
light
and
informal,
and
so
we'll
just
talk
whatever
comes
to
mind
and
if
needed,
then
we
don't
have
like
a
formal
agenda
for
today.
So
we
can
do
what
everybody
wants
to
do.
So,
let's
get
started
with
arsh
top
left
on
my
screen.
B
A
Yes,
we
are
not
going
to
touch
it,
so
this
is
like
the
character
in
the
stick,
we're
going
to
tell
them
that
we
are
not
going
to
merge
anything
until
they
fix
it.
On
their
end,
however,
one
thing
that
you
could
do
is
in
the
prometheus
common
repository
in
their
go
mod.
Add
a
replace
with
the
link
to
the
new
directory.
A
The
goku
folks
stripped
out
one
directory
and
they
created
a
separate
repository
with
a
separate,
go
mod
file
and
it's
called
go
kit.
Go
dash,
kit,
slash
log
right
and
the
prometheus
common
still
refers
to
the
main
goalkit
repository
so
add
a
replace
and
see
if
everything
works
and
create
a
work
in
progress,
pr
pointing
to
the
new
repository
which
is
goku
log.
A
If
you
couldn't,
if
you
can't
figure
out
just
ping
me
and
I'll
walk
you
through
it,
okay,
so
this
is
like
forcing
the
function
and
telling
the
prometheus
common
folks
like
look.
We
got
to
do
this,
so
here
is
a
pr
that
uses
the
new
repository
instead
of
the
old
repository
kind
of
thing:
okay,
okay,
so
that's
good
eddie!
Did
you
have
anything
today
that
you
wanted
to
talk
about.
C
A
Okay,
well
there's.
There
was
one
thing
that
I
wanted
to
talk
to
you
about
at
some
point,
was
we
got
to
push
the
sig
release
folks
to
come
up
with
a
plan
for
when
cube
cuddle
gets
out?
A
You
know
command,
slash,
cube
cuddle
getting
out
into
the
cube
cuddle
repository
kubernetes
flash
cube
cutter
right.
So
how
would
then
cube
cuttle
be
included
in
the
kubernetes
release?
A
Right
yeah?
That's
a
good
question,
so
we
need
a
plan
for
that
and
the
simplest
plan
that
I
can
think
of
is
we
would
vendor
in
cucurl
and
then
the
kk
will
look
in
the
go
mod
and
just
pick
up
whatever
is
in
there
right
and
then
we
would
update
just
like
we
update
any
other
things
that
we
render.
C
A
Yeah,
but
this
is
the
last
thing
that
we'll
have
to
do
is
so
the
idea
here
is
q.
Cuddle
should
have
its
own
release,
cadence
one
specific
version
of
cube
cutter.
We
will
include
in
kkk
just
like
c
advisor
just
like
a
cd,
so
same
same
fashion,
okay,
so
check
with
mate
and
see
if
we
can
write
something
up.
A
A
Sounds
good,
hey
mike
navi
go
for
it.
D
So
you
mentioned
like
we
would
render
cucuta
as
a
as
any
other
dependency.
So
does
this
mean
like
the
releases
will
need
to
be
then
in
order
like
for
cube,
cuddle
needs
to
go
first,
so
that
that
the
release
button
can
then
be
rendered
into
kk
nope.
A
A
And
that's
why
I'm
saying
we
should
write
it
up
so
we'll
know
what
the
pros
and
cons
are
right.
Naveed.
Does
that
answer
your
question.
F
No,
I
was
just
coming
in
and
see
it
see
how
things
were
going.
I
know,
there's
been
a
lot
of
this
kind
of
discussion
happening.
All
right.
I
needed
to
check
in
and
see
what's
going
on.
I
think
this
version
skew
stuff
is
just
going
to
get
more
and
more
complicated.
We
probably
need
to
start
requesting
some
of
our
vendor
tools
to
to
provide
a
version,
inversion
information
for
the
version
of
kubernetes
that
they
support.
F
For
example,
if
I
do
a
you
know,
container
d
version,
I
should
be
able
to
see
yeah,
it's
run
c
95,
it's!
You
know,
container
d.
You
know
this,
the
you
know
here
and
then
kubernetes
here
right,
because
we're
being
used
more
and
more,
of
course,
and
but
yeah,
I
think
the
version
skew
is
going
to
be
an
issue
right.
Certainly
we
can.
We
can
also
show
we
already
show,
I
think,
the
cry
version.
F
I
have
to
go
check,
but
they
give
him
that
for
rendering
that
which
one
do
we,
I
think
we
want
to
support
more
than
one
cry
version
at
a
time
now
I
gotta
do
that
pr
but
yeah
I
don't
know
and
then
there's
the
there's
a
lot
of
reorganization,
that's
going
to
need
to
happen
for
confidential
computing.
So
I
always
wanted
to
check
in
to
see
how
things
were
going
in
that.
A
F
E
A
F
A
So,
on
that
front,
what
we
ended
up
doing
was
we
were
stuck
because
hcd
was
not
able
to
update
its
own
grpc
version
so
between
ip
and
continuity
and
kk.
We
had
this
cycle.
F
A
F
A
F
Are
you
gonna?
Are
you
or
somebody
else
already
pushing
the
grpc
update
in
container
d,
so
you
can
do
the
same
thing.
A
A
So
so
on
that
front
we
still
have
to
deal
with
hcs,
shim
and.
F
A
Yeah,
oh,
oh,
the
run
c1
has
kind
of
settled
down
with
the
rc-95,
so
we're
gonna
do
that
for
sure.
Okay,
so
run
c,
hcs
continuity
version.
Those
kinds
of
things
need
to
get
updated
in
c
advisor
and
then
then
the
whole
enchilada
will
move
to
kk.
So
so
that's
it
so
yeah,
not
a
problem,
the
other
one.
I
don't
know.
If
did
continuity
already
move
to
latest
pause
image?
3.5,
I
don't
think
so.
Okay,
we
we
made
the
change
last
week
here.
A
So
we
can
go
and
do
the
same
change
in
continuity
and
you
know
get
it
ready
for
the
next
release.
Okay,
that's
about
it
with
respect
to
continuity!
A
So,
mr
delorme,
how
do
I
say
your
name?
I'm
sorry,
gucci.
G
A
I
can
try,
go
tear
yeah,
perfect
sounds
good,
close
close,
so
I
I
know
you
saw.
I
saw
your
interest
in
that.
A
Pr
from
paco
with
respect
to
having
a
set
of
dependencies
that
are
forbidden
and
having
some
that
we
are
trying
to
like
keep
out.
A
Can
you
recap
for
everybody
here
over
there
so.
G
Basically,
the
goal
is:
there
are
two
goals.
The
first
one
is
to
to
have
a
way
to
state
like
we
don't
want
specific
dependencies
in
humanities,
mostly
that
dependencies
we
removed
from
kk,
and
we
don't
want
them
to
come
back.
A
good
example
recently
is:
go
open.
Api
spec,
for
example,
or,
like
other
graphing,
go
api
project,
but
also
like
the
old
version
of
k-log
and
some
like
go
mongodb
driver
or
like
some
weird
dependencies
like
that
yeah.
G
The
the
other
side
of
the
of
the
of
the
pr
of
the
challenge
is
to
be
able
to
specify
what
dependency
we
have
today,
but
we
don't
want
in
the
future.
Basically-
and
today
I
think
it's
only
like
the
dependency
you
were
talking
about
earlier
with
arch,
that's
go,
git
dependency,
ego
kits
the
dependency
so
yeah.
That's
that's
what
we
are
working
on
right
now,.
B
A
So
I
I
haven't
looked
at
paco's
pr
yet,
but
a
one
additional
thought
that
I
had
was:
we
need
to
have
a
way
to
say
that
we
shouldn't
go
to
an
older
version
of
something
that
we
already
use
a
new
version
for
you
know
what
I'm
saying.
G
Yeah
I
mean
you
are
talking
about
like
major
version
changes
right
like
I'm
just
talking
about
like
the
version
because,
like
you
know
like
in
go
like
when
you
change
like
the
major
version,
it's
actually
like
changes
like
the
the
moodle
name
right
I'll.
A
So
we
need
a
way
to
express
that
in
whatever
you
know,
files
that
will
come
up
did.
A
Yeah,
it
has
happened
before
where
not
specifically,
with
grpc
we've
had
other
places
where
you
know,
people
were
updating
something
and
we
you
know.
A
So
there
was
a
time
when
we
were
not
paying
too
much
our
attention
to
the
dependencies
as-
and
you
know,
we
were
basically
a
boat
load
of
dependencies
were
getting
in
and
most
of
the
dependents
didn't
even
have
tags,
but
for
the
last
year
or
two
we've
been
paying
attention
to
exactly
what
we
are
rendering
in
and
we
ended
up
requesting
the
continuity
folks
to
break
some
pieces
apart.
We
requested
the
docker
folks
to
break
some
things
apart.
A
We
cleaned
up
how
c
advisor
is
structured,
we've
and
plenty
of
other
things
like
we.
There
was
even
a
dependency
loop
between
continuity
and
kubernetes
kubernetes,
so
we
broke
that
yeah.
So
there
was
a
lot
of
work
that
was
done
to
like
streamline
the
dependencies
and
then
make
sure
that
we
go
to
tagged
versions
and
one
example
of
a
problem
we
used
to
have
before
was
run
c
was
we
were
using
a
shot?
A
Continuity
was
using
a
shot
and
then
kubernetes
was
using
a
sha,
and
so
if
there
was
a
problem
in
run
c,
then
just
figuring
out
like
how
many
places
you
have
to
fix
and
not
fix
it
was.
It
was
a
problem,
so
we
ended
up
pushing
for
like
using
a
specific
tag
and
we
are
trying
to
hold
the
line
on
like
when
new
dependencies
are
coming
in.
We
are
basically
telling
them.
No,
we
are
not
going
to
just
throw
a
random
shot.
You
need
to
give
us
a
tag
so
yeah.
G
A
Exactly
so,
we
are
trying
to
do
that
a
little
bit.
That's
good!
I
like
the
effort
right.
So
in
the
process.
We
are
doing
other
things
like
you,
know,
cleaning
up
licenses
and
keeping
track
of
what
we
drag
in
as
well,
so
anyway
long
story
but
yeah,
the
the.
So
the
point
here
is
like:
let's
try
to
see
what
else
we
could
be
doing
in
that
pr
and
the
main
focus
would
be
I'm
still
debating
whether
we
should
like
for
any
of
the
dependencies.
A
We
should
allow
a
regression
to
an
older
version
or
not,
and
but
then
at
least
we
should
start
with
making
sure
that
we
can
specify
don't
go
back
to
an
older
version
of
grpc.
If
not,
you
know,
at
least
that
dependency,
because.
A
Yeah
that
happened
with
the
yaml
repository,
for
example,
exactly.
G
A
Couldn't
go
to
the
newer
version
and
we
had
to
like
get
them
to
create
a
branch
and
backboard
something
revert
something
and
give
us
a
new
tag
there
so
yeah.
That
is
something
that
I
would
like
to
express:
okay,
just
work
with
paco
and
see
if
you
can
get
that
on.
A
To
pick
up
or
yes
mike
in
so
in
container
d,
there
was
some
error
propagation
we
wanted
to
do,
and
that
was
only
in
the
newer
version
of
grpc
and
we
basically
held
back
at
least
one
pr
or
two
pr's.
Regarding
that
I'll
dig
that
up
and
I'll
point
them
to
you.
So.
G
James,
like
do
you
really
want
like
to
block
like
paco
pr
on
this,
because
it
seems
like
the
kind
of
stuff
like
we
can
add
like
in
a
follow-up
pr
or
right?
Can
you
repeat
that
do
you
like?
Do
you
want
to
block
like
paco
pr
on
this
request
like
this
new,
like
basically
issue
because,
like
okay,
the.
A
Idea
here
is
when
somebody
opens
a
a
gomod
change.
We
should
just
tell
them
right
away
that
this
is
no
good
right.
You
are
doing
something
wrong
and
we
don't
even
want
to
spend
cycles
reviewing
it,
because
we
know
that
we're
going
to
say
no
right,
so
the
early
like,
for
example,
like
in
the
china
time
zone,
if
paco,
does
a
pr
right
right
away.
He
gets
like
a
thumbs
down
on
it.
A
Saying
like
we
don't
allow
this
you,
you
are
going
yeah
yeah,
so
it's
just
for
that
yeah
to
save
some
review
time.
G
A
Yeah
we
should
always
have
like
a
safety
valve
some
somehow
or
maybe
we'll
remove
the
line,
get
it
merged
and
yeah.
Add
it
back
something
like
that,
but
anyway,
let's
see
how
that
pr
shapes
up
from.
G
G
Yeah,
so
I
think
it
changed,
it
was
actually
because
of
an
issue.
I
hate
like
at
work
like
in
humanities,
like
when
you
submit,
like
a
crd,
and
you
get
like
the
error
because,
like
the
format
is
not
right,
the
value
was
always
empty.
Like
the
the
bad
value
was
empty,
it
was
an
empty
string
which
is
like
bad
because
it
says
like.
Oh,
this
value
doesn't
doesn't
respect
this
regex,
but
it
doesn't
show
like
the
value.
G
H
H
Oh
no,
I
just
never
joined
this
meeting,
so
I
thought
it
would
be
great
to
drop
in
and
see
what
people
were
talking
about
because
it's
been
on
my
calendar,
but
I've
never
had
time
before.
So
I'm
just
just
spying.
A
Welcome
welcome
thanks
for
joining
yeah
this
this
meeting,
we
just
talked
about
like
things
that
we
need
to
clean
up,
move
around
maintenance
that
kind
of
things
rather
than
any
specific
features
for
any
any
things.
A
So
what
we
were
talking
about
was
like
updating
the
go
dependencies,
because
this
is
a
good
time
in
the
cycle
to
update
dependencies,
because
you
know
code
freezes,
feature,
freeze
is
gonna,
come
poor,
freezer
is
gonna
come
and
then
we
can't
make
any
changes
right
like
so
the
earlier
we
make
all
the
changes
that
we
need
for.
You
know
going
to
a
newer
version
of
dependencies,
the
more
big
time
it's
going
to
have
in
the
ci
jobs,
so
we
can
spot
any
errors
and
mistakes.
A
Those
kinds
of
things,
so
some
of
the
things
happening
here
is
like
we
are
trying
to
do
an
hcd
upgrade,
and
then
we
were
talking
about
like
the
continuity
runs
the
upgrades
and
things
like
that,
so
that
that's
what
we
do
here
in
this
meeting.
H
A
F
A
another
pull
request
coming
in
for
open
telemetry
to
to
modify
the
version
of
grpc
that
we're
using
inside
the
container.
It's
a
I'm
just
saying
I
I
was
like
wait
a
minute.
I
I
thought
I
saw
some
yep
there's
one
so
heads
up
the
there
may
be
more
than
one
reason
you
know
to
be
upgrading
grpc
coming
coming
forward.
F
A
A
Up,
I
didn't
check
what
version
that
we
needed.
No,
we
haven't
merged
that
yet,
as
far
as
I
know
so,
okay,
you,
you-
can
go.
F
A
Right
anyway,
yeah
we
crossed
that
bridge.
So
the
other
thing
that
I
wanted
to
ask
you
mike
was:
was
there
any
larger
changes
that
are
coming
down?
The
fight
from
continuity
from
the
meeting
that
happened
this
morning.
F
Yeah,
well
we're
we're
going
to
put
that
sandbox
change
in
that's
going
to
come.
We
haven't
yet
decided
how
we're
going
to
handle
the
cata
changes
that
are
being
requested
which,
but
I
think
that's
definitely
going
to
affect
kublaid
today
today,
when
kubla
insures
images
it
it
does,
it
does
so
in
in
a
in
the
context
of
the
pod
that
needs
those
images
insured,
but
it
doesn't
pass
to
the
image
services.
F
A
That
doesn't
sound
like
anything
is
going
to
change
in
terms
of
like
there
is.
Continuity
is
going
to
run
as
a
service.
Container
d
is
going
to
have
a
grpc
api
right.
A
Extra
binaries
are
there
extra
things
that
need
to
be
built
or
shipped
hi.
So
in
this
call
we
worry
about
like.
Are
there
extra
binaries
extra
services
to
run
those
kinds
of
things,
so
that
doesn't
sound
like
anything
like
that.
F
Not
yet
okay
yeah,
no!
No!
We
have.
We
haven't,
got
that
far
on
the
confidential
computing
side
and
the
and
then
and
yeah.
Even
we
you'll
see
a
little
bit
less.
F
You
know
so
we're
gonna
go
to
the
sandbox,
so
I'll
refactor
the
crack
code,
so
that
when
you
run
a
pod,
we'll
we'll
actually
be
you
know
be
able
to
use
the
sandbox
code
underneath
the
covers
and
not
have
to
have
a
pause
container.
So
I
guess
from
a
code
organization
perspective,
I
think
we're
going
to
end
up
losing
the
the
pause
containers,
but
I've
been
seeing
some
features
coming
in
and
out
of
that.
So
so
we
need
to
make
sure
that
whatever
is
happening
there
and
so
forth.
F
You
know
a
pause
containers
that
you
can.
You
can
hook
in
that
we're
doing
that
more
at
the
v2
shim
api
layer.
That
or
you
know
we're
at
the
pod.
You
know
the
run
pod
layer
or
update
pod
layer.
You
know
services
and
cry
if
we,
if
we
can
do
it
there,
so
maybe
maybe
try
try
to
start
rejecting
pause
container
changes
might
be
a
good
idea.
Yeah.
I
think
I
think
we're
gonna
go
that
route,
but
we'll
we'll
we'll
go
ahead
and
accept
the
sandbox
changes,
then
we'll
refactor
cry.
F
A
A
I
I
don't
want
to
do
that
yet
so,
if
if
we
want
to
do
that
at
some
point,
I'll
come
up
with
like
a
umbrella
issue
with
a
list
of
checked
items
and
we
can
like
spread
the
you
know,
load
around
and
get
everybody
to
file
some
changes
and
watch
the
ci
jobs
and
see
see
where,
where
we
go
from
there,
okay,
so
well,
one
thing.
D
D
The
ci
jobs
require
secret
external
secret
crd
to
be
running
in
build
cluster,
which
rna
is
working
on,
so
it's
currently
on
iron.
So
once
the
external
secret
cr
is
running
in
the
build
cluster,
then
we'll
be
able
to,
you
know,
create
the
crs
and
then
have
the
ci
job
consuming
it
so
ci
job
pr
is
open.
It's
just
waiting
for
rn
to
have
that
external
secrets
here
in
place.
Okay,.
A
Oh
then,
the
other
question
I
had
for
universities
did
you
run
a
stick
scan
recently.
Did
anything
new
come
out.
D
D
A
Okay,
okay,
so
thanks
a
lot,
everyone
see
you
all
in
two
weeks,
bye,
bye,.