►
From YouTube: Working Group: August 22nd, 2023
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
A
The
outstanding
rfcs,
like
I,
said
not
a
lot
has
changed
here,
so
this
will
probably
be
pretty
quick
start
with
the
couple
dependencies.
Rfc
I
still
haven't,
had
any
opportunity
to
work
on
this
I've
been
getting
bogged
down
with
other
work
I.
My
hope
is
to
have
some
time
later
this
week
to
dive
in
and
try
and
get
something
useful
out
of
this
and
try
and
get
this
back
on
track
and
back
into
discussion.
A
This
is
the
proposal
for
published
multi-arch
build
packs
is
still
a
draft
Jericho.
Do
you
have
any
updates
that
you'd
like
to
give
on
this?
Are
we
good
all
right
perfect,
and
then
we
have
the
proposal
to
introduce,
introduce
new
growl
VM.
The
last
update
that
we
heard
about
this,
oh,
was
from
Anthony
last
week
and
he
was
saying
that
there
was
some
discussion
happening
within
Oracle.
A
It
looks
like
there
is
an
update
here,
so
anyone
who's
invested
in
that
please
feel
free
to
take
a
look.
Do
Anthony.
Do
you
have
any
updates
you
want
to
give
on
that.
C
Yeah,
it's
a
small
one
though,
but
yeah
you're
right.
There
have
been
some
activity
from
Fabio
from
Morocco,
but
from
a
legal
standpoint.
C
C
Basically,
so
I
contacted
Ram
from
the
cloud
Foundry
Foundation,
you
know
the
developer
Advocate
and
he
told
me
that
he
was
going
to
reach
out
to
some
people
from
what
did
you
tell
me,
Linux
Foundation
yeah,
he
told
me
about
Linux
Foundation
was
handling
a
league
of
stuff
for
cloud
Foundry,
so
I
didn't
get
any
news
from
me
from
him
since
then
I
think
it
was
yesterday
morning.
C
Actually
so,
but
yeah,
let's
kind
of
blocked
with
I
mean
talking
about
you
know
merging
anything,
because
maybe
you
have
seen
that
there
is
a
alternative
from
Daniel,
as
you
can
see
other
record
VM,
so
this
yep
yep,
exactly
you
have
this
one.
So
I
was
just
saying:
oh,
it
was
from
this
discussion.
If
this
thing
could
be
merged
as
is
and
right
now
yeah.
We
just
want
to
make
sure
that
that
everything
is
fine
from
a
legal
standpoint,
so.
B
A
Jan
I
just
saw
that
you
joined
I,
basically
pretty
quickly
glossed
over.
This
do
a
couple
dependencies
because
I
have
not
really
had
any
time
to
work
on
it.
As
of
recent,
do
you
have
any
updates
or
anything
that
you'd
like
to
talk
about
on
that.
D
A
A
Awesome
in
that
case
then
c,
b,
updates
and
questions
I
mentioned
this
last
week.
I
might
have
the
only
relevant
update
that
I
can
think
of
from
the
CMB
front.
Is
that
Pac
30
0.30
came
out?
A
It
has
official
support
for
a
couple
of
platform
life
cycles
that
the
previous
version
didn't
have.
So
if
you
were
trying
to
do
things
like
embed
like
run
time
environment
variables
into
the
Builder
by
like
mounting
a
there,
there's
there's
a
process
but
effectively
if
you're,
making
a
builder
and
you've
added
the
new
sort
of
layer
that
allows
you
to
have
sort
of
builder-specific,
runtime,
params
I,
believe
that
will
now
work
on
pack.
A
Although
I
haven't
used
that
feature
yet
it
also
has
that
with
the
I
guess
the
lifecycle
supports
kind
of
like
independent
from
that,
but
flatten
is
now
fully
out
as
well
as
part
of
this,
which
is
something
that
could
be
interesting
to
look
into.
A
Thank
you
all
of
the
default
or
not
default.
The
suggested
Builders
yeah.
C
A
Updated
to
include
all
of
the
new
paghetto
Jammy
Builders,
including
I,
believe
the
static
Builder.
If
I
recall
correctly,.
D
A
D
Was
an
interesting
for
for
the
for
the
Ubi?
Oh
yes,
yeah.
B
Definitely
we
there
were
some
fixes
and
some
stuff
in
there
that
that
helped
us
out
for
sure
and
that,
like
the
Builder,
our
our
initial
Builder
was
updated
to
use
30
and
17,
which
was
good.
A
On
the
topic
of
extensions,
this
is
kind
of
almost
a
sidebar,
but
the
the
a
team
from
CNB,
mostly
around
Heroku,
came
to
our
working
group
a
couple
weeks
ago
and
they
were
having
that
they
had
a
lot
of
questions
about
how
we
were
using
extensions,
especially
when
it
came
to
like
what
extensions
like
imply
or
the
implications
of
extensions
on
the
ability
to
rebase
I
informed
them.
A
That
I
was
not
the
most
intimate
when
it
came
to
using
extensions
in
general,
because
I
haven't
been
pursuing
any
of
that
work.
I
did
drop
your
name
Michael
at
one
point,
but
if
you
have
any
feedback
or
opinions
on
extensions,
the
I
think
that
it's
very
much
in
a
state
where
it
was
like
plastic
and
very
general,
and
they
aren't
necessarily
opposed
to
moving
it
to
be
a
bit
more
Concrete
in
terms
of
some
of
its
definitions
and
features.
A
So
if
that's
something
you're
interested
in
they've
been
talking
about,
you
know
removing
stacks
and
extensions.
In
the
last
four
or
five
c
b
working
group
meetings.
B
Okay,
yeah
I,
think
I
checked
with
I,
couldn't
make
the
last
one,
but
I
did
check
with
Ozzy
who
said
he
was
going
to
make
that
and
like
in
terms
of
the
rebasing
I.
Think
the
the
original
support
for
extensions,
because
the
like
the
Run
image
was
simply
a
like
you.
You
couldn't
provide
anything
in
your
darker
file
so
that
sported
actually
having
rebases
be
done,
I
think
there's
some
new
functionality.
That's
that's
being
worked
on
that
that
I
hear
through
Aussie.
That
would
allow
you
to
to
extend
that.
C
A
Their
biggest
fear
is
that,
like
extensions
are
used
at
such
a
high
level
of
proliferation,
that
effectively
makes
rebasing
impossible,
and
so,
like
you
know,
rebasing
is
a
big
I,
guess
I
guess
it
is
a
big
feature
of
cnbs
and
so
like.
If,
if,
if
there's
like
I,
guess
a
feature
that
they've
introduced
that
hasn't
been
defined
well
enough
to
cause
like
a
huge
impact
on
rebasing,
they
would
like
to
understand
that
more.
B
A
D
A
B
No,
that's
good
and
I
know.
I
have
a
few
things
I
need
to
get
to
in
terms
of
some
of
the
other
projects,
like
there's
at
least
one
that
was
being
updated.
That
I
think
we
agreed
I
should
just
turn
off
and
a
bit
a
bit
of
sort
of
housekeeping
stuff
like
that
that
we
need
to
do
and
then
we'll
start
to
get
back
to
some
of
the
the
other
work.
A
Yeah
but
but
hopefully,
I
will
be
able
to
at
least
get
a
one-off
release
out
so
that
we
can
continue
to
use
that
yep.
A
No
problem
all
right,
it
doesn't
sound
like
there's
any
other
project
updates,
so
we
can
move
to
open
mic
discussion
topics.
A
E
That's
fine,
so
I'll
just
give
the
context
so
I'm
working
on
creating
multi-arch
stacks
and
it's
failing
to
push
multi-arch
stack
to
PCR
and
I,
like
I
need
to
do
a
little
bit
more
testing,
but
it
sounds
like
that's.
E
The
problem
is
that
this
is
some
of
the
features
in
the
new
Google
artifact
registry
aren't
supported
in
GCR,
so
I
don't
know
that
we're
going
to
be
able
to
get
around
this
so
after
I
do
my
testing
I
just
wanted
to
bring
this
up
here,
just
kind
of
the
push
to
pkg.dav
Dot,
whatever
it's
like
the
the
new
version
of
Google
container
registry.
F
E
I
mean
I
feel
like
it's
kind
of
a
can
of
arms
and
I
just
wanted
to
put
it
out
there
so
that
you
guys
are
aware
and
yeah
I
I'm
literally
doing
some
testing
after
this
call
just
to
see
if
that
solves
the
problem.
But
if
it
does,
then
somebody
will
have
to
address
this.
A
So
I
guess
I.
The
only
question
I
have
around
this
is
like.
Do
we
want
to
continue
to
push
to
a
registry
that
is
hosted
by
Google
I
know
that
there's
advantages
to
us
having
like
like
proxy
Registries
to
reference,
because
for
some
people
it
makes
downloading
more
acceptable.
I,
don't
know,
that's
the
word
I
wanted,
but
it's
the
word
that
came
out
of
my
head.
A
C
Yeah,
so
just
just
to
make
sure
Jacob,
you
said
that
Google
I
mean
gcr.io
does
not
support
images.
C
Okay,
well,
okay!
Well,
that
sucks,
okay!
Well,
this
is
bad
I
believe
because
you
know
this
is
definitely
where
we
want
to
go.
Also.
C
On
the
other
hand,
what
I
wanted
to
mention
duplicating
gcr.io
is
that
we
got
to
Docker
herbs.
I
mean
it's
not
I.
Definitely
didn't
do
a
scientific
study,
but
people
around
me,
expertise,
friends
and
so
on.
They
regularly
tell
me:
oh
I,
couldn't
download
packet
of
build
packs
and
then
I'd
say:
oh
yeah,
did
you
download
from
your
own
IP,
or
did
you
go
through
a
CI
machine?
Oh
yeah,
you're
right,
it's
working
for
my
machine,
but
not
from
VCI
umbrella
anyways.
C
You
know
that
Docker
Hub
has
been
applying
IP
rate,
limiting
basically
I
mean
download
limiting
per
IP
and
it
kind
of
sucks,
because
I
mean
when
we're
on
GCR
I
guess
nobody
complained
about
that
right.
So,
basically,
we
got
ourselves
in
in
a
situation
where
yeah
sure
Docker
hobby
is,
you
know
the
place
to
be
somehow.
C
But
this
rate,
limiting
a
thing
is
I
believe
is-
is
buffering
many
of
our
users
once
again,
I
didn't
do
a
full
Studio,
but
from
one
from
what
I've
heard
so
far,
that
I
mean
it
can
create
a
friction
and
oftentimes.
What
I
would
reply
to
was
friends
is
well,
you
know
what
we're
also
published
to
gcr.ios
just
so.
C
E
Yeah
I
mean
I
I,
think
it's
possible
that
we
can
just
push
to
the
pkg.dev
URL
and
start
using
I.
Don't
know
this
yet
I
need
to
do
some
testing
that
it
actually
solves
the
problem,
but
I
think
it
will
because
I
found
other
I
found
like
discussions
in
other
people's
repos,
where
they
hit
the
same
error.
That
I
saw
and
it
seems
to
be
the
solution.
E
So
I'm
not
saying
stop
pushing
to
Google
container
registry.
You
can
kind
of
push
whatever
you
want.
I
do
think
it's
probably
similar
to
have
more
than
one
place
but
yeah.
Let
me
kind
of
like
finish
the
investigation
or
whatever,
but
once
I
do
then
you'll
have
the
decision
to
make
on
how
you
guys
are
moving
forward.
So.
A
Yeah
I
think
I
think
just
to
be
clear:
I'm
not
encouraging
us
to
abandon
a
Google
hosted
registry
I,
just
figured
I
would
throw
it
out
there
like.
If
we're
gonna
have
to
do
some
form
of
transition.
Maybe
that's
not
the
transition
we
want
to
make,
but
if
it's
yeah
I
think
it's
totally
fair
for
us
to.
If
it's
just
as
easy
as
pointing
to
a
different
like
registry
like
URL,
basically,
then
I
I
don't
know
that
I
would
be
opposed
to
continue
just
kind
of
keeping
things
the
same.
A
E
Only
challenge
I
have
is
like
your
your
registry,
your
repo
exists,
and
it's
clearly
been
there
for
a
while,
so
like
I'm
gonna
go
and
create
a
new
one
and
push
to
it
and
I'll
clearly
be
using
the
new
artifact
registry.
So
I,
don't
I,
don't
have
a
way
to
simulate
the
transition,
the
migration
from
like
GCR
to
the
new
one,
so
that
that's
the
part
that
I'm
not
sure
I
can
really
vet
before
you
move
forward.
So
somebody
might
have
to
help
with
that.
F
So
yeah
got
it
yeah
I
think
just
a
plus
one
will
Anthony
and
Farah
said,
like
I,
definitely
get
feedback
from
folks
inside
of
VMware
that
rate
limit,
but
it's
so
bad
that
you
can't
even
set
up
a
C8
and
Mirror
to
your
own
like
internal
registry.
So
it's
like
you,
don't
even
have
the
ability
to
like
get
yourself
out
of
the
whole
of
Docker
Hub,
so
I
yeah
I
think
it
I
think.
F
If
we
were
to
to
like
only
use
Docker
Hub,
we
would
very
rapidly
get
a
lot
of
feedback
that
it's
unusable.
So
again,
it's
not
a
scientific
study,
but
I
do
have
probably
like
order
of
10
examples
of
people
telling
me
that
that
was
a
problem
for
Stacks
in
particular
and
Builders.
The
other
thing
I
was
going
to
say
and
I
think
really
for
us.
This
is
kind
of
on
you
right,
like
as
a
steering
Community
person
is
I.
F
Think
you
are
the
only
the
steering
committee,
the
only
people
who
have
the
ability
to
actually
try
out
this
migration
path
right,
like
I,
don't
think
the
rest
of
us
have
people,
let's
just
go
muck
around
with
our
GCI
DCR
accounts,
so
separately
from
Jericho's
investigation
of
pushing
into
a
new
artifact
registry.
It
sounds
like
if
we
want
to
maintain
any
support
or
backwards
compatibility
with
a
GCR.
I
o
hostname
that
we're
going
to
have
to
do
some
exploration
and
migration
there
anyway
right,
even
if
you
weren't
doing
multi-arch
stuff.
F
This
has
come
up
because
multi-art
is
like
a
forcing
function,
but
even
if
we
didn't
have
that
we're
still
going
to
either
have
to
decide
to
migrate
or
drop
it.
If
we're
not
going
to
drop
it,
we
have
to
migrate.
If
we
have
to
migrate,
I
think
the
steering
committee
person
has
to
like
Spike
on
this
right.
D
For
what
it's
worth
I
would
probably
continue
to
to
double
publish
stuff
until
Google
actually
removes
and
makes
the
gcrio
links
dysfunctional,
which
is
like
where
then,
then,
that's
just
going
to
happen
right
and
for
what
it's
worth.
We
even
have
gcrao
links
on
a
lot
of
build
pack
a
remix,
so
we
are
still
kind
of
advertising.
This
I've
just
quickly
looked
up.
We
actually
consumed
from
gcirio
at
the
sap
site
to
get
the
the
component
built
packs.
Well,
it's
it's
a
it's
even
more
funny.
D
D
Definitely
kind
of
keep
this
yeah
definitely.
F
Think
like
it's
prudent
to
do
that
sooner
rather
than
later,
it
sounds
like
we
have
about
nine
months.
Maybe
a
bit
less
right.
I
said
like
May
2024
was
when
they're
deprecating
or
sorry
removing
GCR.
So
what
is
it?
August
like
it's
like
eight
months,
so
it
sounds
like
we
should
do
that
sooner
rather
than
later,.
D
F
E
One
follow-up
question
or
the
stacks
release
that
failed
because
of
this
I
guess:
I
just
want
to
make
sure
I'm
helpful
if,
like
I'm
gonna,
do
this
test
it
and
then
allow
the
steering
Community
to
decide
they
want
to
move
forward.
But
then,
after
that,
do
you
plan
to
not
published
to
DCR
and
published
a
Docker
Hub
or
just
not?
Do
the
release
for
now,
like
I,
just
want
to
make
sure
I
I'm
helping
and
getting
staying
out
of
the
way?
And
not
bothering
you
guys
we're
going
to
wait.
F
We
continue
to
be
published
to
GCR
like
that
sucks.
If
that's
the
case,
that's
annoying,
but
I
think
that's
the
commitment
we've
made
right.
The
alternative
is
like
we
start
publishing
DCR
with
no
notice,
not
gonna
work
for
the
reasons
we
mentioned
before,
or
we
just
abandoned
multi-arch
until
we
have
a
better
plan
which
doesn't
really
start
say
your
use
case
and
kind
of
takes
us
back
to
where
we
were
three
months
ago.
So
I
think
it
unfortunately
sounds
like
we
have
to
go
down
that
path
of
kind
of
both
right.
E
Yeah,
actually
there
is
a
potential
work
around
I
suspect
it
would
support
Docker
manifests
so
you
could.
You
could
have
like
the
script
that
publishes
the
stuff
like
just
published
a
Docker
Hub
and
then
have
another
little
function
that
will
convert
the
oci
image
to
a
Docker
manifest
and
then
push
that
I
mean
like
you
know,
I
know
how
to
do
that.
If
you
guys
want
to
do
something
like
that,
you
know,
there's
there's
some
options,
but
we
don't
need
to
figure
it
out
on
this
call.
A
I
think
that
will
conclude
this
working
group
meeting
then.