►
From YouTube: Implementation Sync: 2021-05-20
Description
Meeting notes: https://bit.ly/38pal2Z
A
A
I
can
start
I've
been
working
this
week
on
an
issue
for
stackpacks
to
not
write
the
sha
files
as
part
of
the
analyzer.
I've
made
some
progress.
I
hope
to
make
pr
in
the
next
few
days.
That's
it
for
me.
B
Sorry
I
got
no
real
updates.
I
do
know
I
went
to
the
working
group
yesterday.
There
was
you
know
there
was
word
about
a
stat
pack,
issues
that
doesn't
have
a
someone
claimed
it.
Yet
you
know
I
was
just
hoping
to
look
at
it
and
see
if
it's
you
know
something
it's
in
my
power
to
get
involved,
but
you
know
nothing
solid.
Yet.
A
A
Else,
anthony
I'm
sending
you
the
issue
in
the
chat.
A
Cool
anyone
else,
so,
let's
move
to
release
planning,
I
don't.
I
think
that
we
have
anything
special
to
share
we're
all
working
on
stack
packs.
These
are
the
issues
for
phase
one.
D
A
No,
we
don't.
I
mean
we
didn't
decide
on
like
a
specific
time.
We
used
to
ship
life
cycle
every
two
or
three
months.
I
think
three
months,
but
I
have
no
idea
about
the
next
release.
I
didn't
discuss
it.
It's
either
natalie
or
jesse.
D
A
Cool
next
another
agenda
is
nip
discussion.
A
I
signed
the
last
few
weeks,
we're
not
talking
about
phases
two
and
three,
so
we
have
only
this
bug
but
natalie
updated.
Last
week
she
said
that
she
is
not
sure.
Is
there
anything
we
could
do
here
without
changes
upstream?
C
A
B
A
B
C
C
If
you
do
a
docker
login,
all
of
this
just
kind
of
works
out,
like
I
think,
docker
login
will
be
more
permissive,
but
then
the
data
it
puts
in
the
config
file
will
be
the
correct
file
and
then
the
problem
is
in
systems
like
tecton,
where
people
are
entering
these
things
directly
and
you
don't
have
the
docker
login
workflow
to
clean
things
up
for
you.
B
Okay,
sorry,
you
know
this,
you
know
you're
triggering
my
memory
here
right,
so
I
want
to
reflect
the
perspective
right.
You
know
this
is
life
cycle
code
which
is
supposed
to
be
agnostic
to
all
these
registries
right.
So,
if
there's
a
condition
statement
that
says
hey
if
docker
index.docker
and
then
put
trailing
slash
here,
it's
just
you
know
would
that
be
appropriate
for
the
lifecycle.
I
think
that
was.
I
think
that
was
the
way
the
argument
was.
A
C
The
one
thing
that
is
docker
specific
about
the
way
it
works
is,
if
you
don't
provide
a
registry
we
default
to
docker
right,
so
we're
already
sort
of
taking
a
like
a
naked
repository
reference
without
a
registry
and
adding
inferring
that
it's
docker,
I'm
sorry
using
gvcr.
To
do
that,
like
we
didn't
write
that
code,
but
it's
like
a
standard
and
that
everyone
expects
no
registry
to
be
docker
so
be
because
that
special
case
is
already
sort
of
built
into
the
ecosystem.
C
C
B
A
A
So
since
nelly
investigated
it,
maybe
we
can
wait
for
her
just
to
make
sure
we
follow
like
what
she
is
already
investigated,
that
make
sense
cool,
so
we'll
leave
it
for
now,
and
I
think
we
don't
have
any
other
issues
to
discuss
this
week
unless
anyone
thinks,
by
the
way.
A
All
right,
so
the
next
thing
is
the
rfc.
So
I
know
that
we're
going
to
discuss
this
right,
sam
you're
here
with
us.
Yes,
okay,
okay,
you
were
very
quiet
today.
E
I
also
wanted
to
figure
out
if
anyone
from
the
implementation
team
wants
to
take
over
this,
because
I
think
at
this
point
it
is
beyond
like
what
I
originally
wanted
to
do
and
we
are
talking
about
a
larger
restructure.
So
I
think
it
would
be
better
for
someone
who
knows
more
details
to
take
over
this
rfc.
C
A
C
A
C
Nothing
new,
since
yesterday
I
have
finished
on
my
workstation.
I
will
get
something
up
before
the
end
of
the
week.
It
might
not
be
finished.
It
might
just
be
a
draft
depending
on
how
time
management
goes,
but
I'll
get
something
that
folks
can
look
at.
C
A
C
A
C
A
Okay,
so
I'll
I'll
think
the
three
of
them
about
this
rc
and
we'll
see.
I
have
a
feeling
that
jesse
will
comment
or
approve
it.
Then
we
can
prefer
work
okay,
great.
So
we
have
this
rfc
and
I
saw
something
else
on.
A
Oh
thank
you
dan.
I
forgot.
Maybe
I
should
add,
like
a
question
to
the
agenda.
I'll,
do
this
so
we'll
decide
who
is
on
samurai.
A
Before
beginning
of
the
meeting,
there
is
also
this:
what
do
we
want
to
discuss?
First,
who
added
this
by
the
way.
A
Okay,
so
maybe
we
can
take
a
look
at
the
benny.
Are
you
okay
with
this
great?
Maybe
we
can
take
a
look
at
what
jesse
added.
C
C
Api
versions
and
stuff,
like
that,
and
I
think,
moving
build
packs,
is
a
big
problem
because
it's
sort
of
built
into
the
packaging
and
the
distribution
spec
right,
like
our
apis,
are
not
really
independently
changeable
in
the
way
that
I
wish
they
were
for
right
now.
You
know
all
build
packs
are
packaged
under
cnb,
build
packs
right.
C
C
A
Can
we
do
something
similar
to
what
we're
doing
under
layers
like
have
a
separate
directory
for
build
packs
for
all
build
plaques
like
it's
regular
path,
or
will
it
be
a
problem
with
the
other
build
pack
that
don't
need
to
look
like?
A
C
Know
I
guess
the
the
problem
is
that
that
cmd
directory
is
typically
going
to
be
a
volume,
so
you
can't
in
builder,
like
you're,
going
to
mount
over
it.
So
you
can't,
in
the
builder,
then
put
things
there,
because
they'll
get
mounted
over
when
you
run
so
build
packs
at
cnb.
Build
packs
work
for
that
reason,
but
there's
there's
some
options
that
I
think
we
need
to
think
through.
C
I
guess
it
can't
be
the
platform's
creating
they'll
have
to
you
know.
The
life
cycle
could
like
set
up
some
sim
links
or
something
depending
on
what
it
finds
make
it
all
look
the
same
like
I
have
not
not
thought
through
all
the
details
here.
I've
only
thought
through
the
fact
that
it's
a
problem,
we
should
probably
think
through
the
details.
C
Rc,
I
don't
have
time
to
do
that
before
the
end
of
this
week.
I'm
like
the
thing
that
I
would
be
putting
out.
I
mean
I
can
add
a
comment
that
it's
a
problem,
but
I
can't
edit
comment
with
a
button.
A
E
C
I
think
it
would
be
good
to
keep
it
independent
because
we're
gonna
want
to
if
it's
a
build
pack
feature
we're
going
to
need
to
work
with
the
older
platform
apis
anyway,
as
well.
To
think
about
where
it
goes.
In
the
existing
structure,
we
could
leave
the
where
it
goes
in
the
new
structure
as
an
exercise
for
this
rfc,
which
is
ballooning
into
a
very
necessary
but
complicated
reorganization,
yeah.
E
Cool,
so
do
you
have
any
preferences
there,
or
should
I
just
shove
it
under
a
new
volume?
That's
called
slash,
shared
layers
or
something
like
that.
C
It
makes
sense
we
got
the
top
level
layers
directory
right
under
there.
We
have
build
packs,
we
also
have
and
we
have
config.
I
think
we
can
make
a
a
shared
layers
and
then
it
sort
of
stays
in
the
same
place
in
the
hierarchy
when
you
move
other
things
around,
but
it
makes
more
sense
in
the
new
hierarchy.
E
D
A
quick
question
ping,
the
other
maintainers
jesse
ben
and
steven
to
take
a
look
at
the
documentation
needs
before
release.
Does
anyone
want
to.
D
A
So
thanks
don,
we
have
four
minutes.
Then
do
you
think
we
can
talk
about
this
previous
image?
An
extra.
D
Let's
try
okay,
so
I
guess
I
like
share
my
screen.
C
That
makes
sense,
I
don't
think
anything
would
break
if
we
took
this
check
out.
I
think
we
maybe
shouldn't
have
put
it
in
in
the
first
place.
The
reason
we
really
encourage
people
to
do
things
in
the
same
registry
is
because
it
is
ten
thousand
and
a
half
times
faster,
because
otherwise
you
have
to
copy
all
the
blobs
from
one
registry
to
another
and
when
we
didn't
put
checks
like
this
in
people
would
do
these
things
and
complain
that
cmb
was
really
slow.