►
From YouTube: CNB Sub-Team Sync: Learning - 4 Feb 2022
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).
B
C
C
B
Yeah,
I'm
gabe
from
bloomberg.
I
work
with
zambov
trying
to
increase
our
build
back,
presents
and
happy
to
be
here.
C
Awesome
happy
to
have
you
all
right.
The
next
thing
is:
do
we
have
any
still
issues?
We
actually
have
quite
a
bit.
C
So
if
it's
okay
with
y'all
sam
and
I
will
basically
kind
of
use
this
to
poke
ourselves
to
get
some
of
this
stuff
done
and
yeah,
we
could
kind
of
skip
this
for
today.
That's
all
right!
C
C
Opposed
I
mean
you,
could
you
could
watch
us
run
through
these?
The
only
other
pr
stealth.
Pr
is
a
draft,
so
that
is
good
for
now.
C
Is
unlabeled
issues
which
we
have
so
we
are
good
there.
C
Feedback
for
boot
pack
concepts,
but
I
think
sam
you
wanted
to
present
some
of
that.
And
then
I
have
an
update
on
the
pac
website
from
our
last
meeting
nice
I'll
I'll,
throw
mine
out
there
just
real
quick
to
get
some
ideas
and
then
really
just
relay
the
information.
I
did
kind
of
view
out
a
tiny
proposal
to
jail
to
kind
of
get
a
temperature
feel
for
whether
or
not
that's
something
that
we
wanted
to
do,
and
he
he's
not
opposed
to
the
idea.
C
I
think
he
that
the
reception
was
positive.
C
The
only
real
kind
of
point
of
contention
is
sort
of
like
the
capacity
and
priority
of
such
thing
right,
so
it
seems
relatively
low,
but
it
is
still
something
that
we
would
have
to
do.
We
would
have
to
buy
the
domain
figure
out
the
hosting
and
then
actually
get
the
design
set
up
and
and
put
everything
there
and
move
stuff
out
of
build
packs,
io
update
references
and
pointers
and
so
yeah.
C
So
that
was
something
where
it
just
seemed
like
it's
nice
to
have,
and
not
so
much
a
high
priority
compared
to
other
things
that
we
have
right
and
especially,
I
think
because
the
people
in
the
learning
team
are
kind
of
the
same
people
that
are
driving
other
changes
that
are
probably
higher
priority.
C
That
was
kind
of
my
thing,
and
I
do
agree
with
that.
So
I
didn't
really
have
a
a
way
to
say
like,
but
we
have
this
individual
here.
That
would
be
more
than
happy
to
and
has
all
the
free
time
in
the
world
to
get
this
done
right
so
yeah,
I
don't
know
if
there's
any
thoughts
on
that,
I
did
see
the
sub
domain
right.
If
you
want
to
do
x,
I
o.
A
At
least
like
at
least
getting
something
like
build
packs,
I
o
slash
back,
would
be
super
easy
as
a
separate
website
like
you
just
create
a
separate
talk,
set
or
whatever
publish
it
there
I
don't
know
yeah,
but
even
getting
a
sub
domain
might
be
trivial.
I
think.
C
Yeah
I
know
steven
owns
the
domain,
so
I
think
that
would
be
super
trivial,
the
other
parts
of
the
design,
the
building
and
the
migration.
I
think
those
are
the
heavy
loads
right,
so
I
don't
know
how
to
tackle
this.
Exactly
at
this
point,
I
think
maybe
I'd
like
to
get
a
more
feedback
or
feelers
on
how
much
we
need
this
and
and
what
the
real
benefits
are.
So
I
might
even
just
have
to
think
about
it
myself.
C
There
are
other
avenues
like
I
know
the
the
mentorship
programs
they
for
web
related
stuff,
there's
there's
quite
a
a
a
few
candidates
there.
So
I
think
they'd
be
more
than
happy
to
take
that
on.
It's
just
marrying
it
with
the
design
aspect
of
it
that
might
be
challenging.
A
Lfx
ended,
but
they
also
sent
out
proposals
for
json,
because
I
know
I
think
cncf
might
have
gotten
there
or
just
about
to
get
their
proposals
in
line.
C
C
No,
the
features
yeah,
I
guess
that's
another
one.
The
state
of
it
is
there
were
a
few.
C
I
would
say
adaptive
changes
necessary,
but
very,
very
minor.
At
this
point
I
guess
you
would
say
like
it
just
needs
somebody
some
one
of
the
maintainers
to
go
in
there
and
polish
it
up
and
publish
it.
I
don't
know
that
there's
sufficient
stuff
in
there
to
make
it
a
project
for
the
mentorship
program.
C
C
All
right
yep
go
ahead.
A
So
just
some
feedback
on
concepts.
I
guess
so
we've
it's
like
seeing
people
who
are
not
app
developers
like
now
trying
to
like
understand
a
bunch
of
buildback
concepts
so
like
from
the
perspective
of
the
app
developer,
they're
very
shielded
right
like
they're.
They
they
just
run
this
command
set
some
arguments.
You'll
have
an
app
at
the
end
of
it.
That's
it,
but
as
someone
who
wants
to
dive
deeper
and
maybe
create
a
buildback.
A
A
What's
like
just
just
people,
I've
seen
people
call
every
one
of
these
things
buildbacks,
because
that's
like
the
project
names
of
the
like
okay,
this.
This
is
what
it's
doing
like
I've,
also
seen
some
people,
some
arm
developers
call
the
output
image
build
packs
like
I
want
to
create
my
own
build
pack,
and
then
they
actually
mean
they
just
want
to
create
an
image
using
those
packs.
A
So
like,
as
anyone
knew
coming
to
the
project,
the
first
point
of
confusion
they
have
is
like
the
overload
of
the
word
or
term
build
it's
like
everywhere,
and
I
think
as
someone
new,
it
can
get
confusing
like
after
a
while.
It's
fine,
you
know
the
differences,
but
as
someone
new,
they
get
extremely
confused
as
to
what
one
thing
means
over
the
other
and
the
using
things
interchangeably
and
then,
when
they're
trying
to
follow
the
documentation,
they
just
get
confused.
A
C
And
and
we're
talking
specifically
from
a
build
pack
author's
point
of
view
right
so
like
neo,
built
by
cultures,
so
kind
of
thinking
about
it,
I'm
I'm
I'm
tasked
with,
or
I'm
researching,
the
idea
of
using
build
packs
to
provide
build
packs
for
my
either
enterprise
or
for
a
language,
or
you
know
some
tooling,
and
I
say:
okay,
let
me
look
at
what
buildpacks
project
is
right,
and
so
they
go
there
and
you're
saying
okay.
C
So
now
I
have
to
learn
about
the
ecosystem
I
like
to
so
that,
like
all
the
components
within
there
and
then
say,
okay
now,
I
maybe
hopefully
get
it
not
to
dive
into
a
build
pack
domain
right
and
so
now,
inside
of
the
build
pack
domain.
There's
these
other,
like
micro
concepts
right
that
you
have
to
also
know
you
have
to
know
when
your
stuff
is
executed
in
the
entirety
of
the
life
cycle
and
what
sort
of
documents
inputs
and
outputs
are
necessary
for
the
build
pack
itself
right.
A
Things
like
how
meta
build
packs
work
or
like
the
the
current
buildback
documentation,
is
very
focused
on
one
buildback,
acting
alone,
which
is
rarely
the
case
right
I
mean
it
might
have
been
the
case
when,
like
it
started
out
with
heroku,
just
detecting
one
build
back
and
running
it,
but,
like
I
don't
think,
that's
the
case
anymore.
A
So,
like
things
like
how
detect
works
like
how
how
buildbacks
in
the
group
are
detected
rather
than
a
single
billback
being
detected
like
art,
I've
seen
causes
a
lot
of
confusion.
How
requirements
and
provisions
are
matched
up?
What
role
do
they
actually
play
into
the
build
process?
A
How
were
the
requirements
actually
fed
so
that
the
build
binary
actually
resolves
it
and
gets
the
correct
whatever
information
out
of
it?
I
think
all
of
that
stuff.
We
don't
talk
about
anywhere,
it's
just
in
the
spec
as
abstract
concept.
So,
unless
you've
looked
at
one
of
the
existing
production
buildbacks,
you
would
have
zero
idea
what
to
do
with
any
of
that.
C
Yeah,
so
I
guess
I'm
trying
to
understand
is
this
something
that
is
a
higher
level
issue,
with
the
way
that
the
the
documents
and
the
websites
are
structured,
that
something
at
a
high
level
could
help.
Or
is
it
just
really
simply
the
need
for
the
build
pack
author's
guide
to
have
more
content
provided
by
the
buildpack
team
sub
team.
A
I
think
both
it's
both
content,
as
well
as
us
as
a
project
recognizing
the
fact
that
a
lot
of
the
terms
we
use
are
overloaded
and
they
can
be
very
confusing
to
someone
new.
We
should
recognize
like
if
we
don't
want
to
change
those
terms.
We
should
recognize
the
confusion
and
address
it
for
a
new
person
coming
in
to
the
documentation.
A
Or
like
just
show
with
examples
like
hey
here's,
what
is
different
about
buildbacks
and
builders
and,
like
I
think
the
presenters
are
currently
presented
as
abstract
concepts.
So,
like
people
will
understand
it
when
they
read
like
half
an
hour,
one
or
later
they'll
be
completely
confused
again
like
what
what
was
one
thing
versus
the
other,
if
they
don't
have
something
to
attach
it
to.
C
C
So
if
I
were
to
go
here,
sorry,
not
github,
sorry
doc's,
homepage,
here
right
and
then
I
go
into
let's
say
concepts.
C
A
I
think
the
other
thing
that
confuses
everyone
when
coming
from
docker
is
like
everything
is
stored
as
an
oci
image
like
they
think
everything.
That's
an
oci
artifact
is
a
docker
image
so
like
just
build
packs
being
stored
in
the
registry
or
like
the
build
image.
The
builder
a
container
image.
Creating
a
container
is
a
very
mind-boggling
concept
for
people.
A
A
A
And
then,
like
the
other
thing
that
confuses
them
is
like
you
know
how
it
says:
build
backs
on
the
right
hand,
side
like
built
by
provided
layers
like
yeah
yeah
yeah,
so
like
to
a
new
person
like
they're,
like
oh,
what
what's
happening,
it's
like
built
by
copying
or
some
of
its
stuff
there
like
is
that
itself
a
well
packed
in
the
output
image,
because,
like
there's,
some
runtime
logic
that
also
gets
executed
right,
they're
like
oh,
something,
is
happening
during
the
wrong
time
as
well.
That's
magical!
C
Yeah,
I
definitely
get
all
the
concerns.
I
don't
know
if
I'm
personally
qualified
to
solve
it.
That's
for
sure
I
think
that's
kind
of
what
I'm
coming
to
the
realization
right.
It's
like
when
there
was
nothing.
I
could
definitely
provide
something,
but
now
that
there's
something
and
people
are
saying
like
well,
it
needs
to
be
improved.
It's
like
okay!
Well,
I
don't
know
how
to
do
that
right.
So
I
guess
what
are
some
action
items
that
you
can
think
of
for
improving
them.
A
I
remember
the
last
time
we
dismissed
it
because
it
had
like
some
prerequisites
around
like
it
was
just
a
long
form
or
something.
I
don't
remember.
Oh
what
sorry
like
you
know
how
they
they
they
had.
You
fill
a
survey
about
asking
a
bunch
of
things
before
you
could
before
they
could
find
you
the
appropriate
person
to
write
the
talks
or
something.
A
C
With
you
with
you,
like,
as
I
was
going
through
this
like
what
do
you
mean
it's
like
right,
there
yeah
exactly,
as
you
explain,
I'm
like
okay,
I
could
definitely
see
where
or
people
could
be.
You
know,
hung
up
the
idea
of
a
container.
You
know
image
creating
another
container
image.
That's
pretty
yeah
crazy.
A
So
I
mean
it's
also
like,
while
teaching
this
stuff
to
people
you
realize.
Oh,
I
I
never
thought
that
people
could
like
take
this
concept
twist
it
and
then
try
to
comprehend
it
in
like
this
entirely
different
fashion
and-
and
I
can
follow
their
line
of
reasoning-
I
can
completely
follow
what
thought
process
they
would
have
gone
through
to
reach
that
conclusion.
B
Yeah,
I
think
defragging
the
concepts
would
be
useful
to
a
lot
of
people
like
it
build.
The
word.
Build
itself
is
just
every
for
every
concept
in
build
text.
There
is
a
component
called
build
somewhere
right
and
I
think
that's
that's
a
confusing
thing
for
a
lot
of
people
right
like
the
fact
that
the
build
stack
is
not
the
thing
that
is
built,
but
the
thing
that
builds
and
the
build
pack
is
different
than
a
builder,
which
is
different
than
a
build
plan,
which
is
it's
a
lot
of
build
words
in
there.
C
But
but
yeah,
I
guess
I'm
trying
to
understand
or
trying
to
think
of
what
a
good
alternative
would
be,
this
being
a
build
system
right
there
you
go
build
right,
it
seems
almost
natural
to
think
about
it
in
that
way,
at
those
different
scopes
to
me,
I
think
a
part
is
is
maybe
not
having
a
clear
sort
of
like
scope
of
conversation
right.
C
It's
like
when
we're
talking
about
it
because,
like
app
developers
right,
they
shouldn't
care
all
too
much
about
what
a
build
pack
is
or
what
a
builder
is
or
like
the
life
cycle
stages,
and
all
this
crap
right,
like
they
really
only
care
about
the
high
level
order,
which
is
I've,
got
build
packs
and
they
do
stuff
right.
That's
it.
I
mean.
A
C
C
C
Well,
right
now
for
back
stuff
related.
Yes
right
like
they
need
to
know,
what's
available
to
them
for
execution.
C
A
Yeah,
I
think,
for
each
person
you
should
the
first
thing
they
should
come
to
is
like
a
problem
they
want
to
solve,
like
as
an
app
developer
if
you're
coming
to
buildbacks.
What
do
you
want
to?
Do?
You
want
to
create
production,
container
images?
The
first
thing
you
want
to
see
is
here's
an
example
of
a
demo
repository
run
back,
build
nothing
else.
You
get
an
output
image,
run
docker,
docker,
run
and
boom.
A
You
have
something
that's
running,
that's
that's
my
goal
as
an
app
developer
to
get
from
my
source
code
to
a
container
image
as
soon
as
possible.
As
a
buildback
author.
My
goal
is:
I
have
this
other
language
ecosystem
or
some
other
functionality
that
I
want
to
provide
as
a
build
pack.
How
do
I
do
that
so,
like.
A
I
think
we
we
should
just
introduce
these
concepts
in
terms
of
like
the
solution.
So
like
we
start
with
the
problem
and
as
we're
going
through
the
steps
of
solving
it,
we
should
introduce
the
concepts
to
the
person
then,
rather
than
just
being
like
here's
all
the
concepts
and
then
come
to
the
to
stay.
B
I
think
I
think
it
needs
to
be
more
of
a
story
which
I
think
is
what
you're
saying,
and
I
think
it
needs
to
map
each
part
to
the
output
that
it
that
it
generates
right,
like
the
when
you
talk
about
the
text
phase
like
what
what
is
it
outputted
by
that,
and
why
do
I
care
when
you
come
to
the
build
phase
like
what
is
outputted
by
that?
What
do
I
care
and
then
you
kind
of
build
it
up
to
this
is
what
a
builder
is.
C
Okay,
so
I
have
a
maybe
in
that
same
vein.
Right,
like
I
feel
like
a
lot
of
these,
especially
for
app
developers.
They
are
problem
focused
right
so
like
if
I
say
I
want
to
specify
the
launch
process.
I
can
go
in
here
and
it'll
run
me
through
how
I
go
about
doing
that
right.
C
It
sounds
like
there's
something
at
like
a
high
level,
though
right
like
so
for
in
this
instance
on
this
one.
We
have
a
building
app
right
and
it's
pretty
straightforward
from
for
an
app
developer.
We
have
a
selected
builder,
select
your
app
and
then
just
go
for
build
pack
authors.
It's
literally
this
like
huge,
create
a
build
pack
thing.
C
So
I
do
wonder
if
maybe
for
build
pack
authors,
a
better
approach
would
be
to
give
them
just
a
very
tiny
taste
of
the
core
functionality,
which
is,
I
think,
what
what
dan
was
saying,
or
it
wasn't
gabe
yeah
it's.
These
are
the
different
phases
right.
This
is
detect.
This
is
build
and
that's
it.
So,
almost
just
like
these
two
steps.
C
C
Now
I
need
to
do
something
a
little
bit
more
elaborate,
and
so
then
we
get
onto
parts
of
creating
a
build
pack.
A
C
Did
you
all
ever
decide
what
to
do
with
like
the
whole
package,
lib
cnb
and
all
those
libraries.
A
Like
eclipsey
and
me
and
stuff,
I
so
we're
with
piccado
folks,
we've
been
trying
to
converge
on
v2,
which
combines
lip
cnb
with
packet,
just
waiting
for
like
reviews
from
emily
daniel
ryan.
A
A
C
So
it
sounds
like
the
way,
I'm
understanding
this
is
we're
using
bash
to
be.
You
know,
completely
agnostic
and
and
kind
of
just
show
you
what's
possible
through
file
system
manipulation,
essentially
but
you're
saying
it's
not
production,
quality
and-
and
I
would
definitely
agree
with
that-
and
it's
probably
not
the
easiest-
to
have
to
go
through
this
process.
A
C
The
part
where
maybe
I'll
rebut
a
little
bit
as
I
don't
know
that
we
need
production
quality,
build
pack
examples
in
the
documentation.
C
Typically,
a
lot
of
these
docs
would
just
scratch
the
surface,
and
if
you
want
to
look
at
something
production
ready,
you
would
go
look
at
actual
implementations
of
things
right,
so
you
could
go.
Look
at
the
potato,
build
packs
at
the
google,
build
packs
or
heroical
build
packs
for
production,
ready
stuff.
C
A
I
think
that's
also,
potentially,
what
makes
it
harder
for
people
to
graduate
from
this
simple
will
back
to
a
production
buildback,
they
don't
know
a
process
like
start
from
here.
Look
at
any
of
the
potato
stuff
you'll
be
completely
lost
same
for
heroku
they've
got
shims
in
there
to
get
me
to
e3
come
back
again.
We
lost
what's
happening.
A
What
about
the
google
ones?
I
have
never
looked
at
those
google
ones
have
like
basilisms
in
there
and
they
have
their
own
style
right.
C
Yeah,
so
I
mean
in
my
ideal
world
it
would
be
like
bash.
Is
nice
right?
It's
agnostic!
I
see
the
value
in
dash,
but
it
would
be
nice
thinking
about
it
from
a
new
build
pack
author's
perspective.
If
I
had
like
really
like
a
specific,
almost
like
go
dsl
right.
That
says
like
this
is
what
you
do
like
a
very
simple
api.
A
A
C
C
Yeah
yeah
yeah-
I
would
be
on
board
with
that.
I
obviously
I
don't
think
that
this
is
my
call
to
make
or
are
called
to
make
the.
I
think
that
the
part
where
I
think
it
makes
it
not
possible
to
do
at
this
point
in
time
is
that
there
is
no
concrete
gold
library
for
us
to
endorse
right
slip.
Sandy
is
it
then,
what
about
packet
or
packet
v2.
A
A
A
C
My
understanding,
with
libc
and
b,
is
that
it's
mostly
just
data
structs
in
a
really
a
really
funky
api.
C
The
the
thing
that
might
make
people
happy
but
definitely
would
increase
our
our
overhead
would
be
like
having
the
option
to
have
both
right,
which
I've
seen
it
in
some
projects
where
it's
like,
like
you're
looking
at
this
right
and
it's
like
bash,
but
then
you
click
and
it's
a
different
language,
yeah
and
then
it'll
switch
it
for
everything.
Essentially.
C
Well,
I
guess
what
I'm
thinking
is
like
okay,
let's
say
that
somebody
comes
in
and
says:
hey.
I
want
to
contribute
a
a
new
guide
for
build
pack
authors
that
is,
you
know
specific
to
whatever
right
and
like
okay
cool
go
ahead
and
do
that,
but
you
have
to
provide
it
in
bash
and
you
also
have
to
provide
it
in
go
and
you
also
have
to
provide
it
in
java.
I
don't
know.
A
C
Yeah,
if
we
could
get
a
tabbed
version
and
and
do
go
and
then
maybe
what
I
guess
this
is
how
I
would
strategize.
This
is
like,
let's
add
a
little
bit
of
go
in
here
for
me,
honestly
again,
I
go
back
to
the
preference
of
like
maybe
not
live
cmb,
but
something
a
little
bit
more
robust
and
user-friendly,
which
would
would
you
know,
be
the
packet
stuff
from
my
understanding
I
haven't
played
with
it,
but
I
wonder
if
I
don't
bring
some
concerns
yeah,
I
don't
honestly,
I
don't
think
it
should.
A
A
C
A
A
But
like
v1,
alpha
1
still
works
because
how
kubernetes
works
right
migration
yeah,
so
this
is
low
maintenance
and
it's
one
spec
and
it's
exposed
at
a
very
high
level
to
an
app
developer.
That's
a
library
that
people
will
be
using
with
lots
of
public
exposure,
different
functions
and
stuff.
So
it's
a
slightly
different
story.
C
But
I
guess
what
your
concern
is:
is
the
the
stability
of
the
api
that
we
would
be
using.
You
don't
think
that
they
would
adapt
something
like
sember
and
make
sure
that
it
doesn't
break,
because
that
would
just
be
you
know
I
I
can
make
these
not
just.
C
Right,
I
think
it's
it's
the
opposite,
and
I
I
hear
the
concern
differently,
which
is
they
may
be
moving
in
such
a
a
pace
that
doesn't
make
it
reasonable
for
us
to
keep
up
with
them
right.
So
if
they
had
a
stable
api
right
and
they
produced
something-
and
you
know
it
was
golden
for
a
while-
it
was
lts
for
like
a
better
words
right.
Then,
in
that
case,
like
we
should
the
concerns
don't
exist.
C
A
A
They
have
sometimes
added
partial
support
for
certain
apis,
like
with
the
features
that
they
need
in
their
build
packs.
They're,
sometimes
just
not
needed
to
support
an
api
because
they
didn't
need
any
of
the
features
in
there.
So
like
let's
say
we
bump
up
a
build
pack.
Api
version,
the
bash
versions
of
today.
Now
you
have
to
go
and
update
the
packet
one.
How
do
you
do
that.
A
A
C
All
right,
I
know
we're
a
little
bit
over
time.
How
did
we
get
through
everything?
Do
you
feel
comfortable
with
the
answers
on
that?
I
wouldn't
get
too
much
concrete
stuff,
but
it
was
a
good
time.
A
C
Okay,
so
what
you're
saying
is
something
like,
let's
say,
bloomberg:
we're
supposed
to
provide
a
potential
like
aka
resource
to
help
out
with
some
of
that.
That's
what
we're
saying.
C
Yeah
yeah,
I
don't
know-
I
know
in
the
past
vmware
had
someone-
I
don't
know
if
you
ever
met
him
or
saw
his
name
around
sander
and
he
was
kind
of
helping
with
some
of
that
and
it
was
pretty
well
received,
because
I
think
we
all
know
that
we're
busy
so
yeah
there
will
be
a
lot
of
pushback.
I
I
will
also
add
this
shouldn't
be
private
because
there's
a
job
posting
out
in
the
public,
but
we're
getting
a
community
manager
so
like
we're
interviewing
for
community
manager.
C
That
will
be
part
of
the
build
packs,
io
pocato
and
maybe
some
of
the
keynative
initiatives.
So
so
I
think
they'll.
I
think
the
hope
there
will
also
be
that
they
may
help
on
the
cncf
front
and
a
lot
of
that
stuff
as
well.
Nice.