►
From YouTube: TAG General Meeting - 2022-09-21
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
D
E
F
Okay,
I
posted
a
link
for
the
agenda.
We
already
have
some
topics
on
the
agenda
right
now.
If
you
have
any
additional
topics,
obviously
feel
free
to
publish
sorry
to
add
them
to
to
the
list.
E
E
F
So
we
have
two
chairs,
One
tech,
lead
here
or
I.
Think
we
can
start
with
the
project
presentation
it's
going
to
be
recorded,
anyways
sort
of
let
you
wait
any
longer
yeah.
We
can
run
some
questions
here.
We'll
publish
the
meeting
later
on
as
well
and
I
would
also
advise
you
to
later
on.
Watch
the
Tag
app
delivery,
select
Channel
for
follow-up
questions,
as
some
people
joining.
D
No
worries,
let
me
actually
start
sharing
my
screen.
D
Are
you
guys
able
to
see
my
screen
cool,
okay,
hi,
everyone,
I'm
Praveen
driver
and
I'm,
one
of
the
maintainers
of
Carvel,
and
we
are
really
excited
to
be
here
today.
I
also
have
John,
Reno
and
Tomic
with
me
today.
Keeping
the
time
constant
in
mind.
I'll
get
started
with
the
stuff,
so
I'll
start
with
how
Carvel
Boss
by
cover
was
created
and
what
is
Carvel
so
usually
the
kubernetes
provides
all
the
configurations
and
machineries
that
we
need
to
get
our
workloads
running
in
our
cluster.
D
But
when
we
want
to
do
things
at
scale,
we
would
need
more
jobs
to
be
done.
So,
for
example,
with
with
more
resources
and
more
manifest
coming
into
picture,
we
would
need
to
extract
boilerplates
and
simplify
the
configuration
API
for
the
end
users,
and
this
is
often
done
by
templating
configurations
and
with
lay
the
threes
in
recent
times.
There
are
microservices
and
then
cachings
and
data
storage
Etc
with
these
requirements.
D
We
also
need
to
manage
a
lot
of
dependencies
and
that
that's
the
one
of
the
jobs
that
needs
to
be
scaled
and
then,
since
we
are
scheduling
workloads
into
containers,
we
would
also
need
images
to
paper
and
then
in
some
cases
we
would
also
need
our
software
to
be
distributed.
So
it
needs
to
be
packaged,
and
then
it
it
needs
to
be
easily
distributable
and
once
it
reach
reaches
the
destination.
D
We
would
also
want
to
configure
the
end
users
to
to
be
able
to
configure
those
deployments,
and
once
they
have
done
that
they
should
be
able
to
easily
deploy
or
upgrade
or
remove
certain
Resources,
with
confidence
so
and
to
scale
all
of
these
jobs
in
in
a
fairly
automated
way,
like
probably
using
guitars.
We
would
need
different
sort
of
tools
for
each
of
these
shops,
and
these
tools
should
also
be
like
really
composable,
so
that
users
can
choose
the
best
tool.
D
That
actually
is
supports
their
use
case.
So
Carvel
tools
actually
covered
each
of
these
shops
and
yeah
Carvel
is
basically
it's
it's
composed
of
like
a
bunch
of
tools,
so
a
tool
suit
and
it
it
is
based
inspired
by
the
Unix
philosophy
of
favoring,
thoughtful
collection
of
small
tools.
Each
of
them
is
focused
on
one
particular
job
and
it's
really
good
at
it,
while
remaining
like
completely
composable.
So
you
can
pick
one
over
the
other
I'll
start
with
the
first
rule,
Itt,
so
ytt
or
vamil.
D
Templating
tool
is
a
general
purpose,
sample
templating
and
patching
tool
and,
unlike
other
approaches,
ytt
operates
within
vammal.
So
it
elevates
the
author
from
like
doing
some
bookkeeping
of
indentation
and
encoding
of
values.
So
it
basically
sees
your
vaml
files
as
data
structures
rather
than
a
text.
So
it's
fairly
easy
to
like
not
get
issues
not
get
errors.
D
Certain
errors
and
authors
generally
begin
with
the
kubernetes
Manifest
and
it
reduce
the
overall
maintenance
and
extract
variables
or
in
some
scenarios,
chunks
of
paramil
or
even
like
whole
sets
of
files
into
functions
and
libraries
and
in
ytt
logic,
is
expressed
in
Python
dialect.
That
is
using
stala,
which
is
a
language
purposeful.
D
For
expressing
configuration
only
and
then
authors
can
also
Define
a
schema
for
input
values
as
precise
constants
to
ensure
values
are
well
informed
and
well,
and
there's
also
the
overlaying
the
patching
part
to
it
like
at
the
end,
once
we
have
distributed
the
packages,
users
can
also
do
some
patching
on
and
do
some
using
overlays,
so
they
can
easily
change
sort
of
things
according
to
the
requirements
at
the
end.
So
why
Diddy
has
like
kind
of
both
the
things
overlaying
and
templating
any
questions
so
far.
D
The
part
of
it
like
I
would
say,
like
ytt,
can
entirely
replace
them.
I'd
say
like
there
are
some
things
so,
for
example,
templating
you
can
use
help
templating
or
you
can
use
580
according
to
your
needs
and
then
next
up
you
can
use
vendor
or
you
can
use
something
else.
So
it's
entirely
composable.
D
Okay,
so
next
up
is
vendor.
Vendor
provides
like
a
means
of
declaring
the
contents
for
a
directory,
so
it's
you,
authors
can
like
specify
any
sort
of
resources
like
repositories
or
oci
images
or
HTTP
locations,
GitHub
releases,
Helm,
charts
or
even
like
other
directories
and
vendors
and
simply
synchronizes
the
contents
of
the
pointed
to
directory
with
the
directory
that
you
want
to
think
of
and
take
in
some
scenarios
when,
when
the
source
has
an
actual
versioning
like
a
git,
repo
Venter
will
remember
the
exact
version.
D
The
revision
fetch
to
ensure
repeatable
bills,
big
region,
we
generally
just
need
to
use
vendor
sync
and
I'll,
resolve
the
dependencies
and
sync
the
source
and
the
version.
D
Next,
stop
is
building
images.
Sorry.
A
D
D
Okay,
thanks
for
pointing
that
out,
John
yeah
now
for
building
images,
a
cable
actually
scans
kubernetes
manifest,
and
it
looks
for
image.
References
and
authors
can
configure
well-known
image.
Building
tools
like
Docker
or
KO
or
tack
and
cable
actually
relies
on
them,
and
it
ensures
that
an
image
is
either
fetched
or
built.
Then
cable
resolve
all
the
image
references
to
their
char-256
stages
and
then
it
kind
of
becomes
a
cryptographic
certainty
that
what's
running
in
production,
is
what
what
was
meant
to
be.
D
Okay,
next
is
image
package
by
the
way
to,
let
me
know
if
I'm
going
too
fast
or
too
slow.
D
Okay,
so
image
package
collects
kubernetes
manifest
and
references
to
all
the
button
images
into
like
a
single
artifact,
a
bundle,
and
then
these
bundles
can
be
copied
from
one
registry
to
another.
It
could
be
the
public
or
it
could
be
internal
and
then
use
can
also
produce
a
completely
self-contained
double
on
the
entire
system
of
the
entire
system
and
take
the
configuration
and
all
the
images
and
all
the
dependent
images
in
a
single
archive.
And
then
this
like
allows
us
to
like
handle
completely
I
kept
environments.
B
Is
that,
are
you
using
oci
format,
yeah.
D
D
That's
right:
next,
stop
is
cap
now
cap,
actually
super
charges
and
operator's
ability
to
apply
manifest
to
the
cluster
and
with
cap
users
can
best
basically
preview
a
precise
plan
of
the
changes
that
are
going
to
be
applied
to
the
cluster
before
they
are
applied,
and
it
it
also
establishes
a
lightweight
ordering
so
that
resources
have
like
the
natural
dependencies
are
deployed
first
and
then
the
needed,
and
then
the
actual
resources
are
deployed.
D
So,
for
example,
namespace
resources
are
deployed
after
namespaces
and
then
it
also
observes
a
live
update
of
the
current
state
of
the
deployment
until
all
resources
are
fully
reconciled.
So,
basically
we
get
to
see
all
the
Recon
resources
reconcile
and
be
sure
that
everything
is
deployed
and
we
can
also
Monitor
and
inspect
logs
of
the
pods
or
anything
which
are
which
is
present
of
in
that
application.
D
Well,
you
can
use
other
tools
with
cap
to
enable
github's
Behavior,
so
cap
would
actually
just
use
the
manifest
that
that
are
provided
to
it
and
then
deploy
it
on
the
cluster
based
on
the
current
state
of
the
cluster.
Now
you
can
use
other
tools
like
let's
say
just
I,
think
GitHub
has
oidc,
so
you
can
use
it
along
with
that
or,
as
you
might
have
seen
in
the
next
slide.
D
There
is
also
an
included
openated
integration
that
we
provide,
which
is
Cap
controller,
so
you
can
use
that
tap
controller
as
well.
I.
B
D
D
Yep
yeah
so
moving
on
to
the
next
tools,
such
as
cap
controller,
so
each
tool
like
what
we
have
seen
is
is
independent.
D
You
can
actually
use
one,
maybe
some
other
tool
over
one
of
the
cover
tools
or
maybe
just
use
one
that
you
actually
like
for,
like
whatever
works
best
for
you,
like
maybe
in
cicd
pipelines
original
scripts,
and
then
we
also
have
cap
controller
cap
controller
basically
provides
custom
resource
definitions
for
an
application
and
a
Carvel
app
basically
has
three
steps:
fetch
you
can
fetch
kubernetes,
manifest
from
wherever
you
want
to
and
then
template
with.
D
Whatever
method
you
would
want
to
and
then
deploy
the
Manifest
finally
onto
the
cluster,
and
then
this
this
step
is
something
that
has
repeated.
So
these
three
steps
are
like
repeated
in
whatever
duration
you
would
want
to
sell
and.
D
Yeah,
so
this
workflow,
basically
configures
each
required
tool
and
of
like
you
can
you
can
also
use
some
other
third-party
Tool,
because
we
do
have
a
large
number
of
tools
these
days
and
then
one
of
the
recent
additions
to
cap
controller
is
K
control,
which
is
a
CLI
native
to
cap
controller
itself.
So
it
it
basically
allows
our
users
to
like
either
authors
or
packages
or
consume
packages
which
are
already
there.
So
it
internally
uses
the
cap
controller
apis
itself,
but
provides
a
CLI
interface
to
users.
D
And
one
of
the
recent
additions
to
cover
tool,
so
it
has
secret
gen
controller,
which
is
again
another
controller
and
it
it
actually
makes
managing
Secrets
like
with
a
lot
of
ease.
So
it
provides
on
clusters,
secret
generation
or
sharing
secrets
between
various
name
spaces
and
then
content
secret
content,
shaping
assembling
secret
contents
from
like
other
resources
and
then
since,
like
this,
functionality
is
also
exposed
by
scr.
So
it
can
also
be
used
with
Cube
CTL
cap
or
Helm
or
arcosity
flux,
CD
or
even
cap
controller.
D
A
No
I
saw
Eloise.
Did
you
have
a
question
about
cap
controller.
F
I
think
the
only
I
think
I
think
I
understand
the
end-to-end
process.
The
only
question
I
have
is
there
was
a
tool
when
you
like
go
into
image
packaging.
That
also
would
allow
me
to
just
automatically
upload
everything
from
that
package
back
to
an
oci
registry
as
well.
F
That
I
kept
building
my
images
and
then
I'm
packaging.
Everything
up
in
itself
contained
entity
which
we
could
probably
also
use
for
air
gap,
type
of
environments
and
then
I'm
sending
them
to
you
Josh,
for
example,
and
you
don't
want
to
obviously
import
everything.
That's
there
and
that's
part
of
this
process.
You
need
to
push
those
images
to
your
oci
registry,
so
whether
there's
there's
also
something
that
image
package
would
allow
me
to
do
as
part
of
the
distribute
step.
Yep.
A
You
can
copy
so
from
registry
to
registry.
You
can
pull
from
a
registry
to
local
either
just
the
manifests
and
a
bill
of
materials
of
the
images
or
you
can
grab
the
whole
thing
as
a
tarball
from
a
registry,
or
you
can
push
so
typically,
typically
in
the
in
that,
in
that
early
build
step,
you'll
your
build
will.
Actually,
the
images
will
typically
get
pushed
up
to
a
registry
right
away.
A
I
mean
if
there
was
only
one
little,
the
only
Nuance
I
can
think
of
to
add
here,
proving
those
really
good
is
just
that.
Those
light.
Blue
boxes
are
all
cli's,
I'm
I,
don't
know
if
that's
blatantly
obvious
and
that
the
the
controllers
were
are
the
the
in
cluster
machinery,
and
the
idea
is
that
you
don't
have
to
use
ytt
to
use
that
you
could
like
to
use
different
parts
of
the
thing.
A
Cap
controller
directly
supports
Helm,
templating,
fetching
Helm,
charts
and
GitHub
releases
Etc,
there's
a
lot
of
stuff
that
it
can
pull
in
and
it'll.
Allow
you
to
template
with
Q
as
well
as
a
more
recent
ad
and
Our
intention
is
to
add
more
of
those
Integrations,
so
that
that
opinionated
integration
in
cap
controller
is
something
that
you
can
bring
more
and
more
of
your
own
tools.
A
But
the
idea
is
to
kind
of
make
space
for
hey
if
there's
something
that
you're
already
doing
that,
you
can
take
advantage
of
how
those
individual
jobs
are
solved
well
by
some
tool
and
carbo.
It's
meant
to
deliberately
not
be
an
all
or
nothing
thing,
but
you
have
to.
If
you
do
that,
you
kind
of
have
to
provide
an
out
of
the
box
integration
of
all
the
things
in
case
someone
wants
to
just
get
started.
A
Is
that
yeah
there's
there's
a
couple
of
things
there's
like
there's
like
an
install
script
that
gets
all
the
tools
on
your
machine.
There's
GitHub
actions
setup.
So
if
you
want
to
do
it
in
a
GitHub
action,
you
can
get
a
a
pre-built
container
with
that
stuff.
D
B
A
We
haven't
yet
we're
still
working
with
talking
with
that
that
maintainer
to
like
see
if
we
can
get
that
as
part
of
the
donation,
but
I'm
not
involved
in
I,
just
know
that
we
haven't
yet
brought
it
fully
under
the
umbrella.
But
Adrian's
like
maintains
it.
F
Yeah
I
think
it's
cool
like
also
I,
just
split
it
up
into
smaller
tools.
I,
really
like
the
apologies,
as
you
mentioned,
like
the
Unix
parallel
mechanic,
use
individual
tools
for
individual
steps
in
my
reward
process,
I'm
just
wondering
like
from
Carvel.
Overall,
that's,
maybe
a
bit
more
of
a
10
000
feet
type
of
questions,
because
sometimes
we
have
to
approach
the
presenting
like
a
lot
of
different
tools.
A
Yeah,
that's
a
good
question.
I
mean
I,
think
I
think
it
fits
all
within
like
this
workflow
I,
don't
mean
to
be
like
that
that
I
guess
I
guess
that's
something
that
I
mean
you
would
take
on
a
Case
by
case,
but
let's
see
what
what's
something
that
wouldn't
be
a
Carvel
thing:
it's
really
about
application,
packaging
and
deployment.
A
So
if
there's
something
that's
outside
of
that,
I
think
maybe
one
one
piece
in
here-
that's
not
well
represented
yet
is
like
the
developer
workflow.
A
So
you
can
use
stuff
like
this
within
some
of
those
inner
loop
developer.
Loop
stuff,
probably
haven't
explored
that
as
deeply
as
we
wish.
We
could
yet
actually.
B
I
I
kind
of
this
kind
of
a
good
segue
into
what
I
want
to
discuss
a
little,
which
is
around
app
platform
like
I.
Look
at
this,
like
you
put
all
these
pieces
together,
maybe
a
little
bit
more,
but
then
you've
got
what
we're
talking
about
as
a
as
a
full,
app
platform
that
you
know
all
your
devops
and
folks
can
all
build
on.
F
F
F
What
you
want
to
be
and
what
you
don't
want
to
be
is
kind
of
helpful
for
people
when
they
look
at
it
like
how
does
it
fit
into
the
overall
ecosystem
and
saying,
like
you,
won't
add
a
service
mesh
most
likely
tomorrow,
and
it
would
like
be
a
number
of
like
first
implementation
and
you
don't
already
and
then
incubation
stage,
and
somebody
would
have
something
incubated
in
there
and
different
and
like
newer
projects
that
that
you're,
adding
at
least
having
some
some
ideas
on
how
to
present
this
I
think
is
useful.
F
But
I
think
that
if
you
said
it's
a
package
and
deploy
workflow,
that
makes
sense,
and
it's
also
very
natural-
for
a
project
to
grow,
especially
two
focused
ones,
and
so
just
like
I
said
as
a
hint
as
you
might
be,
as
you're
submitting
it.
Thinking
about.
Okay,
like
what's
like
in
scope
and
what
you
consider
out
of
scope
is
very
often
helpful,
like
just
making
it
explicit.
I
know
you
have
like
every
everywhere
assume.
You
have
a
very
good
idea.
What
you
want
to
see
in
scope
and
out
of
scope
here.
F
I
think
Mighty
would
fit
I
think
that
that's
what
the
users
of
projects
will
figure
out
themselves
like
when
they
want
to
use
certain
tablets.
But
if
you
say
like
this
is
a
cncf
projects
and
suddenly,
like
you're,
adding
entirely
different
things,
you
want
to
have
at
least
some
understanding.
What
does
this
project
actually
stand
for?
Is
it
just
a
collection
of
tools,
or
do
they
have
like
a
bigger
idea
of
why
the
individual
tools
exist?
F
Obviously,
projects
evolve,
but
it
it's
just
just
helpful
to
understand
it
like
not
to
name
specific
approaches,
but
there
were
projects
that
were
sometimes
considering
of
several
sub-projects
of
different
tools
at
a
different
majority
level
and
also
like
having
different
use
cases,
and
you
were
like
wondering
like
how
should
you
judge
like,
for
example,
adoption
of
this
or
maturity
of
it
all?
If
you
keep
adding
tools,
they
they
do
very
different
things
and
so
forth.
F
I
think
it's
just
helpful
for
somebody
actually
reviewing
your
proposal
to
like
have
a
good
understanding
of
where
you
want
to
take
this
and
just
being
explicit
about
it
that
they
say
now
we're
not
adding
this,
and
this
is
like
the
focus
so
Josh
I
hope
this
answers.
Your
question
as
well.
Yeah
I,
think
we're
ready
to
put
it
on
the
landscape
might
be
good
analogy
like
can
you
identify
that
one
space
in
a
landscape
where
you
would
put
it,
and
rather
than
the
case,
it's
like
a
collection
of
500
tools
to
do.
A
Yes,
I'm
hearing
there
that
there's
this
risk
that
you've
seen
in
the
past,
where
projects
don't
have
focus
and
Clarity
of
their
Charter
and
then
it
becomes
difficult
for
anyone
to
understand.
Well,
what's
inside
the
Box,
you
got
a
label
on
the
outside,
but
why
would
I
even
open
it
up
and
what's
in
there,
and
if
it
becomes
a
mishmash
of
of
solutions,
then
it
doesn't
help
out
the
community.
It
makes
it
confusing.
F
It
would
be
totally
obvious
what
you're
doing,
but
you
like,
adding
a
lot
with
deployment,
you're,
adding
stuff
around
templating
and
so
forth,
like
making
the
picture
wider,
which
makes
a
lot
of
sense.
I
would
just
say,
like
everything
you
need
from
like
packaging
to
deploying
an
application
that
would
be
super
helpful
but
like,
rather
than
making
it
one
big
project,
it's
a
pick
and
choose
approach
of
the
set
of
individual
tools.
F
That
would,
for
example,
help
me
or
Harry
understood
your
presentation
today
to
see
okay.
Where
should
I
put
this
project.
F
Begin
with
and
then
the
direction
yeah
think
about
it.
That
way
you
have,
but
you
will
obviously
have
or
already
have
at
some
point
a
governance
committee
that
and
will
have
a
charter
of
what
the
approach
would
do.
What
would
this
Charter
look
like?
How
would
a
governance
committee
decide
what
you
put
in
there
and
what
is
the
out
of
scope
for
your
project
again,
not
required,
but
I
think
like
just
helpful
overall
as
you
as
you
build
the
approach
and
help
people
understand
what
you
want
to
have
where
you're
headed
with
the
project.
E
B
And
there
were
a
little
over
time
on
this,
but
I
I
just
wanted
to
ask.
How
do
you
see
your
relationship
with
tanzu?
Is
this
I
mean
I,
assume
it
works
on
all
kubernetes
is,
is
it?
Is
it
like
tightly
coupled
with
tanzu,
like
yeah,
just
wondering
if
it
would
work
on
its
own
nicely,
and
you
know
vanilla,
kubernetes
and
your
thoughts
on
that.
A
Yeah,
it's
probably
probably
more
accurate
to
describe
it.
The
other
way
around
that
tanzu
is
built
on
Carvel
in
a
sense
like
all
the
all
of
like
the
component
delivery
is
done
using
this
tooling,
and
so
something
that
you
see
explicitly
like
if
you
go
and
look
at
like
their
install,
docs
and
stuff
you'll
see
mentions
of
these
tools.
A
Other
pieces
have
been
facited
to
try
and
improve
sort
of
a
more
simplified
experience,
because
it's
a
different.
It's
a
different,
a
slightly
different
segment.
Folks,
who
part
one
one
big
part
of
the
the
underlying
Philosophy
for
Carvel
from
the
beginning,
has
been
not
to
get
into
the
way
of
you
understanding
the
kubernetes
Machinery
to
gently.
Introduce
you
to
it.
A
So
you
see
it,
you
don't
have
to
understand
all
of
it
in
order
to
to
use
Carvel,
but
it
the
more
that
you're
exposed
to
it.
The
more
likely
you're
going
to
develop
your
mental
model
of
what's
Happening
under
the
covers.
A
It's
your
choice
to
go
deeper,
but
with
with
a
commercial
product,
that's
aimed
more
toward
trying
to
present
higher
level
abstractions
like
tonsu
is
there
there
might
be
some
layers
on
top,
but
it's
absolutely
independent
of
talking.
B
B
F
Yeah,
maybe
out
of
curiosity
from
icon
preview
to
base
who
who
outside
of
VMware,
do
you
have
as
like
maintainers
on
the
project,
or
do
we
have
maintenance
outside
of
VMware.
A
Yeah
we
have
a
smattering
of
folks
from
the
community
and
we've
just
started
doing
some
work
in
Earnest
with
twilio
we're
just
in
the
beginning
of
making
the
project
much
more
open
and
easy
to
contribute
to
it's.
Actually,
one
of
our
big
items
on
our
roadmap
is
to
make
contribution
much
simpler.
A
That's
where
we're
at,
but
we're
very
Community
oriented.
So
we
we
we're
proud
of
being
very
responsive
on
slack
and
and
GitHub.
When
the
shoes
come
up
and
whatnot,
we
try
and
help
people
get
involved.
However,
deeply
they
want
to
we've
like
paired
with
people
as
they're
doing
contributions
at
times.
So
there's
the
individual
intention,
but
there
needs
to
be
more
systemic
support
and
we
know
that.
F
E
F
Mean
if
there
are
any
are
any
other
questions.
Otherwise,
you
can
also
continue
on
it's
like
after
we
share
it
with
other
people,
we're
taking
timing.
F
F
B
Yeah
I'm
the
main
problem:
working
collaborative
delivery
right
now,
yeah
I,
think
adjustment
by
the
way
for
looking
over
that
article
and
anybody
else
that
did
so
yeah
I
just
wanted
to
get
eyes
on
this.
This
is
kind
of
summarizing.
You
know
where
we
are
for
better
and
worse
with
Cooperative
delivery.
It's
been
a
lot
me
and
Robert
and
palaces.
Thank
you
for
joining
us
and
a
couple
other
folks,
so
I
kind
of
wanted.
B
To
sum
up,
you
know
what
we've
done
in
the
past
six
to
12
months
or
what
we've
learned
you
know
with
our
we
haven't
had
as
many
user
interviews
as
we
might
have
wanted,
but
a
lot
of
learning
is
coming
from
conversations
and
then
that
the
projects
actually
so
there's
kind
of
in
this
article
I
just
want
to
make
sure
everyone's
okay
with
it
I
just
basically
put
down
three
kind
of
ideas
that
I
think
come
out
of
the
projects
that
we
work
with
and
reflect.
Some
of
that
and.
B
Yeah,
if
everyone
is
you
know,
I
want
to
take
some
more
input,
but
if
everyone
is
is
generally
okay,
I'm
thinking,
we
publish
this
on
our
tag,
blog,
which
I
don't
know.
If
everybody
even
knows
about,
but
we're
working
on
that,
but
also
the
cncf
blog,
because
our
tag
blog
is
still
you
know,
we're
still
ramping
that
up
and
then
the
last
thing
which
I
put
in
the
notes
Here,
is
that
I'm
kind
of
thinking
of
using
this
I.
B
F
Yeah
I
think
my
comment
was
just
as
we're
kind
of
like
revamping
it
and
pushing
it
out
to
The
Wider
Community,
whether
we
should
really
pick
a
name
that
people
understand
like
historically
the
people
who
were
like
in
favor
of
Cooperative
delivery
are
not
even
part
of
that
idea.
Anymore
and
at
least
I've
always
run
into
a
situation
where,
when
I
talk
about
Cooperative
delivery,
people
don't
actually
know
what
you're
talking
about.
E
F
I
think
we
just
have
a
very
confusing
name
for
something
that
everybody
understands
if
it
would
just
call
it
slightly
differently,
because
if
we
talk
about
composable
development
platforms
or
delivery
platforms,
whatever
we
want
to
call
them,
I
think
everybody
understands
what
we
mean
from
the
from
from
the
get-go.
That's
that's
why?
I'm
still
struggling
with
with
the
name
out
of
experience,
I
was
never
the
biggest
fan
to
be
fair,
like
in
full
transparency,
but
I
really
figured
it
out.
F
Whenever
I
was
explaining,
it
people
didn't
get
what
it
was
about,
but
then
I
was
talking
about
composable
delivery
platforms
and
everybody
said
Oh.
You
mean
like
compiling
tools
together
to
build
a
delivery
platform
that
developers
can
use
and
attack.
You
see
it
almost
flowing
out
of
people.
That's
why
I
was
kind
of
wouldn't
it
be
a
good
time
to
maybe
rename
it.
B
F
Mean
otherwise
you
would
you
read
the
blog
post
about
composable
Cloud
native,
develop
delivery
platforms,
or
would
you
read
one
on
Cooperative
delivery,
yeah
I
think
we
might
need
to
end
up
using
the
first
two
paragraphs,
just
doing
exactly
the
explanation
which
we
have
been
doing
for
almost
like
the
last
year
and
also
leaves
the
opportunity
there
are
lots
of
people
out
there
building
their
their
their
platform,
their
engineering
platforms
or
an
action
delivery
platforms.
F
I
mean
we
just
talked
about
Carvel
and
obviously
tanzu,
which
is
sitting
on
top
of
it,
which
basically
goes
in
that
direction
as
well.
I
think
that
that's
a
that's
a
topic
that
people
are
already
dealing
with,
that
we
could
simply
attach
to
it
at
the
end
of
the
day.
It's
what
we're
doing,
rather
than
introducing
a
new
term
that
nobody
is
actually
familiar
with.
F
Even
if
it
intellectually
sounds
right,
it
might
be
easier
to
have
that
to
push
it
in
that
direction,
that
that's
just
why
I'm
bringing
it
up
right
now
and
as
we
release
it,
the
blog
post,
which
I
think
is
great
I,
think
it
could
help
with
attraction
to
to
tell
people
what
we
actually
want
to
do.
F
Not
wanting
to
delay
anything
here,
but
if
we
could
like
agree
on
the
name
and
that
in
that
area,
I
think
it
would
be
way
more
or
way
easier
to
digest.
Unless
somebody
really
thinks
that
Cooperative
delivery
is
what
we
want
to
introduce
to
the
wider
audience.
I
mean
I
have
a
strong
opinion
here,
but.
B
I
think
the
product
as
a
platform
says
that
you
know
that
oh
sorry,
platform
as
a
product
kind
of
says,
Cooperative
delivery.
It
says:
deliver
your
platform
as
a
product
incorporate.
You
know
be
aware
of
your
customers,
so
yeah
the
term
platform
I
think
for
sure
belongs
there.
What
kind
of
platform
we
could
probably
debate?
I
heard
you
proposable
delivery,
always
baby
I
won't
bike
shed
too
much.
F
B
That
to
me
that
I
don't
want
to
blog
the
meeting,
but
that
is
still
a
little
bit,
not
the
term.
It's
a
little
like
intellectual,
like
I,
know
why
you're
saying
that,
but
it's
I
I
to
me
what
you're
saying
is
you're,
bringing
all
these
capabilities
from
infrastructure
and
underneath
together,
you're
composing
them,
maybe
I
like
app
what
happened.
You
know.
I've
mentioned
this
in
the
chat,
but
like
a
slight
difference
on
that.
Is
it's
reflecting
the
end
goal
like
we're?
B
F
C
I
think
it
I,
like
I,
like
George,
said
you
to
have
to
have
to
blog
post
as
a
moralist
as
a
start
of
the
of
the
of
the
restart
of
the
work
group
so
rename
and
chose
to
use
it
to
rename
it
and
and
for
the
consequence
to
to
drive
it
a
bit
more
or
to
get
more
Direction.
B
Okay,
I'm
going
to
take
a
action
item
so
to
to
get
that
integrated
in
there
I
think
we
all
agree
that
platform,
let's
let's
debate
a
little
and
slack
and
finalize
the
the
adjective
or
whatever,
also
FYI,
I'm
meeting
with
Luca
galante
today,
how
I
quoted
in
that
article-
and
he
runs
platformengineering.org
I-
just
figured
get
some
of
his
thoughts
too.
So
I'll
share
those
if
they're,
if
they're
relevant,.
F
C
Yes,
I
can
so
at
least
I
can
start.
So
we
are
very
well
present
on
the
on
the
North
America.
We
have
one
one
project
meeting
on
the
25th
of
of
October
in
in
the
afternoon,
and
at
the
moment
we
are
looking
for
a
project
such
as
Carver
who
want
to
present
there
so
for
a
15-minute
talks
and
so
on
about
it
could
could
be
their
projects
or
the
things
they
are
working
on
just
to
give
people
and
users
and
ideas
what
what
they
they
get.
Delivery
is
about.
C
So
at
the
moment
we
have
a
cfp
open
I
think
we
can
post
a
link
somewhere
here.
C
C
C
So
we
will,
we
will
be
there
for
for
some
hours
every
day
and,
furthermore,
projects
will
also
have
the
possibility
to
present
their
their
projects
there,
if,
if
they
don't
have
their
own
pools,
so,
for
instance,
also
looking
at
Carver.
If
you
don't,
if
you
reach
the
sandbox
Pro
level
in
the
last
few
months-
and
you
didn't
have
the
possibility
to
get
your
own
booth,
you
could
you
could
simply
reach
out
to
us,
and
we
could.
We
could
schedule
some
some
presentation
there
and
just
to
also
promote
your
project.
C
It's
also
about
end
user
stories.
So
if
someone
of
you
has
a
cool
continuously,
has
a
cool,
app
delivery
solution
and
wants
to
Showcase
this.
You
are
welcome
at
every
time
in
at
our
project
meeting,
but
also
at
our
booth.
If
you
want
to
stay
directly
directly
on
YouTube,
did
I
forget
something
always.
F
F
Also,
if
you
want
to
bring
other
topics
to
discussion
that
you
might
be
interested
in,
feel
free
to
do
so,
and
everybody
deliberately
looked
into
the
project,
you
usually
have
a
bit
of
a
harder
time
to
be
present
at
cubecon
and
to
have
discussions
and
then
well
that's
a
good
idea,
like
we
have
like
this
entire
four
hours
that
we
revealed
there
and
plus
to
boost
and
giving
people
some
more
opportunities
to
do
something
there
again.
F
If
you
want
to
bring
some
of
your
end,
users
who
have
interesting
talks
feel
free
to
do
so
as
well.
So
we're
trying
this
the
first
time
this
year
also
provide
us
feedback,
what's
valuable
to
you.
What
what's
is
working?
What
is
not
working
for
you,
we're
all
learning
here,
but
it's
a
slightly
different
approach
in
the
past,
people
are
like
only
talking
about
okay,
what's
happening
in
Tech
app
delivery.
F
This
time
we
want
to
write
more
of
a
discussion
more
of
an
exchange
between
projects,
because
we've
obviously
hasn't
seen
that
a
lot
of
projects
in
the
cncf
work
very
independently.
Some
of
them
do
collaborate,
but
also
see
this
as
an
opportunity
to
engage
with
other
projects
to
collaborate
and
maybe
get
some
collaborative
discussions
going.
F
It
will
be
very
competitive
to
be
fair
with
kubecon
like
there
will
be
a
lot
of
events
going
on
at
the
same
time,
so
no
promises
they
will
that
there
will
be
hundreds
of
people,
but
what
we
actually
learned
yeah,
you
never
know
it
might
also
that
a
lot
of
people
show
up.
But
what
we
have
seen
in
Valencia
for
booth
is
a
lot
of
people
show
up.
Actually.
F
So,
if
you
think
of
presenting
at
the
booth,
usually
a
lot
of
people
stop
by
more
than
probably
in
the
presentations,
because,
yes
yeah,
it's
a
very
packed
agenda,
as
you
might
have,
as
you
might
know,.
A
Well,
Thomas
said
we're
definitely
planning
on
being
there.
We've
got
a
couple
folks
speaking
at
get
Ops
con
as
well,
and
we're
we're
wet
behind
the
ears
I
think
we
were.
We
were
accepted
as
a
Sandbox
project
last
week,
so.
A
Every
win
well
definitely
excited
to
be
part
of
the
community.
I
think
that's
the
big
part
for
us,
but
yeah.
We
we,
we
were
looking
part
of
what
we
were
hoping
to
find
out
today
was
like
what
was
the
intent
around
those
projects.
It
sounds
like
we
have
some
stuff
that
we
can
share.
That's
pragmatic
in
in
in
that
project
meeting
and
excited
to
be
there
and
yeah.
We'll
talk
about
pulling
something
together
for
a
demo
for
the
the
presentation
for
for
the
booth,
I.
F
B
F
B
F
I
think
we
will
review
already
along
the
way,
but
that's
maybe
the
last
one
when
you
can
submit
yeah
or
we
say
Friday
the
7th
and
then
probably
extends
because
people
who
will
present
well
for
50-minute
presentation.
People
don't
need
to
prepare
that
much
to
be
fair,
yeah
I
think
we
can
make
the
16th.
F
E
F
Yeah,
thanks
for
the
presentations
today
very
insightful
and
again,
if
you
have
a
brad,
Josh's
blog
post,
do
so
and
also
provide
feedback
to
him
if
you
want
to
so
he's,
always
happy
to
get
feedback.
Yes,.